Copy of final_distrib_inv by zhangyun

VIEWS: 3 PAGES: 117

									                                                                                                   COLLECTION
                                   CLASSIFICATION                                                                          POLITICAL
                                                     YEAR(S)                   DEPENDENT         INSTRUMENT FOR                                OTHER INDEPENDENT
   AUTHOR      YEAR     COUNTRY    (FROM REVIEW                 UNITS (N)                                                INDEPENDENT                                      FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                    ANALYZED                    VARIABLE            DEPENDENT                                      VARIABLES
                                      ESSAY)                                                                               VARIABLES
                                                                                                     VARIABLE

Akhmedov      2004    Russia      Political         1995-2003 region-       budgetary           Goskomstat (State    months to election,                                Overall, strong evidence of sizable
and                               business cycles             month         expenditure,        Committee of         incumbent term,                                    opportunistic cycles in fiscal policies and
Zhuravskaya                                                   (6767)        budget revenue,     Statistics),         dummy for                                          no evidence of cycles in economic
(2004)                                                                      growth inflation,   Ministry of          incumbent governor                                 growth, in particular high spending occur
                                                                            regional            Finance of the       is supported by the                                a 1-2 months before elections. Shift of
                                                                            budgetary wage      Russian Federation   Communist coalition                                public expenditures to what is most
                                                                            arrears, wage                                                                               visible to voters (1318) "Total budget
                                                                            level, and                                                                                  expenditures experience the first
                                                                            income                                                                                      significant jump up of 7 percent nine
                                                                                                                                                                        months before elections. After that there
                                                                                                                                                                        are no significant changes until one month
                                                                                                                                                                        before elections when expenditures rise
                                                                                                                                                                        significantly by 13 percent." (1316)
                                                                                                                                                                        "Magnitude of the cycle decreases with
                                                                                                                                                                        education, urbanization, level of
                                                                                                                                                                        democracy, transparency of the
                                                                                                                                                                        government, and freedom of media."
                                                                                                                                                                        (1334)

Alesina,     1999     Italy       Redistributive    1995        individual log of hourly        Bank of Italy                                high school degree,        "In summary, the North redistributes to
Danninger,                        consequences                  s (23924) earnings of full      survey 1995                                  college, year work         the South in a way that creates negative
and Rostagno                                                               employed                                                          experience, female,        cultural and social side effects. But, if this
(1999)                                                                     workers                                                           married mid                is the case, why is this redistributive
                                                                                                                                             management                 system chosen? The answer may be that
                                                                                                                                                                        redistribution through public employment
                                                                                                                                                                        is less visible than direct transfers,
                                                                                                                                                                        therefore it is politically less costly and it
                                                                                                                                                                        is more effective at creating patronage and
                                                                                                                                                                        political benefits for local politicians." (4)
                                                                                                                                                                        "we see that public sector wages are not
                                                                                                                                                                        statistically different between the South
                                                                                                                                                                        and the North. On the contrary the results
                                                                                                                                                                        for the private sector are quite
                                                                                                                                                                        different….Southern residents earn on
                                                                                                                                                                        average about 18.9 less than their northern
                                                                                                                                                                        counterparts." (14)



Alperovich    1984    Israel      Political returns 1976 and    localities general              "Local Authorities   percent of residents    population, ratio of the   "As expected, the coefficient of POL is
(1984)                                              1978        (52)       government           in Israel -          who voted for parties   number of people 65        positive, highly significant, and extremely
                                                                           participation in     Financial Data"'     which formed the        years old or above and     similar for both governments. A
                                                                           budgets of local     and "Local           government              14 years old or below to   significant coefficient confirms the basic
                                                                           authorities          Authorities in                               the number of people       hypothesis that in forming their policy
                                                                                                Israel - Physical                            aged 19 to 65, per         towards local authorities, governments are
                                                                                                Data"                                        capita annual deficits     responding, among other factors, to the
Ames          1995a   Brazil      Programmatic      1987-1988             voting in             Constituent          spatial distribution of constituency attributes,   "deputies receiving of the
                                                                                                                                                                        political orientation pork benefits voted to
(1995a)                           versus pork                   municipal Constituent           Assembly, Kinzo      electoral support       career paths of            weaken the legislature and strengthen the
                                                                ity (403) Assembly              (1989)               (dominance and          politician (local,         executive, and tended to oppose statism-
                                                                          (Congressional                             clustering), seniority business, bureaucratic),    welfarim, and popular democracy….In
                                                                          Power, Support                             and electoral                                      sum, pork buys deputies." (339) Deputies
                                                                          for Executive,                             insecurity, state unity                            with clustered vote bases supported
                                                                          Statism-                                   and state interest,                                congressional power and statis-welfarist
                                                                          Welfarism,                                 political party, pork                              issues. Wealth and industry had
                                                                          Support for                                barrel                                             antiexecutive, economically statist-
                                                                          Popular                                                                                       welfarist, supportive of popular
                                                                                                 COLLECTION
                                  CLASSIFICATION                                                                        POLITICAL
                                                    YEAR(S)                   DEPENDENT        INSTRUMENT FOR                              OTHER INDEPENDENT
   AUTHOR    YEAR     COUNTRY     (FROM REVIEW                 UNITS (N)                                              INDEPENDENT                                      FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                   ANALYZED                    VARIABLE           DEPENDENT                                    VARIABLES
                                     ESSAY)                                                                             VARIABLES
                                                                                                   VARIABLE

Ames        1995b Brazil        Swing versus       1986-1990 deputies      whether deputy    Joint Commission municipal                  distance from               "The higher the percentage of a
(1995b)                         core                                       submitted         on the Budget    dominance (vote            municipal center            municipality's votes a deputy won in 1986,
                                                                           budgetary         (432)            share)                     (spacial center of vote),   the more likely that deputy was to pursue
                                                                           amendment to                                                  distance from personal      more support in the same place in 1990.
                                                                           particular                                                    center (where candidate     The negative slope on the squared term
                                                                           municipality in                                               got most of his votes),     means that deputies at some point regard a
                                                                           1989 and 1990                                                 percent of vote to          municipality as 'locked up', thus meriting
                                                                                                                                         retired deputies, percent   no additional effort...the actual inflection
                                                                                                                                         migrants, social            points were beyond nearly all cases"
                                                                                                                                         "match" variables on        (421).
                                                                                                                                         income distribution,
                                                                                                                                         government employees,
                                                                                                                                         population; interparty
                                                                                                                                         fragmentation,
                                                                                                                                         intraparty
                                                                                                                                         fragmentation, rank in
                                                                                                                                         party list, career
                                                                                                                                         variables




Ariga (2005) 2005   Australia,  Electoral system  Australia    district- electoral           original dataset                                                        "electoral cohesiveness is higher under the
                    Ireland,    impact            (1969-83),   level or cohesiveness (a                                                                              electoral rules without intra-party
                    Finland,                      Ireland      candidate- measurement of                                                                             competition [Norway, Australia, the U.K.]
                    Japan,                        (1973-89),   level      last margin that                                                                           than under those with it [Ireland, Japan,
                    Norway, the                   Finland                 depends on                                                                                 Finland]." (27)
                    United                        (1995-                  electoral rules                                                                            "Time series of the estimated electoral
                    Kingdom                       2003),                  which may                                                                                  cohesiveness under electoral systems
Arriola     2009    Africa                        Japan
                                Political returns 1960-99      40         include intra-
                                                                          time to an         Africa South of the cabinet size, polity,   GDP, oil exporter,          without intra-party competition (Australia
                                                                                                                                                                     "Recruiting more ministers into the
(2009)                                                         African     extraconstitution Sahara              executive years in      mineral exporter, ethnic    cabinet is an effective coup-inhibiting
                                                               states      al change in the                      power, civil war        fractionalization,          strategy for the leader who fears being
                                                                           chief executive                                               French colony,              overthrown. Each additional cabinet
                                                                           (regime duration)                                             Urbanization, trade         appointment lowers coup hazard by 23%
                                                                                                                                         (%GDP),                     to 25%, with all else being equal,
                                                                                                                                         population(log)             depending on the regime variable used in
                                                                                                                                                                     the model." (17)
Arulampalam 2009    India       Swing versus       1974–75 to major        grants (central   Reserve Bank of    swing dummy,             annual rainfall, state      "A state that is aligned and swing in both
, Dasgupta,                     core               1996–97    states       plan scheme and   India Bulletin     alignment dummy          population, proportion      the last Vidhan Sabha and the last Lok
Dhillon, and                                                  (14)         centrally         1974-1996,                                  of state population         Sabha elections is estimated to receive
Dutta (2009)                                                               sponsored         Finance                                     characterized as            19.6% (exp(0.179)−1) higher central
                                                                           scheme)           Commission,                                 scheduled caste, per        grants than a state that is unaligned and
                                                                                             National Accounts                           capita state of             non-swing with respect to both election
                                                                                             Statistics                                  agriculture in state        types. State that is aligned and swing in
                                                                                             (Government of                              domestic product,           last election receives more grants
                                                                                             India, Ministry of                          state's per capita          regardless of swing characteristics.
                                                                                                   COLLECTION
                                  CLASSIFICATION                                                                         POLITICAL
                                                    YEAR(S)                   DEPENDENT          INSTRUMENT FOR                               OTHER INDEPENDENT
  AUTHOR      YEAR     COUNTRY    (FROM REVIEW                 UNITS (N)                                               INDEPENDENT                                        FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                   ANALYZED                    VARIABLE             DEPENDENT                                     VARIABLES
                                     ESSAY)                                                                              VARIABLES
                                                                                                     VARIABLE

Baldwin      2005    Mali        Political         1992-1994 electoral     change in            L'Enquete          percentage of votes      income per capita,          "No support for the hypothesis that
(2005)                           favoritism                  districts     percentage of the    Malienne           obtained by Adema        unemployment, literacy      legislators distribute primarily to their
                                                             (42)          population with      Conjoncture        (political party) in     rate, urbanization level,   core supporters…little support for the
                                                                           government or        Economique et      the second round of      percent of population       theory that legislators redistribute
                                                                           public sector        Sociale            elections, difference    speaking various            primarily to swing districts" (20).
                                                                           jobs                                    between Adema's          subnational languages       "…greater support for the hypothesis that
                                                                                                                   vote share and 50%,                                  President Konare targeted swing voters.
                                                                                                                   turnout (all for both                                The coefficient on unemployment is
                                                                                                                   presidential and                                     positive" and coupled with a positive
                                                                                                                   legislative elections)                               coefficient on income, "something other
                                                                                                                                                                        than a benign attempt to reach the most
                                                                                                                                                                        needy is at work here" (23) "Because
                                                                                                                                                                        Konare won more than 50 percent of the
                                                                                                                                                                        second round votes in all of the regions in
                                                                                                                                                                        my sample, the finding that government
                                                                                                                                                                        jobs and presidential support [vote share]
                                                                                                                                                                        are negatively correlated is equivalent to
                                                                                                                                                                        the finding that more government jobs
                                                                                                                                                                        went to areas where the presidential race
                                                                                                                                                                        was closer" (23). Posits her own
                                                                                                                                                                        alternative explanation, beyond
                                                                                                                                                                        core/swing, that government jobs went to
                                                                                                                                                                        potentially destabilizing regions (24)




Banerjee and 2007    India       Political         1971, 1991 parliamen 1971-1991         Census of India     political             land inequality,                    "Most striking difference between 1971
Somanathan                       favoritism                   tary      change in the                         fragmentation, social scheduled tribes,                   and 1991 is gains registered by the
(2007)                                                        constitue fraction of                           fragmentation         scheduled castes,                   Scheduled Castes relative to other groups.
                                                              ncies     villages with the                                           Brahmans, Muslims,                  Higher shares of Scheduled Castes in
                                                                        public good                                                 Christians, Sikhs, Avg.             1991 are associated with more high
                                                                        (education/health                                           village population,                 schools, health centers, dispensaries, and
                                                                        /water)                                                     number of villages                  piped water." (308) Social fragmentation
Banerjee,    2005    India       Political               1991 districts share of villages 1991 Census of      Non-British dummy, geographic factors                     Non-British coefficient in 1991 (unlike
                                                                                                                                                                        has no systematic effectpositive for 14 in
Iyer, and                        favoritism                   (391 in   with a particular India               fraction of land not affecting public good                goods out of 26 and negatively significant
Somanathan                                                    16 major public good (26                        under landlord        transport ie. Average               for 4, nonlandlord effect positive for 15
(2005)                                                        states of categories)                           control,              annual rainfall,                    and negative for 3, and caste and religious
                                                              India)                                          fragmentation index maximum and                           fragmentation coefficient negative for 10
                                                                                                                                    minimum temperatures,               and positive for 4. "British areas do
                                                                                                                                    dummy for coastal                   significantly worse than non-British
Banful (2008) 2008   Ghana       Political         1994 to    district per capita         internal documents binary variable        districts, proportion of
                                                                                                                                    population, population              areas." (646)
                                                                                                                                                                        "DACF formula rules are followed in
                                 favoritism        2005       (approx District            from the            which is 1 if ruling density, ethnic                      calculating the districts' allocation,
                                                              1000-    Assemblies         headquarters of the government won the fragmentation, literacy                however, the fund that the districts
                                                              1300)    Common Fund District Assembly district in last               rates                               receive, their disbursements, is generally
                                                                       (DACF)             Common Fund in presidential election;                                         different from this amount…
                                                                       disbursement       Accra, Ghana        political cycle                                           Counterfactual allocations calculated
                                                                       logged, per                            variables including                                       suggest that the DACF formula was
Barkan and   1989    Kenya       Redistributive    1983-88    province District Focus for private             Distribution patterns                                     "the significance of District Focus as a
Chege (1989)                     consequences                 (7)      Rural              correspondence      of roads and health                                       mechanism to achieve a redistribution of
                                                                       Development        and Development (compared these                                               resources and power in Kenya is probably
                                                                       Fund (non-         Estimates for the figures to Rural                                            greatest in political rather than economic
                                                                       regression         Year 1987/88        Development Fund                                          terms." (450) "Allocations for roads and
                                                                       analysis)                              figures, non-                                             health, however, indicate that significant
                                                                                                              regression analysis)                                      redistribution of development expenditure
                                                                                                                                                                        requires the intervention of central
                                                                                                                                                                        ministries, because it is they which
                                                                                                                                                                        unilaterally allocate the most funds to the
                                                                                                                                                                        rural areas." (450)



Berman,       2008   Iraq        Miscellaneous     2004-2007 district      intensity of         "significant       spending by          district characteristics,       "spending on public goods is
Shapiro,                                                     (832 =        insurgent activity   activity"          Coalition forces on year indicators,                 unconditionally correlated with greater
Felter (2008)                                                104distric    (rate of attacks     (SIGACT) reports   small-scale          interactions of these           violence…. once we condition on
                                                             ts*8half-     per capita against   by Coalition       reconstruction                                       community characteristics, we find that
                                                             years)        Coalition and        forces, MNF-I-     projects through                                     greater service provision leads to less
                                                                           Iraqi government     SIGACTS III        programs, such as                                    violence. This violence-reducing effect
                                                                           forces)              Database           CERP, intended to                                    appeared in the second half of 2007, when
                                                                                                                   provide local public                                 operational changes meant that Coalition
                                                                                                        COLLECTION
                                   CLASSIFICATION                                                                                 POLITICAL
                                                        YEAR(S)                    DEPENDENT          INSTRUMENT FOR                                    OTHER INDEPENDENT
   AUTHOR      YEAR     COUNTRY    (FROM REVIEW                     UNITS (N)                                                   INDEPENDENT                                         FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                       ANALYZED                     VARIABLE             DEPENDENT                                          VARIABLES
                                      ESSAY)                                                                                      VARIABLES
                                                                                                          VARIABLE

Besley and    2002    India       Redistributive      1958-1992 states          public               Bulletin on Food       food grain                ratio of urban to total     Turnout in previous election does not
Burgess                           consequences                  (16)            distribution of      Statistics, Ministry   production per            population, population      affect responsiveness of state
(2002)                                                                          food and state       of Food and            capita, real per          density, log population,    governments. Greater political
                                                                                government           Agriculture,           capita flood damage       Consumer Price Index        competition associated with higher levels
                                                                                expenditures on      Directorate of         to crops, newspaper       for Agricultural            of public food distribution, but not
                                                                                calamity relief      Economics and          circulation (media        Laborers and Industrial     calamity relief. Levels of food
                                                                                                     Statistics,            development),             Workers                     distribution respond to political
                                                                                                     Government of          ownership of                                          competition and timing of elections, but
                                                                                                     India, Reserve         newspapers, turnout,                                  calamity relief does not. (1435) 10% drop
Besley,       2005    India       Political returns         2002 househol holders of BPL             Bank of India
                                                                                                     survey conducted       political competition,
                                                                                                                            political                 village characteristics     In food production resources politicians
                                                                                                                                                                                  in targeting public will bring 1% increase
Pande, and                                                       ds       (Below Poverty             by authors in 2002 opportunism                   such as village literacy    exhibit group preferences and are
Rao (2005)                                                       (approx Line) cards                                                                  rate 1991 and whether       opportunistic, and individual and village
                                                                 5400) in                                                                             village had at least one    characteristics mediate the extent of
                                                                 259                                                                                  Gram Sabha meeting in       opportunism. (28) "Selection is
                                                                 villages                                                                             the last year, individual   manifested in more educated politicians
                                                                 in 4                                                                                 characteristics such as     being less opportunistic. Incentives are
                                                                 southern                                                                             education, a                shaped by village meetings in which
                                                                 states                                                                               household's political       villagers ratify beneficiary lists chosen by
                                                                                                                                                      history, SC/ST              politicians" (25)
                                                                                                                                                      household
Besley,       2004    India       Political           survey       househol     which                "Independent audit     Household                 Household regressions:      "Living in a reserved GP increases a
Rahman,                           favoritism          taken in     ds,          households           of village             regressions: SC/ST        dummies for Christian       SC/ST household's likelihood of getting" a
Pande, and                                            September-   villages     received "low-       facilities", surveys   dummy (see                or Muslim household,        "low-spillover" public good (423).
Rao (2004)                                            November     (396)        spillover" public    of households          comments), SC/ST          household size, age,        "Household public goods have low
                                                      2002                      goods (i.e., had a                          reservation dummy,        literacy, occupation of     spillovers and are targeted toward SC/ST"
                                                                                house or toilet                             Pradhan's village         household head,             (425). "…The Pradhan enjoys agenda
                                                                                built), which                               dummy, government         household of elected        setting power in resource allocation"
                                                                                villages received                           headquarter dummy.        official, village fixed     (425)
                                                                                "village-level"                             Village-level             effects, Village-level:
Betancourt    2000    India       Political           1977-78      districts    allocation of        Indian                 % of eligible             % of scheduled caste,       "features of each state common to all
and Gleason                       favoritism                       (325)        educational and      Development            population voting in      % Muslim, agricultural      districts within a state are important
(2000)                                                                          medical services     District Data          1977-8 state              productivity, proportion    determinants of the provision of these two
                                                                                (doctors/nurses/t    compiled from          assembly elections,       landless, rural land per    public inputs to rural households. All 17
                                                                                eachers per 10       1981 Indian            female/male voter         person, proportion          state dummies are significantly different
                                                                                population)          Census and other       turnout ratio, number     urban                       from zero at the 1% level." (2175)
                                                                                                     sources                of voting                                             "Existence of statistical discrimination in
Bettcher      2005    Japan and   Political           Japan:      factions      experience of        (non-regression        organizational                                        "clientels and factions of interest are
(2005)                Italy       favoritism          1947 to     (2)           factions in          analysis)              dimension (strongly                                   based primarily on patronage, while
                                                      1993                      Japan's Liberal                             organized (clientel)                                  tendencies and factions of principle are
                                                      Italy: 1964               Democratic Party                            or weakly organized                                   based primarily on policy beliefs. Both
                                                      to 1982                   and Italy's                                 (tendency)), and                                      factions of interest and faction of principle
                                                                                Christian                                   motivational                                          are more highly developed
                                                                                Democracy                                   dimension (factions                                   organizationally than clienteles or
                                                                                (non-regression                             of interest or factions                               tendencies." (344)
                                                                                analysis)                                   of principle)                                         "Overall these groups [LDP factions and
Bodenstein    2008    European    Redistributive      2000-2006 region          allocation of    European                   Federalism, electoral     GDP per capita,             DC factions]funding (selection of eligible
                                                                                                                                                                                  Objective-1 are best classified as factions
and                   Union       consequences                                  Structural Funds Commission                 competition               unemployment,               regions) is basically driven by regional per
Kemmerling                                                                                                                  (divergence between       agricultural funds per      capita income, whereas the selection of
(2008)                                                                                                                      two largest parties),     capita                      Objective-2 (structural funds allocation to
                                                                                                                            effective number of                                   eligible regions) regions is determined by
                                                                                                                            parties                                               the regional unemployment rate. (15)




Bouvet and    2010    European    Redistributive      1994-99      region       allocation of    European                   left-wing national        GDP per capita,             "Left-wing governments and more EU-
Dall'erba             Union       consequences                     (120)        Structural Funds Commission                 and regional              unemployment,               sceptical countries get more objective 3
(2010)                                                                                           (1989–93 are from          governments,              agricultural funds per      and 4 and 5 funds, but they get less aid
                                                                                                 ‘Community                 percentage of EU-         capita                      through objective 2." (523) "Political
                                                                                                 Structural                 sceptics in national                                  bargaining will always be part of the
                                                                                                 Interventions’,            populations,                                          allocation process because
                                                                                                 Statistical Report         National and                                          there are too many potential recipient
                                                                                                 No. 3 and No. 4            regional electoral                                    regions, and the decision process involves
                                                                                                 (European                  margins, Dummy                                        interaction between several levels of the
                                                                                                 Commission,                variable for a safe                                   political arena." (524)
                                                                                                 COLLECTION
                                   CLASSIFICATION                                                                         POLITICAL
                                                      YEAR(S)                   DEPENDENT      INSTRUMENT FOR                                   OTHER INDEPENDENT
   AUTHOR      YEAR     COUNTRY    (FROM REVIEW                  UNITS (N)                                              INDEPENDENT                                       FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                     ANALYZED                    VARIABLE         DEPENDENT                                         VARIABLES
                                      ESSAY)                                                                              VARIABLES
                                                                                                   VARIABLE

Brown and     2009    multi-      Redistributive    1973-1997 countries agricultural/resid World Bank,              democracy indicator, population, urban              "democratization increases the residential
Mobarak               country     consequences                (57)      ential/"other"     World                    democracy*GDP        population, electricity        sector's share of consumption relative to
(2009)                                                                  sector's share of Development               interactions         production, electricity        industry" (200)
                                                                        electricity        Indicators                                    privatization, total
                                                                        consumed                                                         electricity consumption,
                                                                        relative to                                                      GDP per capita, GDP
Bruhn (1996) 1996     Mexico      Political returns 1988-1991 states    industry
                                                                        amount of          Tercer Informe de                             per capita squared
                                                                                                                    PRI vote 1988, left "marginalization index"         "Only the left vote reaches at least .1
                                                              (31)      PRONASOL           Gobierno, 1991:          vote 1988 (PRI      which is a government           significance in models of spending
                                                                        spending per       Anexo                    opponent), PAN vote measure and includes            reorientation. Curiously, PRI vote reaches
                                                                        capita,                                     1988                poverty, living                 .1 significance in explaining spending per
                                                                        "reorientation" of                                              standards, and access to        capita" (160)
                                                                        funds (change                                                   services; GDP per
                                                                        from 1988                                                       capita, change in GDP
                                                                        levels)                                                         per capita



Brusco,       2004    Argentina   Swing versus            1999 individual Probability of a surveys conducted        Peronist handout,    income, education,             Only about 7% of whole sample received
Nazareno,                         core                                    Peronist Vote in by authors in 2001-      Peronist             housing, gender age,           goods and 12% of them are low-income
and Stokes                                                                Presidential and 2002                     sympathizer, radical population                     class, majority of respondents said they
(2004)                                                                    Gubernatorial                             sympathizer                                         pay attention to programmatic appeals
                                                                          Elections and                                                                                 more than in the past. "The greater vote-
                                                                          types of targeting                                                                            buying effort by the Peronist Party, and
                                                                          behavior (i.e.                                                                                the greater effectiveness of these efforts,
                                                                          patron, gift,                                                                                 reflect, we believe, the party's deeper
                                                                          Puntero, job)                                                                                 penetration of lower- class social
                                                                                                                                                                        networks and hence its greater ability to
                                                                                                                                                                        monitor voters." (73)
Bungey,       1991    Australia   Redistributive    1956-57 -- states (6) special purpose     Commonwealth          proportion of state's     proportion of urban       "Grants may well go to states where the
Grossman,                         consequences      1985-86               recurrent grants,   Governments'          seats held by the         population,               federal government's need to increase
and Kenyon                                                                special purpose     Budget Paper          federal government's      unemployment rate,        voter support is greatest. However, as the
(1991)                                                                    capital grants,     Number 7,             party, proportion         taxable income per        position of the government in the state
                                                                          special purpose     Payments to or        squared, dummy if         capita, population        improves, pork-barelling behavior is likely
                                                                          total (sum),        For the States, the   federal and state         density in rural areas,   to take over" (negative slop on proportion
                                                                          general purpose     Northern Territory    governments are of        population, school        of state seats held by federal government
                                                                          grants              and Local             the same political        pupils as proportion of   party but positive slope on term squared)
                                                                                              Government            party, proportion of      population,               (665). Marginal seats not significant.
                                                                                              Authorities           seats held with less                                Dummy on state and central government
Cadot, Röller, 2006   France      Political         1985-1992 regions        infrastructure   Federation            Number of large,          region GDP, region        "Public goods, even if imperfectly
and Stephan                       favoritism                  (21)           investments      Nationale des         established firms in a    employment, non-          'targetable'…are used by politicians as
(2006)                                                                                        Travaux Publics,      region (as proxy for      residential private       redistribution instruments" (1147). More
                                                                                              SNCF (national        lobbying activity).       capital stock,            extremist voters (Le Pen, hunter party) in
                                                                                              railway company)      Proxies for swing         transportation            a region means the region gets more
                                                                                              for railroads,        voters : (1) Absolute     infrastructure stock,     investment per worker -- significant in all
                                                                                              Laboratoire           difference between        transportation            specifications. Sharing a party with the
                                                                                              d'Observation         vote for right and left   infrastructure net        national government gets a region more
                                                                                              Economique et des     parties (which, they      investments               investment, but insignificant most of the
                                                                                              Institutions          admit, is not a                                     time Absolute difference between right
                                                                                              Locales (OEIL)-       "swing voter"                                       and left vote totals is significant and
                                                                                              gathered data from    variable but is "a                                  negative.
                                                                                              SNCF for roads        measure of how
                                                                                                                    heated the electoral
                                                                                                                    race was"(1135))
                                                                                                                    and (2) Vote for Le
                                                                                                                    Pen and another
                                                                                                                    extreme party
                                                                                                                    (disaffected voters
                                                                                                                    are ripe for buying).
                                                                                                                    Also a dummy where
                                                                                                                    local and national
                                                                                                                    political congruence
                                                                                                                    (for targeting "local
                                                                                                                    political allies"), and
                                                                                                                    incumbent's margin
                                                                                                                    (safe candidates get
                                                                                                                    more lobbying, have
                                                                                                                    greater incentives to
                                                                                                COLLECTION
                                   CLASSIFICATION                                                                    POLITICAL
                                                     YEAR(S)                 DEPENDENT        INSTRUMENT FOR                             OTHER INDEPENDENT
  AUTHOR       YEAR    COUNTRY     (FROM REVIEW                 UNITS (N)                                          INDEPENDENT                                   FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                    ANALYZED                  VARIABLE           DEPENDENT                                   VARIABLES
                                      ESSAY)                                                                         VARIABLES
                                                                                                  VARIABLE

Calvo and       2004 Argentina    Political returns 1987,      provinces share of           Ministry of the    Peronist vote share,    logged median income,   "Both positive and significant effects of
Murillo                                             1990,      (24) and expenditures        Economy            UCR vote share          over-representation     the Peronist vote on both the amount of
(2004)                                              1995, 2000 Buenos financed by the                          (both logged)           (share of legislators   federally financed expenditures and on
                                                               Aires     federal                                                       over share of           revenue sharing…By contrast, the UCR-
                                                                         government,                                                   population), logged     Alianza has no significant effect on either
                                                                         relative share of                                             population share        federal financing or the relative revenue-
                                                                         fiscal resources                                                                      sharing ratio" (749)
                                                                         (revenue sharing                                                                      "patronage can serve as a distributive
                                                                         ratio) received by                                                                    mechanism to benefit partisan
                                                                         each province                                                                         constituencies with higher electoral
                                                                                                                                                               returns for political parties whose
                                                                                                                                                               constituencies are more dependent on
                                                                                                                                                               public largesse (demand side). We also
                                                                                                                                                               show how institutionally defined
                                                                                                                                                               territorial biases can affect access to
                                                                                                                                                               patronage if overlapping with geographic
                                                                                                                                                               concentration of electoral support for
                                                                                                                                                               political parties (supply side)." (754)


Calvo and     2009    Argentina   Political               2007 individual voters'           survey conducted   size and structure of   socio-demographic       "Our results provide ample support for the
Murillo               and Chile   favoritism                   (5600=28 distributive        by authors         voters' personal and    status of respondents   importance that voters attach to local
(2009)                                                         00*2       expectations                         political networks,     (measured through       benefits in Argentina -- given more
                                                               countries)                                      political behavior of   questionnaire)          clientelistic linkages and broader
                                                                                                               voters (measured                                networks of the PJ and UCR -- and the
                                                                                                               through                                         importance of global benefits in shaping
                                                                                                               questionnaire),                                 the distributive expectations of Chilean
                                                                                                               likelihood of                                   voters -- under more programmatic party-
                                                                                                               receiving handouts,                             voter linkages. Proximity to party
                                                                                                               being offered a job                             networks, therefore, is a stronger predictor
                                                                                                               in public sector                                of voters' distributive expectations in
                                                                                                               (patronage), or                                 Argentina, whereas Chilean voters giver
                                                                                                               witnessing increased                            more importance to ideology in perceived
                                                                                                               public investment in                            access to benefits." (25)
                                                                                                               community if a given
                                                                                                               party wins the
                                                                                                               election (pork),
                                                                                                               global and local
                                                                                                               benefits perceived
                                                                                                               by voters : proximity
                                                                                                               of voters to each
                                                                                                               party's network,
                                                                                                               ideological proximity
                                                                                                               between voters are
                                                                                                               parties



Canes-Wrone           U.S.        Political               2008 individual % homes sold      zillow.com,        gubernatorial           real income growth,     On average, gubernatorial election
and Park                          business cycles              s (1000                      Conventional       election year and its   change in               depresses home sales by two and three-
(2013)                                                         per                          Mortgage Home      interaction with        unemployment,           tenths of a pp.
                                                                                            Price Index        electoral               demographic change,
                                                                                            published by the   competitiveness         lagged foreclosure ,
                                                                                            Federal Home                               years indicators
                                                                                            Loan Mortgage
                                                                                            Corporation
Canes-Wrone           10 OECD     Political         1975-2006 country nongovernment OECD.Stat                  polarization between interest rates, country    Nongoverment investment declines in the
and Park              countries   business cycles             by quarter gross fixed                           major coalitions,    and year fixed effects     preelection period if polarization between
(2012)                                                        (936)      capital formation                     electoral                                       major parties is high or the race is
                                                                         (GFCF)                                competitiveness                                 competitive (103). When polarization is
                                                                                                                                                               high, 2-3pp decrease of private fixed
                                                                                                                                                               investment growth in the quarter before
                                                                                                                                                               the election. Private fixed investment
                                                                                                                                                               decline in preelection period occurs when
                                                                                                                                                               parties are within 15pp of each other
                                                                                                                                                               (113).
                                                                                                 COLLECTION
                                    CLASSIFICATION                                                                    POLITICAL
                                                      YEAR(S)                   DEPENDENT      INSTRUMENT FOR                               OTHER INDEPENDENT
   AUTHOR       YEAR     COUNTRY    (FROM REVIEW                 UNITS (N)                                          INDEPENDENT                                         FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                     ANALYZED                    VARIABLE         DEPENDENT                                     VARIABLES
                                       ESSAY)                                                                         VARIABLES
                                                                                                   VARIABLE

Case           2001    Albania     Political         1992-1996 rural   block grants to   Ministry of Labor      Voting records in         number of families          "..more assistance allocated to swing
(2001)                             favoritism                  commune communes of                              1992 local                requesting assistance,      communes…" (408), also more money to
                                                               s (191) social assistance                        government                commune population,         "pivotal" communes (districts that voted
                                                                       funds                                    elections and 1994        family land holding         for the winning party, which she considers
                                                                                                                constitutional            patterns                    to be a measure of pivotal, got larger
                                                                                                                referendum.                                           block grants). "Our predictions are
                                                                                                                Absolute value                                        consistent with those that would obtain
                                                                                                                between Democratic                                    in…a Dixit and Londregan framework."
                                                                                                                vote and 50% as                                       (409)
                                                                                                                measure of swing.
                                                                                                                Support of 1994
                                                                                                                constitutional
                                                                                                                referendum taken as
                                                                                                                measure of
                                                                                                                Democratic support


Castells and   2005    Spain       Redistributive    1987-1996 electoral infrastructure       Fundacion BBVA Position in the              output growth,              (i) "regional governments need more time
Solé-Ollé                          consequences                regions    investments                        electoral cycle, plus        population growth,          to adjust to the long-run equilibrium, (ii)
(2005)                                                         (NUTS3)                                       Electoral                    growth in ratio of trucks   are inclined towards efficiency, and (iii)
                                                               (50),                                         Productivity: pivotal        to output, growth in        are more responsive to infrastructure
                                                               which                                         representative               vehicle-km per year,        needs" (1200) Vote margin (inverse of
                                                               receive                                       dummy, number of             growth in rail              number of votes needed to gain/lost a
                                                               allocation                                    votes needed for             passengers per year,        seat) and turnout are mostly positive and
                                                               s from                                        incumbent to                 growth in tons              statistically significant. The measure of
                                                               both the                                      gain/lose a seat,            transported per year,       swing voters (cut point density) is not stat
                                                               central                                       turnout in legislative       growth in air               sig. When combined into an electoral
                                                               and                                           elections, cut-point         passengers per year,        productivity composite, they become
                                                               regional                                      density of legislative       growth in regional          positive and stat sig. Thus evidence "in
                                                               governme                                      elections,                   general revenues,           favour" of the Dixit-Londregan model
                                                               nts.                                          Partisanship:                growth of regional          (1199). Leftist regional governments
                                                               Central                                       Incumbent's share of         capital revenues,           (sharing party with center) and a higher
                                                               and                                           votes in last election,      growth in net debt of       share of socialist regional representatives
                                                               regional                                      dummy for regional           region                      attract more investment. The effect of vote
                                                               allocation                                    socialist government                                     share for incumbents is positive but not
                                                               s                                             (socialists always the                                   stat sig.. "Constituency support (i.e., vote
                                                               investigat                                    central government                                       share) does not influence investment (Cox
                                                               ed                                            incumbent), socialist                                    and McCubbins 1986)..." (1199).
                                                               separately                                    share in regional
                                                                                                             government




Chhibber       1995    India       Political returns 1967-84    states       loan             Reserve Bank of   the "political            revenue accruing to the     "The coefficient for the political variable
(1995)                                                          (15)         disbursements    India Bulletin    situation of the          center (national            was positive and significant, suggesting
                                                                             from centre to                     central party in a        revenues), per capita       that the central government used political
                                                                             state, food                        state", determined by     income                      considerations to disburse resources to the
                                                                             disbursements                      three criteria: "first,                               states" (85) Also, "political factors have a
                                                                             from centre to                     whether a state faced                                 positive and significant influence on the
                                                                             state                              either a national or                                  distribution of food from the centre" (87)
                                                                                                                state election;
                                                                                                                second, whether the
                                                                                                                state government and
                                                                                                                national government
                                                                                                                were controlled by
                                                                                                                the same party prior
                                                                                                                to the election; third,
                                                                                                                whether there was
                                                                                                                serious intra-party
                                                                                                                factional conflict or
                                                                                                                not." (84)
                                                                                                    COLLECTION
                                   CLASSIFICATION                                                                          POLITICAL
                                                       YEAR(S)                   DEPENDENT        INSTRUMENT FOR                                 OTHER INDEPENDENT
   AUTHOR      YEAR     COUNTRY    (FROM REVIEW                   UNITS (N)                                              INDEPENDENT                                    FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                      ANALYZED                    VARIABLE           DEPENDENT                                       VARIABLES
                                      ESSAY)                                                                               VARIABLES
                                                                                                      VARIABLE

Chhibber and 2004     India       Programmatic        1967-1997 major         proportion of      Reserve Bank of     dummy for party           sources of income,     "multiparty states have higher levels of
Nooruddin                         versus pork                   states        state's budget     India Bulletin      system, one year lag,     state's per capita     expenditures on club goods or civil
(2004)                                                          (15)          allocated to civil                     nature of ideological     income, growth rate,   administration spending than their two-
                                                                              administration                         competition in the        revenue from income    party counterparts. Center-right
                                                                              (as a club good)                       assembly,                 taxes, grants from     competition generates more bloated state
                                                                                                                     competitiveness of        central government,    apparatuses when compared to a single-
                                                                                                                     election, dummy for       loans from central     party dominant system than competition
                                                                                                                     coalition                 government             between two centrist parties or a center
                                                                                                                     government, voter                                and leftist party." (18) Two party systems
                                                                                                                     turnout                                          need to win majority of vote forcing
                                                                                                                                                                      political parties to build coalitions,
                                                                                                                                                                      whereas in multi-party systems, parties
                                                                                                                                                                      have an incentive to focus on their voting
                                                                                                                                                                      blocs since any loss of support can result
                                                                                                                                                                      in defeat. (24)


Chong, De La 2010     Mexico      Political returns         2009 electoral turnout              field experiment     treatment,                demographics           "Our results so far indicate a robust and
O, Karlan,                                                       precincts                      conducted by         interventions with                               negative effect of dissemination of
and                                                              (1298)                         authors in 2009 in   information on                                   information about extensive corruption on
Wantchekron                                                                                     Jalisco, Morelos,    Fondo para la                                    voter turnout. The drop in participation
(2010)                                                                                          Tabasco              Infraestructura                                  translates into a decrease in the
                                                                                                                     Social Municipal                                 incumbent's vote share, although this last
                                                                                                                     (FISM)                                           result is not statistically significant in all
                                                                                                                                                                      specifications.... Precincts that were
Cingano and   2009    Italy       Redistributive      1985-97    manufact     firm revenues     Italian National     fraction of sample                               informed that our results, access to
                                                                                                                                                                      According to the municipality exercised
Pinotti                           consequences                   uring        deflated using 2- Accounts             period in which firm                             political connections increases firm
(2009)                                                           firms        digit industry                         was connected to a                               revenues by approximately 5%, yielding to
                                                                 (1227)       indexes                                politician (measured                             an almost equivalent change in current
                                                                                                                     by indicator variable                            profits. These gains only accrue to firms
                                                                                                                     of at least 1                                    establishing a connection through
                                                                                                                     employee of firm                                 politicians appointed with the party that
                                                                                                                     being appointed in a                             won the elections: firms connected
                                                                                                                     local government),                               through other politicians see no increase
                                                                                                                     politician elected                               in market shares, just as non-connected
                                                                                                                     with a party that won                            firms." (5-6) "estimates from alternative
                                                                                                                     elections (measured                              production function specifications
                                                                                                                     by an indicator                                  indicate that political connections do not
                                                                                                                     variable for at least 1                          have any impact on productivity. Rather,
                                                                                                                     employee of firm                                 the average effect on revenues turns out to
                                                                                                                     being appointed in a                             be driven by firms operating in markets in
                                                                                                                     local government                                 which public demand plays a major role."
                                                                                                                     with the winning                                 (7)
                                                                                                                     coalition)


Cole          2009    India       Political           1992-1999 districts     all credit,       "Banking             years until next       loan characteristics      Credit is targeted toward districts in which
(2009)                            business cycles               (412)         agriculture       Statistics" by       scheduled election (1- (loan size, number of     the majority party just won or just lost the
                                                                              credit,           Reserve Bank of      4 yrs)                 agricultural loans,       election, only in election years. Loan
                                                                              nonagricultural   India                                       interest rates),          write-offs are greatest in districts in which
                                                                              credit                                                        nonperforming loan        winning party enjoyed the most electoral
                                                                                                                                            variables, district       success, only following an election. (248)
                                                                                                                                            characteristics           Agriculture credit issued by banks is
                                                                                                                                            (population, literacy     lower in off election years (2-4 yrs prior to
                                                                                                                                            etc)                      election) than the year prior to an election
                                                                                                                                                                      year. The difference is about 8 % points.
                                                                                                                                                                      (230) "Electoral cycles serve as an
                                                                                                                                                                      instrument for identifying the effect of
                                                                                                                                                                      marginal loans on output" (230)
                                                                                                      COLLECTION
                                    CLASSIFICATION                                                                             POLITICAL
                                                       YEAR(S)                    DEPENDENT         INSTRUMENT FOR                                 OTHER INDEPENDENT
   AUTHOR       YEAR     COUNTRY    (FROM REVIEW                  UNITS (N)                                                  INDEPENDENT                                     FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                      ANALYZED                     VARIABLE            DEPENDENT                                       VARIABLES
                                       ESSAY)                                                                                  VARIABLES
                                                                                                        VARIABLE

Cole, Healy,   2008    India       Redistributive    1977-99     administr     crop yield, state   Sanghi, Kumar,        weather measured by                               "Extreme amounts of rain lead to higher
and Werker                         consequences                  ative         expenditure on      McKinsey (1998,       rainfall                                          amounts of spending -- particularly
(2008)                                                           district*el   disaster relief,    crop), Reserve                                                          droughts. Budget transfers from central
                                                                 ection in     vote share for      Bank of India,                                                          government for the purpose of calamity
                                                                 28 states     incumbent ruling    Election                                                                relief were often provided as matching
                                                                 (21,532       party candidate     Commission of                                                           funds, directly proportional to the amount
                                                                 elections,                        India                                                                   spent by the state." (14) Ruling party does
                                                                 594                                                                                                       very poorly during extreme droughts but
                                                                 districts)                                                                                                its performance increases steadily with
                                                                                                                                                                           rainfall, reaching an optimum at a point
                                                                                                                                                                           between 0 and 1 standard deviations
                                                                                                                                                                           above the mean. As rainfall exceeds this
                                                                                                                                                                           optimum, support for the ruling party
                                                                                                                                                                           declines." (14-15) Voters punish ruling
                                                                                                                                                                           coalition even for circumstances beyond
                                                                                                                                                                           its control. "Voters are twice as sensitive
                                                                                                                                                                           to relief spending during extreme weather
                                                                                                                                                                           as they are to relief spending during bad
                                                                                                                                                                           weather" (19)


Costa-i-Font, 2003     Mexico      Swing versus      1990-95     states        real per capita     statistical annexes   share of local        income, population,         robust and positive effect of public
Rodriguez-                         core                          (32)          public              to the Presidential   municipalities        percentage rural            investment on the support for PRI, the
Oreggia, and                                                                   investment          Address to the        affiliated to central                             more educated population, the lower the
Lunapla                                                                        allocated           Nation (various       government’s party,                               support, negative and significant effect of
(2003)                                                                                             years)                dummy for states                                  the governor of the state being of a
                                                                                                                         governed by political                             different political affiliation to the PRI
                                                                                                                         party other than PRI,
                                                                                                                         dummy for
                                                                                                                         Campeche and
                                                                                                                         Tabasco (two major
                                                                                                                         recipients of public
                                                                                                                         investment for oil
                                                                                                                         field exploitation)


Crampton       2004    Canada      Swing versus      1996-2000 electoral       job grants by       Transitional Jobs     MP's margin of          unemployment              "I find that the strong hypothesis that
(2004)                             core                        district        Human               Fund, Canada Jobs     victory in previous     variables including       districts represented by the governing
                                                               (595)           Resources           Fund by Human         election, whether       whether threshold level   party will fare better than those
                                                                               Development         Resources             MP runs for             was met, and by how       represented by the opposition is not
                                                                               Canada per          Development           reelection in the       much threshold was        supported. Overall, Liberal districts fared
Crisp and      2004    Colombia    Swing versus      July 1998- senators       which               Senate Travel         vote share, security    destination appeal        "Senators prefer maintaining their existing
Desposato                          core              July 1999 (100)           "department"        Office                of seat (margin of      (political kidnappings,   constituency to developing a new one"
(2004)                                                                         (district)                                victory), political     beach resort,             (147) "Where senators do branch out,
                                                                               incumbent                                 ambition,               bogota/cundinamarca)      they avoid seeking votes in others'
                                                                               senators traveled                         "dominance" (sum of                               bailiwicks" (148) "Politicians go where
                                                                               to                                        all candidates'                                   the voters are" (they travel more to more
                                                                                                                         squared vote                                      populated districts) (150) Candidates
                                                                                                                         percentages,                                      running for reelection visit their core
                                                                                                                         indicating whether                                constituencies more often than those who
Crisp and      2002    Columbia    Programmatic      1986-90     senators      vote                bill (Hirschman-      prior success of                                  "Prior success of senator and increase
Ingall                             versus pork       (pre-       (300)         concentration       Herfindahl Index      senator, experience                               experienced is positively associated with
(2002)                                               reform) and                                   used to calculate)    with the new rules,                               vote concentration. Negative sign on
                                                     1994-1998                                                           previous victory in a                             interaction term of experience and
                                                     (post-                                                              subnational district,                             traditional party interaction indicates that
                                                     reform)                                                             membership in                                     while patterns of support for candidates
                                                                                                                         traditional party,                                from traditional parties are becoming
                                                                                                                         experience*tradition                              more concentrated over time, they are
                                                                                                                         al party                                          doing so more slowly than those for
                                                                                                                                                                           candidates from new and traditionally
                                                                                                                                                                           small parties." (738)
                                                                                                 COLLECTION
                                  CLASSIFICATION                                                                         POLITICAL
                                                    YEAR(S)                   DEPENDENT        INSTRUMENT FOR                                 OTHER INDEPENDENT
   AUTHOR     YEAR     COUNTRY    (FROM REVIEW                 UNITS (N)                                               INDEPENDENT                                       FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                   ANALYZED                    VARIABLE           DEPENDENT                                       VARIABLES
                                     ESSAY)                                                                              VARIABLES
                                                                                                   VARIABLE

Dahlberg and 2002    Sweden      Swing versus            1998 municipal    grants/transfers Ministry of            To test                  Vacancy rate (vacant       "Strong support for the prediction derived
Johansson                        core                         ities that   for sustainable  Environment            "LW/DL" cutpoint         jobs divided by            from the LW/DL model saying that the
(2002)                                                        applied      ecological                              density of               unemployed persons),       incumbent government purchases votes by
                                                              for grants   development and                         ideological              tax base, social welfare   investing in those municipalities where
                                                              (115)        for increased                           distribution,            spending per capita,       there are a lot of swing voters" (38).
                                                                           municipal                               closeness of last        cash flow, population      Support using both "cutpoint density" and
                                                                           employment                              election as given        share under 16,            distance between blocs. No support for
                                                                                                                   absolute value of        population share under     "the hypothesis that the incumbent
                                                                                                                   difference in vote       64, environmental          government purchases votes by investing
                                                                                                                   shares. To test Cox-     rating, vote share of      in their own supporters...presented by Cox
                                                                                                                   McCubbins : dummy        environmental party        and McCubbins" (38).
                                                                                                                   if socialist majority
                                                                                                                   in municipal council,
                                                                                                                   vote share for
                                                                                                                   incumbent (Social
                                                                                                                   Democrats)


Dall'erba    2005    European    Redistributive    1989-1999 145           regional per       New Cronos Regio                                                         "We first used Moran’s I to detect the
(2005)               Union       consequences                regions at    capita GDP         by Eurostat                                                              presence of positive global spatial
                                                             NUTS II       series in Ecu                                                                               autocorrelation in the distribution of per
                                                             (Nomencl      current prices                                                                              capita GDP to explore whether the rich
                                                             ature of      (spatial                                                                                    (poor) regions have a propensity to be
                                                             Territoria    distribution)                                                                               clustered close to other rich (poor)
                                                             l Units       (exploratory)                                                                               regions. Global spatial autocorrelation
                                                             for                                                                                                       also characterizes the regional growth
                                                             Statistics)                                                                                               rate, structural funds and Community
                                                             over 12                                                                                                   projects total cost." (146) Evidence shows
                                                             EU                                                                                                        this persists over time.
de Figueiredo 2009   Brazil      Political returns 2008-10   countries
                                                             individual    number of          survey conducted     (experiment                                         "Bolsa Familia has had little or not effects
and Hidalgo                                                  s (1000)      politicians from   by authors in 2008   embedded in survey                                  on political attitudes beyond generating
(2009)                                                                     the PT or PSDB     after October        with all voters, non-                               support for President Lula for voters in
                                                                           that received a    Municipal            Bolsa voters group,                                 survey sample." (19) "Consistent with our
                                                                           positive           elections            Bolsa voters) receipt                               matching estimates, Bolsa Familia
                                                                           evaluation                              of PT message                                       recipients do not appear to attach any
                                                                           (feeling                                treatment, PSDB                                     positive affection to the PT party label.
                                                                           thermometer)                            message treatment,                                  However, the partisan message is
                                                                                                                   cues treatment                                      substantial." For PSDB politicians,
                                                                                                                                                                       partisan cues are not significant. (19)


de la Fuente, 1995   Spain       Redistributive    1981--1990 regions      infrastructure     National Institute   number of years          expected return to         infrastructure investments are not higher
Vives,                           consequences                 (17)         investments        of Statistics,       regional government      investment, income per     in regions that have the same party as the
Dolado and                                                                                    national Census,     controlled by            capita, effective public   central government (Socialist) and are not
Faini                                                                                         Municipal Census,    Socialists (the          capital stock, dummy       higher in regions with
(1995)                                                                                        Labor Force          national incumbent),     1986--1990                 nationalist/regionalist parties (taken to be
                                                                                              Survey, Mas et al    number of years                                     more active in lobbying for regional
                                                                                              (1993)               regional government                                 funding). "The party in power remains
                                                                                                                   controlled by                                       insignificant in all specifications" (34)
                                                                                                                   nationalist/regionalis                              Negative slope for income -- "on average
                                                                                                                   t parties (more active                              redistributive concerns outweighed the
                                                                                                                   in lobbying), dummy                                 political influence of the richer regions"
                                                                                                                   for Basque-Navarre
                                                                                                 COLLECTION
                                 CLASSIFICATION                                                                            POLITICAL
                                                   YEAR(S)                   DEPENDENT         INSTRUMENT FOR                                  OTHER INDEPENDENT
  AUTHOR     YEAR     COUNTRY    (FROM REVIEW                UNITS (N)                                                   INDEPENDENT                                    FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                  ANALYZED                    VARIABLE            DEPENDENT                                        VARIABLES
                                    ESSAY)                                                                                 VARIABLES
                                                                                                   VARIABLE

De La O     2008    Mexico      Political returns 1994 and   electoral    voting turnout      evaluation survey,     share of people        population 1994           "I show that average turnout and vote
(2008)                                            2000       seccion      (difference in      ENCASEH 1997           treated early by       logged, population        shares in the 1994 election do not predict
                                                             (precinct)   turnout 2000-       ran by Progresa's      Education, Health      change                    enrollment in the randomized early
                                                             (355)        1994)               implementers,          and Nutrition                                    treatment but the latter does not influence
                                                                                              Federal Electoral      program (Progresa)                               future electoral outcomes. The estimates
                                                                                              Institute, Census of   measured by                                      suggest that a section fully treated twenty-
                                                                                              Population and         multiplying a                                    one months before election time increased
                                                                                              Housing (2000)         variable that is one                             its turnout in 2000 by five points and its
                                                                                              and Partial            when the precinct                                incumbent vote share by four points. For
                                                                                              Cenusus (1995)         has localities that                              an average section with 578 potential
                                                                                              produced by            were part of the early                           voters the estimations imply a change in
                                                                                              INEGI (used for        treatment, and share                             in the probability of turnout from 0.60 in
                                                                                              experimental           of people living in                              1994 to 0.65 in 2000; and a change in the
                                                                                              setup)                 localities that were                             probability of voting for the incumbent
                                                                                                                     randomized with                                  from 0.38 to 0.42. I also show that the
                                                                                                                     respect to the                                   conditional effect of Progresa on the
                                                                                                                     precinct population;                             strength of the PRI in 1994 is smallest
                                                                                                                     vote share by                                    among those precincts where the PRI had
                                                                                                                     political party                                  more than eighty-five percent of the votes
                                                                                                                                                                      in 1994 compared to precincts where the
                                                                                                                                                                      PRI strength was moderate or low." (43)


De Rynck and 2001   European    Redistributive    1999 to    EU           allocation of                              cohesion policy                                  "While the evidence points to a stronger
McAleavey           Union       consequences      present    regions      Structural Funds                                                                            emphasis for social and local problems
(2001)                                                                    (non-regressional                                                                           and more intense interaction between the
                                                                          analysis,                                                                                   Commission and local actors, the
                                                                          exploratory)                                                                                dominant focus on economic problems
                                                                                                                                                                      and regional disparities remained intact in
                                                                                                                                                                      the 1999 reform." (553) GDP per capita
                                                                                                                                                                      remains central to the Commission in
                                                                                                                                                                      deciding redistribution.


Dellmuth    2009    European    Redistributive    2000-2006 EU            amount of           European               ratio of paid SF in     GDP, employment,         GDP per capita and unemployment rate,
(2009)              Union       consequences                regions       structural          Commission             relation to potentially share of public sector   are relatively identical and statistically
                                                            (148)         funding (SF)        decisions (various     available funds         employees                significant. "The measure introduced to
                                                                          granted to          years)                 (information), left                              capture the capability of regions to cater
                                                                          regional                                   parties vote shares,                             to the Commission’s demand for
                                                                          development                                political centre of                              information about the implementation of
                                                                          plans (log)                                gravity proposed by                              SF on the ground supports the argument
                                                                                                                     Gross and Siegelman                              that the Commission distributes higher
                                                                                                                     (1984), effective                                funding to regions which can help
                                                                                                                     number of parties,                               overcoming informational asymmetries
                                                                                                                     support for the EU,                              and increase the prestige and reputation of
                                                                                                                     Federalism index,                                the Commission. The variable is
                                                                                                                     EU support, election                             statistically significant across all models
                                                                                                                     year dummy                                       and has the expected sign." (17) "The
                                                                                                                                                                      empirical results show that the nature of
                                                                                                                                                                      the allocation process appears to ensure
                                                                                                                                                                      that national politicians cannot use the
                                                                                                                                                                      Community resources to build electoral
                                                                                                                                                                      support." (19)


Denemark    2000    Australia   Political returns 1990-1993 constitue total sports            Australian auditor- preelection            unemployment,                "…marginal electoral divisions in both
(2000)                                                      ncies     grants dollars          general             marginality            household income             1990 and 1993 can be seen as receiving
                                                                                                                  (measured by                                        significantly more funds than safer seats"
                                                                                                                  absolute value of 2                                 (905) yet districts with "very safe"
                                                                                                                  party preferred swing                               electoral margins (17.5-25%) also
                                                                                                                  required to lose the                                received larger grants (903)
                                                                                                                  seat), seat controlled
                                                                                                                  by ALP, member of
                                                                                                                  the preelection
                                                                                                                  Labor Cabinet
                                                                                               COLLECTION
                                  CLASSIFICATION                                                                       POLITICAL
                                                    YEAR(S)                 DEPENDENT        INSTRUMENT FOR                                OTHER INDEPENDENT
  AUTHOR      YEAR     COUNTRY    (FROM REVIEW                 UNITS (N)                                             INDEPENDENT                                     FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                   ANALYZED                  VARIABLE           DEPENDENT                                      VARIABLES
                                     ESSAY)                                                                            VARIABLES
                                                                                                 VARIABLE

Diaz-Cayeros, 2006   Mexico      Political         1989-1994 municipal share of private                          level of development time trend, lag of           Clientelism reduces risk for politicians
Estévez, and                     favoritism                  ity (2400) good provision                           using deprivation      dependent variable         because they can use preexisting networks
Magaloni                                                                (Pronasol)                               index from                                        to target core constituencies. Clientelism
(2006)                                                                                                           CONAPO, number                                    in poorer areas, and degree of clientelism
                                                                                                                 of parties, margin of                             declines as municipality grows. Against
                                                                                                                 victory, risk that                                expectations of modernization theory,
                                                                                                                 shows how fast PRI                                greater recourse to clientelism in
                                                                                                                 is losing votes in any                            municipalities at middle levels of
                                                                                                                 municipality relative                             development. Clientelism less prevalent
                                                                                                                 to the national trend                             in politically heterogeneous
                                                                                                                 for that party                                    municipalities. (201)


Diaz-Cayeros, 2008   Mexico      Swing versus      1989-1994 municipal per capita                                core voter, erosion     spatial lag, population   total per capita transfers averaged over six
Estevez,                         core                                  municipal level                           (decline), census-      (lag)                     years shows weak relationship with the
Magaloni                                                               expenditures, per                         based measure of                                  municipal welfare levels (CONAPO),
(2008)                                                                 capita private                            poverty or social                                 larger transfers are strongly associated
                                                                       and public                                marginality                                       with PRI's traditional strongholds in
                                                                       goods, share of                           (CONAPO), PRI                                     municipal elections. "Public goods should
Doménech,   2000     European    Redistributive    1986-199            transfers per     Commission                                      per capita income         Regional funds are negatively related to
Maudes, and          Union       consequences                          capita            accounting system                                                         per capita income and redistributive effect
Varela                                                                                   (SINCOM)                                                                  has increased over time. Social funds also
(2000)                                                                                                                                                             negatively related but redistributive effect
                                                                                                                                                                   is constant over time. Total expenditures
                                                                                                                                                                   show a negative relationship with per
                                                                                                                                                                   capita income. "we have found that per
                                                                                                                                                                   capita contributions increase
Drazen and   2005    Columbia    Political         1987-2000 municipal votes                Contraloria          vote share in         per capita GDP, time        proportionally to per capita income as a
                                                                                                                                                                   Decrease in some expenditures such as
Eslava                           business cycles             ities                          General, a public    previous election,    trend, social indicators    transfers to retirees, payments to
(2005)                                                       (1100)                         agency monitoring    per capita investment                             temporary workers, payments of debt
                                                                                            public finances      spending (log), per                               service. Pre-election increases in
                                                                                                                 capita current                                    development projects such as
                                                                                                                 spending (log), per                               infrastructure, power, and road
                                                                                                                 capita government                                 construction. " Columbian voters penalize
                                                                                                                 deficit fiscal debt                               the incumbent party for running high
                                                                                                                                                                   deficits. Furthermore, high capital
                                                                                                                                                                   expenditures increase the share of votes
                                                                                                                                                                   obtained by the incumbent party relative
                                                                                                                                                                   to the challenger." (25)


Dunning      2010    India       Political returns 2007-8     individual received a         survey conducted Scheduled Caste or          political knowledge and "while there is evidence that reservation
(2010)                                                        (968)      benefit or job     by authors in 2009 Scheduled Tribe           engagement              increases the propensity of member of the
                                                                         from the village                      participation                                     target groups to say that the council
                                                                         council in                                                                              prioritizes them, reservations does not
                                                                         previous 1-2                                                                            strongly boost perceptions that the council
                                                                         years                                                                                   effectively serves SC and ST
                                                                                                                                                                 communities." (20) "we fail to reject the
                                                                                                                                                                 null hypothesis of no effect for most of the
                                                                                                                                                                 many variables that tap distributive or
                                                                                                                                                                 policy outcomes... In Karnataka,
                                                                                                                                                                 reservation of village council presidencies
                                                                                                                                                                 does not appear to have strong distributive
                                                                                                                                                                 or policy effects." (40)


Dunning and 2008     Argentina   Swing versus            2003 individual received items   Survey                 usually' vote for                                 "Argentina's political machine targets
Stokes                           core                         s (500)    (clientelist                            Peronist party,                                   supporters when it uses its resources to
(2008)                                                                   benefits)                               'usually' vote for                                simulate turnout and opponents when it
                                                                         (formal model on                        Radical (UCR) party                               seeks to buy the vote of people who it
                                                                         certain vs.                                                                               knows are highly likely to vote." (24)
                                                                         potential voters
                                                                         and empirical)
Edwards and 2007     multi-      Programmatic      1970-2000 democrac government            World Bank's         district magnitude      lagged GDP, lagged        "Increases in district magnitude in
Thames               country     versus pork                 ies (77) spending              World                (log), indicator for    GDP per capita, lagged    candidate-centered systems lead to
(2007)                                                                measured by           Development          whether election        GDP growth, lagged        decreases in public goods spending, not
                                                                      general               Indicators (2004),   occurred in a           trade openness as a %     simply total expenditure. In the model, the
                                                                      government final      UN Common            previous time period,   of GDP, % of the          coefficient for the log of district
                                                                      consumption           Database             dummy variable for      population above the      magnitude is positive and statistically
                                                                      expenditure and                            presidential or         age of 65                 significant. This means that there is a
                                                                      education                                  parliamentary regime                              positive relationship between increasing
                                                                                                 COLLECTION
                                 CLASSIFICATION                                                                          POLITICAL
                                                   YEAR(S)                   DEPENDENT         INSTRUMENT FOR                                 OTHER INDEPENDENT
   AUTHOR    YEAR     COUNTRY    (FROM REVIEW                 UNITS (N)                                                INDEPENDENT                                         FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                  ANALYZED                    VARIABLE            DEPENDENT                                       VARIABLES
                                    ESSAY)                                                                               VARIABLES
                                                                                                   VARIABLE

Evans       2006    Canada      Political returns 1958-2000 electoral     share of vote for Elections Canada,      federal spending per     avg percent change in        "Approximately 8% of all majority party
(2006)                                                      district      the majority      Parliament of          capita in an electoral   the GDP per capita           candidates win or lose their electoral
                                                            (990)         candidate, or     Canada's website       district two years       during the previous          district races by less than 3% of the vote,
                                                                          opposition                               prior to a federal       electoral cycle; average     and 26% of the time the margin of victory
                                                                          incumbent                                election                 change in the provincial     or defeat is less than 10% of the vote. A
                                                                          candidate                                                         GDP per capita during        $100 increase in per capita spending in an
                                                                                                                                            the previous electoral       electoral district will increase the share of
                                                                                                                                            cycle, electoral district    the vote for the majority party by
                                                                                                                                            dummy variable, year         approximately 2% of the total vote cast
                                                                                                                                            dummy                        and at best leave unchanged the share of
                                                                                                                                                                         the vote for the incumbent opposition
                                                                                                                                                                         party candidate" (246)


Fattore     2007    European    Redistributive    1975-1995 state         percentage of the European               % of regions within GDP, New State Year,              "The coefficient for the lagging region
(2007)              Union       consequences                              European           Commission            a state that the       year dummies                   variable is positive as expected, as well as
                                                                          Regional                                 European                                              statistically significant. Substantively, for
                                                                          Development                              Commission has                                        every one unit increase in the percentage
                                                                          Fund (the largest                        identified as eligible                                of lagging regions within a specific state,
                                                                          portion of the                           for ERDF grants via                                   it can be expected that that state will also
                                                                          current Structural                       Objective 1 or 2 of                                   receive a .08% increase in their share of
                                                                          Funds) yearly                            SF.                                                   the ERDF disbursements for that year."
                                                                          budget allotted to                                                                             (15) "richer states contribute more to the
                                                                          a specific state                                                                               EU budget, and, therefore, exert more
                                                                                                                                                                         influence over decision-making regarding
                                                                                                                                                                         monetary issues such as the Structural
                                                                                                                                                                         Funds. However, this hypothesis was not
                                                                                                                                                                         supported by the model’s results." (15-16)



Ferraz      2006    Brazil      Political         1996-2003 municipal     cost of enforcing   Sao Paulo            election dummy for       municipal, economic          "environmental regulation is implemented
(2006)                          business cycles             ities in      environmental       Environmental        mayor, years to          controls                     less stringently whenever incumbent
                                                            Sao Paulo     regulation or       Agency and           election, reelection,                                 mayors have electoral incentives to distort
                                                            (645)         number of           election data from   election*reelection                                   such regulations in order to reduce the
                                                                          environmental       Tribunal Superior                                                          costs of regulation to local industries."
                                                                          licenses            Electoral
                                                                          distributed
Finan       2004    Brazil      Political         1996-2000 municipal federal outlays,        Federal Chamber      municipal vote           deputy characteristics:      "A deputy's share of votes in a
(2004)                          favoritism                  ities     and amount of           of Deputies and      share, dummy if          gender, education,           municipality influences both the choice to
                                                                      federal                 the Senate           mayor and federal        occupation, age, party       reward the municipality and the size of the
                                                                      expenditure                                  deputy are in the        affiliation, civil status,   public work" (13). "Being of the same
                                                                      conditional on                               same political party     incumbency status.           party is positively associated with more
                                                                      receiving an                                                          Municipality                 and larger public works." (16) "we
                                                                      outlay                                                                characteristics: gender,     document a strong association between
                                                                                                                                            education, party             the allocation of public works during the
                                                                                                                                            affiliation, winning         1996-9 electoral cycle and the degree of
                                                                                                                                            percentage, electoral        electoral support received in the 1994
                                                                                                                                            competition for mayor,       elections. A mere 10 percentage point
                                                                                                                                            electoral competition        increase in political support implies an
                                                                                                                                            for deputies, whether        increase of R$75,174 in public works for
                                                                                                                                            municipality has paved       a municipality during the electoral
                                                                                                                                            roads, shanty-town,          cycle."(32) Because these public goods
                                                                                                                                            incentives to foster         vary in their degree of exclusion,
                                                                                                                                            economic activity, a job     politicians prefer to allocate goods that are
                                                                                                                                            program, a constitution,     less excludable to municipalities where
                                                                                                                                            zoning laws, budget          they received more political support." (32)
                                                                                                                                            plan, property laws,
                                                                                                                                            work laws, percentage
                                                                                                                                            of budget devoted to
                                                                                                                                            public investment,
                                                                                                                                            number of municipal
                                                                                                                                            guards, whether the
                                                                                                                                            municipality has state
                                                                                                                                            judge, percentage of
                                                                                                                                            population living in
                                                                                                                                            urban areas, an
                                                                                                   COLLECTION
                                  CLASSIFICATION                                                                           POLITICAL
                                                     YEAR(S)                  DEPENDENT          INSTRUMENT FOR                                OTHER INDEPENDENT
   AUTHOR     YEAR     COUNTRY    (FROM REVIEW                 UNITS (N)                                                 INDEPENDENT                                     FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                    ANALYZED                   VARIABLE             DEPENDENT                                      VARIABLES
                                     ESSAY)                                                                                VARIABLES
                                                                                                     VARIABLE

Fukui and    1996    Japan       Political         1993-1994 prefecture local grants-in-       personal              influence of Diet                                 "Of the four factors discussed above that
Fukai                            favoritism                  s          aid funds and          interviews with       members and                                       appear to affect significantly, if not
(1996)                                                                  public works-          local politicians     national bureaucrats,                             determine a Diet member's role as an
                                                                        related funds          and aides to          party organization,                               intermediary between a local constituency
                                                                        (descriptive, non-     members of            impact of 1994                                    and the national treasury, personal
                                                                        regression             Parliament            electoral reform                                  reputation and "tribal" connections are
                                                                        analysis)              conducted in                                                            likely to remain important assets under
                                                                                               Okayama and                                                             any electoral system.... The electoral
                                                                                               Toyama in 1994                                                          Keiretsu or line and the noenkai, on the
                                                                                               summer                                                                  other hand, are likely to undergo a
                                                                                                                                                                       significant change, since they are devices
                                                                                                                                                                       geared primarily to electoral competition
                                                                                                                                                                       among candidates of the same party."
                                                                                                                                                                       (284)


Gagliarducci, 2008   Italy       Programmatic      1994-2006 individual    fraction of bills online archive of Margin of victory,            demographic               "Our empirical results show that
Nannicini,                       versus pork                 members       targeted to the    bills for legislative appointments in          characteristics, self-    majoritarian elections are strongly
and                                                          of Italian    region of election activity              parliament and in        declared previous job     associated with a greater amount of
Naticchioni                                                  House of      over the total                           government, party                                  targeted policies, such as pork-barrel
(2008)                                                       Represent     number of bills                          affiliation, political                             projects favoring local constituencies...
                                                             atives        presented as                             experience, local                                  proportional representation, however,
                                                             (1699)        main sponsor,                            government                                         would produce other side effects, such as
                                                                           fraction of                              experience, vote                                   rent extraction by elected officials who are
                                                                           parliament votes                         share, system of                                   less accountable to their voters." (3)
                                                                           missed without                           election, electoral                                "Being elected in the majoritarian system
                                                                           any legitimate                           district                                           entails an increase in the share of
                                                                           reason over the                                                                             geographically targeted bills of 8.2%, this
                                                                           total number of                                                                             is more than doubles the share of targeted
                                                                           electronic votes                                                                            bills with respect to the predicted value
                                                                                                                                                                       for 6.4 for proportional representatives at
                                                                                                                                                                       the threshold." (22) "...being elected in
                                                                                                                                                                       the majoritarian system entails a fall in the
                                                                                                                                                                       absenteeism rate equal to 14.9%, that is, a
                                                                                                                                                                       fall of more than 30% with respect to the
                                                                                                                                                                       predicted value of 47.7 for proportional
                                                                                                                                                                       representatives..." (23)


Gauthier and 2008    Chad        Redistributive          2004 health       whether health      Healthy Facilities     frequency of         location, no. of            "results show discretion of district and
Wane                             consequences                 centers      center received     Survey by World       supervision visits by employees                   regional administrators has a significant
(2008)                                                        (281) and    positive amount     Bank 2004             regional or district                              impact on probability of receiving public
                                                              hospitals    of public                                 officials, foreign                                resources, in that the probability of
                                                              (30)         resources                                 donor support                                     receiving material resources increases
                                                                           (financing,                                                                                 significantly for health centers that have
                                                                           drugs, material)                                                                            been visited by the regional delegate."
                                                                                                                                                                       (70)
Gimpelson,   2000    Russia      Political         1992-8      region      public              Goskomstat RF         electoral cycle         Per capita gross          "a negative correlation between
Treisman,                        business cycles               (78)        employment                                variables for           regional product, the     development (measured either by per
and                                                                        (share of total                           gubernatorial and       urban share of the        capita gross regional product or
Monusova                                                                   employment in                             governor elections,     population, the shares    urbanization) and public employment.
(2000)                                                                     education,                                governor affiliated     of the population under   public employment growth was faster in
                                                                           health care and                           with communists         16 and over 55, ethnic    ethnically-defined subunits of the Russian
                                                                           sport, social                                                     minorities, previous      state.... The evidence on electoral effects
                                                                           protection,                                                       year unemployment,        is mixed.... Political ideology or
                                                                           culture, art and                                                  change in                 affiliations of the governor do not appear
                                                                           science, and                                                      unemployment,             to influence his public employment
                                                                           public                                                            previous year/change in   strategy." (13-15)
                                                                           administration)                                                   federal transfers and
                                                                                                                                             loans, Year dummies
                                                                                                     COLLECTION
                                   CLASSIFICATION                                                                             POLITICAL
                                                      YEAR(S)                    DEPENDENT         INSTRUMENT FOR                                OTHER INDEPENDENT
   AUTHOR      YEAR     COUNTRY    (FROM REVIEW                  UNITS (N)                                                  INDEPENDENT                                    FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                     ANALYZED                     VARIABLE            DEPENDENT                                      VARIABLES
                                      ESSAY)                                                                                  VARIABLES
                                                                                                       VARIABLE

Giraudy       2007    Argentina   Political         1993-2002 23              percentage of       Instituto Nacional dummy for Peronist        roadblocks,               Partisanship at both the federal and
(2007)                            favoritism                  provinces       total population    de Estadisticas y  President and             unemployment              provincial level, together with institutional
                                                              and             that received       Censos (INDEC) Radical/Alianzist             (lagged), Economic        features of Argentine federal system
                                                              Ciudad          employment                             Governor in office,       crisis (2001-02)          strongly effect cross-provincial allocation
                                                              de              programs in                            presidential election,                              of employment programs. Positive and
                                                              Buenos          province                               Senate                                              significant relationship between the
                                                              Aires /                                                overrepresentation                                  number of roadblocks, level of
                                                              year                                                   (log), House                                        unemployment, 2001-2 economic crisis,
                                                                                                                     overrepresentation(lo                               and % of people getting emergency
                                                                                                                     g)                                                  employment programs. Presidential
                                                                                                                                                                         elections are statistically insignificant.
                                                                                                                                                                         (46)
Golden and    2012    India       Political         2000-2009 districts       electricity use     Uttar Pradesh     state assembly         division fixed effects,       Line loss higher in election years than non-
Min                               business cycles             in Uttar                            Power Corporation election year,         H야                            election years in UP. No evidence that
(2012)                                                        Pradesh                             Ltd.              candidates' criminal                                 line loss is greater in districts with
                                                              (1661)                                                record, voter turnout,                               criminal candidates.
                                                                                                                    win margin of last
                                                                                                                    election
Golden and    2008    Italy       Swing versus      1953-1994 electoral       total               government-                                                            goods go to districts with influential
Picci                             core                        district        infrastructure      administered                                                           legislators and not to core districts of
(2008)                                                        (32)            investments         quarterly survey of                                                    ruling parties
                                                                                                  subnational public
                                                                                                  offices
Gonzalez-     2010    Nicaragua   Political                2008 individual incidents of           survey conducted      political              income, education, age, Although we have no conclusive answer
Ocantos,                          favoritism                    (approx individual gift,          by authors            participation,         gender, urban/rural     in relation to the partisan determinants of
Kiewiet de                                                      3000)      neighborhood                                 support FSLN or                                clientelism, the results suggest that there
Jonge,                                                                     gift                                         PLC, welfare and                               are two central mechanisms for
Meléndez,                                                                                                               public works,                                  clientelistic linkages in the Nicaraguan
Osorio, and                                                                                                             attended Citizen                               context: 1) voter's perception that parties
Nickerson                                                                                                               Power Councils,                                can monitor their vote and 2) voter's
(2010)                                                                                                                  monitor                                        participation or proximity to state
                                                                                                                                                                       sponsored community like the Consejos
                                                                                                                                                                       de Poder Ciudadano." (27)


Hainmueller   2008    Germany     Programmatic      1950s-       district     SMD vote share Caramani (2000),           margin of victory in district-level covariates   For both parties under PR, incumbency
and Kern                          versus pork       2000s (8     (248 up      of party, PR vote Statistisches           each district to votes                           status has positive and significant effect
(2008)                                              federal      to 1990      share of party    Bundesmat               in the next election,                            on party vote share. Incumbency status
                                                    elections)   and 328      (Social                                   party incumbency,                                increases PR vote share by 1.7% for SPD
                                                                 after        Democratic                                legislator                                       and 1.4 % for CDU on average. Also,
                                                                 unificatio   Party(SPD), and                           incumbency                                       incumbency has substantial effect on
                                                                 n)           Christian                                                                                  party's SMD vote share, and is slightly
                                                                              Democratic                                                                                 larger for legislator incumbency.
                                                                              Union (CDU))
Healy and     2008    United      Programmatic      1988-2004 counties        presidential vote   General Election      disaster damage        county-level variables    "When disaster damage increases by 1%,
Malhotra              States      versus pork                 (3141)          share for           Data for the U.S.,    measure (natural       that can change over      the regression results suggest that relief
(2008)                                                                        incumbent party     Congressional         disasters), disaster   time (economic            expenditures increase by 0.072%. There
                                                                              in county           Quarterly's Votings   relief expenditure,    conditions), county       appears to be a significant political
                                                                                                  and Elections,        prevention spending    demographics              component to disaster relief spending too.
                                                                                                  ICPSR                 (also lagged),                                   Counties that more strongly supported the
                                                                                                                        incumbent party's                                incumbent party in the previous election
                                                                                                                        previous vote share                              received more relief spending than did
                                                                                                                        in the county                                    other counties, an effect that would not
                                                                                                                                                                         occur if relief spending only reflected
                                                                                                                                                                         response to disasters." (17)
                                                                                                    COLLECTION
                                CLASSIFICATION                                                                            POLITICAL
                                                   YEAR(S)                     DEPENDENT          INSTRUMENT FOR                                OTHER INDEPENDENT
   AUTHOR    YEAR    COUNTRY    (FROM REVIEW                    UNITS (N)                                               INDEPENDENT                                       FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                  ANALYZED                      VARIABLE             DEPENDENT                                      VARIABLES
                                   ESSAY)                                                                                 VARIABLES
                                                                                                      VARIABLE

Hirano      2005    Japan      Programmatic      1980-2002 municipal        targeted             account settlement minimum distance to       % of workforce             "under the MMD/STV system subsidies
(2005)                         versus pork                 ity              subsidies            (Chiho Zaisei      any LDP member's          engaged in first tier      were lower for municipalities farther from
                                                           (approx          measured by per      Kenkyu Kai),       home office               industries (agriculture    an LDP incumbent's hometown. … the
                                                           3300*yea         capita central-to-   various issues                               and fisheries), % of       municipalities far from LDP incumbents'
                                                           rs)              municipality                                                      population considered      home offices received more subsidies
                                                                            government                                                        dependent (over 65,        after the electoral reforms." (26) "As the
                                                                            transfers (Kokko                                                  under 15), income          model predicts both the distribution of
                                                                            shishutsukin)                                                                                candidates' electoral support and the
                                                                                                                                                                         distribution of government resources are
                                                                                                                                                                         found to be more concentrated in MMD
                                                                                                                                                                         systems than SMD systems." (27)



Hirano      2006    Japan      Political returns 1977-1992     district-    per capita central- Nikkei NEEDS        legislators who pass      1st Tier workers/legislators who pass away in office:
(2006)                                           (includes 5   level        to-local            database and        away in office : effect   workers (log),   "There is no statistically significant
                                                 elections     (1760)       government          various issues of   of individual                              evidence that losing any type of Diet
                                                                                                                                              population density
                                                 for Lower     and          transfers           the                 representative                             member affects the overall amount of
                                                                                                                                              (log), per capita income
                                                 House of      municipal                        shichosonbetsu      measured by death of                       transfers directed to particular electoral
                                                 the           ity-level                        kessan jyoko        a legislator, whether                      districts." (15) "Contrary to popular
                                                 Japanese      (49895)                          shirabe             deceased incumbent                         perception, there is no evidence that
                                                 Diet)                                                              was elected for                            senior LDP incumbents had a statistically
                                                                                                                    several terms,                             significant effect on the distribution of any
                                                                                                                    whether deceased                           type of transfer." (16) close elections :
                                                                                                                    incumbent did                              "victory of an LDP incumbent affects the
                                                                                                                    poorly in the election                     distribution of central government
                                                                                                                    to legislative session,                    transfers to municipalities which are part
                                                                                                                    whether LDP                                of the incumbent's core electoral support
                                                                                                                    representative passes                      in the first year following the election."
                                                                                                                    away                                       (18)
                                                                                                                    close elections :                          "individual LDP representatives are not a
                                                                                                                    indicator for whether                      free-flowing "pipeline of pork".... There is
                                                                                                                    LDP candidate won                          no evidence that representatives, even
                                                                                                                    the last seat in                           senior ones, affect the distribution of
                                                                                                                    district, difference in                    central government transfers across
                                                                                                                    government transfers                       districts. Electoral margins, rather than
                                                                                                                    received from being                        seniority, is the main characteristic
                                                                                                                    part of LDP                                determining which Diet representative
                                                                                                                    candidate's core                           affects year-to-year central government
                                                                                                                    support                                    transfers to Japan. Only marginal
                                                                                                                                                               representatives - the ones with the strong
                                                                                                                                                               incentives and perhaps the ones the LDP
                                                                                                                                                               would like to help most - appear to have
Horiuchi    2003    Japan      Electoral system FY1994-5       districts    per capita grants account settlement representative in    per capita income (log), Under Japan's parliamentary system, and
(2003)                         impact           (FY runs       (126)        from central to   (Chiho Zaisei      ruling coalition     total population (log), the single non-transferable vote electoral
                                                Apr to                      municipal         Kenkyu Kai),       measured by whether age over 65 population rule, strong evidence supporting
                                                March)                      government for various issues        he voted for the     ration, agriculture      "universalism within districts" or
                                                                            specific projects                    leader of the ruling sector workers ratio,    members of the opposition share
                                                                            (Kokko                               coalition in the                              distributive benefits within each district.
                                                                            shishutsukin,                        election for the                              Also, districts with more "powerful"
                                                                            differenced, log)                    Prime Minister,                               members of the ruling coalition receive a
                                                                                                                 seniority, num. of                            larger amount of grants than other
                                                                                                                 former bureaucrats,                           districts, namely the number of
                                                                                                                 num of former local                           attendances to Budget Committee,
                                                                                                                 officials, num. of                            seniority, and whether a representative is a
                                                                                                                 former local                                  former high-ranking official in local
                                                                                                                 politicians, num. of                          government are particularly important.
                                                                                                                 former secretaries                            (17)
                                                                                                COLLECTION
                                 CLASSIFICATION                                                                         POLITICAL
                                                    YEAR(S)                 DEPENDENT         INSTRUMENT FOR                               OTHER INDEPENDENT
   AUTHOR    YEAR     COUNTRY    (FROM REVIEW                 UNITS (N)                                               INDEPENDENT                                     FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                   ANALYZED                  VARIABLE            DEPENDENT                                     VARIABLES
                                    ESSAY)                                                                              VARIABLES
                                                                                                  VARIABLE

Horiuchi    2007    Japan       Political returns FY 1994 -   municipal per capita           account settlement   number of              municipality fiscal        "number of ruling-coalition members
(2007)                                            FY 1995     ities     amount of total      (Chiho Zaisei        incumbents who         strength index,            (divided by the district magnitude has a
                                                  (August     (3342)    transfers (general   Chosa Kenkyu         belonged to the        municipality population    negative (-0.030) and highly significant
                                                  1993-                 and specific         Kai, various         ruling coalition       (log), municipality        effect in the first-difference regression….
                                                  March                 subsidies) from      issues).             during each budget population density             Coalition leaders allocate, ceteris paribus,
                                                  1996)                 the central to       Municipality         cycle, divided by the (log), taxable income       disproportionately larger amounts to
                                                                        municipal            population at the    district magnitude. per capita (log), ratio of    districts with more opposition-party
                                                                        governments (in      end of previous      Dummy one if           population aged below      members to secure smooth operation of
                                                                        log)                 fiscal year from     incumbent is a         14 and of population       their rule." (21-22)
                                                                                             Kokudo Chiri         member of              over 65, ratio of          "...districts with powerful members of the
                                                                                             Kyokai (various      conferences for        number of persons          ruling coalition (that is, districts with
                                                                                             issues)              coalition leaders;     employed in the            Cabinet ministers, party leaders, and those
                                                                                                                  number of              agricultural sector        who attended the Budget Committee
                                                                                                                  attendances by ruling- against total number of    frequently) tend to receive the larger
                                                                                                                  party members at the employed persons, ratio      amount of per capita transfers, but these
                                                                                                                  Budget Committees. of the population in           effects are statistically insignificant." "The
                                                                                                                  number of seats        densely inhabited          number of seats per capita, a measure of
                                                                                                                  divided by the         districts                  malapportionment of seats in Japan’s
                                                                                                                  district population                               Lower House, is positive and significant
                                                                                                                  (logged)                                          in all regression estimates."(22)




Horiuchi and 2008   South Korea Swing versus      1993-2003 municipal total transfers per    Annual Report of vote share of             fiscal independency         "Winning president's vote share has a
Lee                             core                        ity/year  capita from            Local Finance      winning candidate       index, population           negative effect on total transfers, whereas
(2008)                                                      (864)     central                (1992-2002),                               density (logged), ratio     its squared term has a positive effect."
                                                                      government to          published by                               of population over 65       (15) The quadratic relationship between
                                                                      municipality           Ministry of                                (logged), agricultural      vote and money is highly significant
                                                                                             Government                                 workers per capita          before 1997, but not significant after.
                                                                                             Administration and                         (logged), manufacturing     Results consistent with their formal model
                                                                                             Home Affairs                               workers per capita          that shows incumbent president allocates
                                                                                                                                        (logged), dummy for         a larger amount not only in his own turf
                                                                                                                                        Seoul                       but also in rival's. (17)


Horiuchi and 2003   Japan       Programmatic      FY 1991 -   municipal total transfers      Chiho Zaisei     LDP's district seat        number of seats divided    "Our regression estimates suggest that in
Saito                           versus pork       FY 1994,    ities     from central         Chosa Kenkyu Kai share                      by district population     most (but not all) years, the LDP's seat
(2003)                                            FY 1995 -   (3300)    government to        (fn 21)                                     (logged, is a measure of   share had a negative effect on the subsidy
                                                  FY 1998               municipal                                                        apportionment),            allocation…we might interpret the
                                                                        government                                                       municipality fiscal        negative estimate as a consequence of the
                                                                        (logged)                                                         strength index             DLP's efforts to buy off marginal voters"
                                                                                                                                         (formula), population,     (677) Yet "after FY1994, the magnitude
                                                                                                                                         income, dependent          of the coefficient dropped and the effect
                                                                                                                                         population, industrial     became statistically insignificant"
Idema       2009    OECD        Swing versus      1991-2001 countries public tertiary   Education at a            percentage of higher total number of students     "This suggests that the higher the
(2009)              countries   core                        (14)      education         Glance, OECD              educated individuals enrolled as a % of           education level of swing voters of parties,
                                                                      spending per                                amongst core and       population, government     the more likely parties are to increase
                                                                      student as a                                swing voters of the partisanship,                 higher education spending when in
                                                                      percentage of per                           parties in             government                 office." (15)
                                                                      capita GDP                                  government, divided expenditures as a % of
                                                                                                                  by the average         GDP
                                                                                                                  percentage in
                                                                                                                  society, average level
                                                                                                                  of education of core
                                                                                                                  and swing voters,
                                                                                                COLLECTION
                                  CLASSIFICATION                                                                        POLITICAL
                                                     YEAR(S)                   DEPENDENT      INSTRUMENT FOR                                  OTHER INDEPENDENT
   AUTHOR     YEAR     COUNTRY    (FROM REVIEW                 UNITS (N)                                              INDEPENDENT                                         FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                    ANALYZED                    VARIABLE         DEPENDENT                                        VARIABLES
                                     ESSAY)                                                                             VARIABLES
                                                                                                  VARIABLE

Indridason   2005    Iceland     Political         1945-2000 cabinet        coalition        Morgunblaoio         prevalence of                                         "It can be hypothesized that in the
(2005)                           favoritism                                 formation and    (1945-2000           clientelistic politics,                               presence of clientelism, coalition
                                                                            termination      various issues),     political institutions,                               agreements will avoid restricting
                                                                                             pjooviljinn,         minority status                                       politicians' ability to serve their clients.
                                                                                             Grimsson (1977),                                                           This is consistent with the freedom in
                                                                                             Gislason (1993),                                                           appointment of non-cabinet positions, the
                                                                                             Jonsson (1969),                                                            use of the inner cabinet as a primary
                                                                                             Eggertsson (1999),                                                         solution mechanism, the general
                                                                                             Alpingi website                                                            vagueness of policy agreements, little
                                                                                             (www.althingi.is)                                                          emphasis on procedural rules, and the use
                                                                                                                                                                        of negative election rule." (458) "The
                                                                                                                                                                        presence of clientelism creates a demand
                                                                                                                                                                        for access to the discretionary distribution
                                                                                                                                                                        of public resources that is concentrated in
                                                                                                                                                                        the hands of the executive. Politicians
                                                                                                                                                                        therefore behave as if they were
                                                                                                                                                                        officeseekers rather than policy-seekers,
                                                                                                                                                                        regardless of their 'true' preferences."
                                                                                                                                                                        (462)


Johansson    2003    Sweden      Swing versus      1981-1995, municipal grants               1991 and 1994        difference between        per capita taxable          Municipalities with many swing voters get
(2003)                           core              1992-1995 ities                           Swedish Election     two blocs in last         income, share of young      larger transfers when using cutpoint
                                                              (255)                          Study, Statistics    election, cutpoint        (<19) inhabitants, share    density (significant at 10%, tested over
                                                                                             Sweden               density                   of old (>64)                1992-1995). Closeness of election has no
                                                                                                                                            inhabitants, population     statistically significant effect (whether
                                                                                                                                            density                     1981-1995 or 1992-1995).
John and     2001    England     Political returns 1981/2-     larger       grant related    former Department     unitary authority,       population, measures of     "We find that there is both targeting of
Ward                                               1995/6      English      expenditure      for the              Westminster election      average individual          extra resources to marginal seats and the
(2001)                                                         local        assessment and   Environment,         dummy, pre-               social need (factory        use of a scattergun approach that spreads
                                                               authoritie   standard         Transport, and the   Westminster election      analysis of census          benefit widely, especially when the
                                                               s (107)      spending         Regions              dummy, interaction        information)                government sees its re-elect under threat."
                                                                            assessment                            lead and                                              Targeting more important in the late
                                                                            (GREA/SSA)                            Westminster election                                  1980s. (331)
                                                                                                                  dummy, county
                                                                                                                  council election
                                                                                                                  dummy, London
                                                                                                                  borough election
                                                                                                                  dummy, control of
                                                                                                                  local authority, Tory
                                                                                                                  flagship, hung local
                                                                                                                  dummy


Kasara       2007    African     Political         1970-1987, crop-     tax rate on a crop Jaeger (1992) and      dummy if president's      scale for democracy,        "When a president comes from a crop-
(2007)               nations     favoritism        1966-1995 country                       McMillan (2001)        ethnic group comes        dummy if political          growing region, farmers face taxes that are
                                                              combinati                                           from a major crop         competition, average        higher by 23 points….changing our proxy
                                                              ons (50)                                            growing region,           duration of tenure of       for ethnic identification of
                                                                                                                  dummy if crop is          leaders, variant of black   farmers…slightly lowers the substantive
                                                                                                                  crown in the region       market premium,             size of the effect, but does not alter the
                                                                                                                  that dominates the        dummy if in CFA franc       basic finding that farmers associated with
                                                                                                                  governing coalition       zone, real per capita       the current regime face higher taxes"
                                                                                                                                            PPP GDP, dummy if           (167) Rulers are better at co-opting local
                                                                                                                                            country under IMF           allies in their home, which means there
                                                                                                                                            program, world price        are fewer alternative candidates to run
                                                                                                                                            for crop (logged),          against the rulers and the rulers can tax at
                                                                                                                                            percentage of total         a higher rate
                                                                                                                                            agricultural land area
                                                                                                                                            devoted to crop,
                                                                                                                                            percentage of total
                                                                                                                                            value of exports of
                                                                                                                                            crop, public investment
                                                                                                                                            as percent of GDP,
                                                                                                                                            distance from capital to
                                                                                                                                            a major town in the
                                                                                                                                            region
                                                                                              COLLECTION
                                  CLASSIFICATION                                                                    POLITICAL
                                                    YEAR(S)                  DEPENDENT      INSTRUMENT FOR                               OTHER INDEPENDENT
   AUTHOR     YEAR     COUNTRY    (FROM REVIEW                UNITS (N)                                           INDEPENDENT                                         FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                   ANALYZED                   VARIABLE         DEPENDENT                                     VARIABLES
                                     ESSAY)                                                                         VARIABLES
                                                                                                VARIABLE

Kauppi,     2004     European    Redistributive    1976-2001 region       country i's EU   European           political power          agricultural production,     "Our results indicate that at least 60% of
Widgén, and          Union       consequences                             budget share     Commission                                  GDP per capita versus        the budget expenditures can be attributed
Carrillo                                                                                                                               EU wide GDP per              to selfish power politics and the remaining
(2004)                                                                                                                                 capita                       40% to the declared benevolent EU
                                                                                                                                                                    budget policies. However, when we apply
                                                                                                                                                                    specific voting power measures that allow
                                                                                                                                                                    correlated preferences and cooperative
                                                                                                                                                                    voting patterns between the member
                                                                                                                                                                    states, our estimates indicate that the
                                                                                                                                                                    power politics view can explain as much
                                                                                                                                                                    as 90% of the budget shares. We conclude
                                                                                                                                                                    that power politics can explain a major
                                                                                                                                                                    part of the Council decisions and that
                                                                                                                                                                    correlated preferences and voting
                                                                                                                                                                    cooperation between EU countries
                                                                                                                                                                    potentially play significant roles in EU
                                                                                                                                                                    decision making." (223)



Keefer and   2009    India       Political returns 1999-2003 constitue actual spending     Indian             party stronghold (1 if   political, caste, religion   "party stronghold variable is a significant
Khemani                                                      ncies     incurred by the     constituency       party won every          fragmentation,               determinant of variation in
(2009)                                                       (483)     MP as a             development fund   election in the 90s),    constituency population      spending….MPLADS disbursements in
                                                                       percentage of       called Member of   candidate                density, proportion of       constituencies with dominant party are 10
                                                                       what the MP was     Parliament Local   stronghold, margin       villages with various        % points lower, more than one-half of the
                                                                       entitled to spend   Area Development   victory, 1993-99         public goods                 standard deviation in spending in the data.
                                                                       on public works     Scheme             spending, reserved                                    " (105) "Party stronghold effects are also
                                                                       in his or her       (MPLADS)           indicator for whether                                 robust to the control for a constituency's
                                                                       constituency                           a constituency is                                     average margin of victory." (107)
                                                                                                              reserved for
                                                                                                              candidates belonging
                                                                                                              to the scheduled
                                                                                                              castes and tribes,
                                                                                                              candidate switched



Kemmerling   2006    European    Redistributive         1999 countries allocation of    EU Commission         Effective Number of GDP, unemployment                 "As expected, GDP per capita has a
and                  Union       consequences                (71)      Structural Funds                       Parties (ENP),         rate, agriculture sector       negative and strongly significant impact
Bodenstein                                                             (SF) per capita                        presence of federalist size                           on the regional allocation of SF. The sign
(2006)                                                                                                        systems (FED).                                        of the coefficient of the unemployment
                                                                                                                                                                    rate (UNEMP) is positive but
                                                                                                                                                                    insignificant. Allocation of Agricultural
                                                                                                                                                                    Funds per capita (AGRAR/CAP) has a
                                                                                                                                                                    positive coefficient, suggesting that
                                                                                                                                                                    beneficiaries of Agricultural Funds also
                                                                                                                                                                    receive more SF....The share of left parties
                                                                                                                                                                    in the European elections at the regional
                                                                                                                                                                    level (LEFTPAR) and the share of
                                                                                                                                                                    Eurosceptic parties (ESCEP) have the
                                                                                                                                                                    expected positive signs and are also
                                                                                                                                                                    significant at the 10% level... Neither the
                                                                                                                                                                    federalism (FED) nor the multiparty
                                                                                                                                                                    system indicator (ENP) performs
                                                                                                                                                                    particularly well" (13)
                                                                                                  COLLECTION
                                   CLASSIFICATION                                                                        POLITICAL
                                                      YEAR(S)                   DEPENDENT       INSTRUMENT FOR                                OTHER INDEPENDENT
  AUTHOR       YEAR     COUNTRY    (FROM REVIEW                  UNITS (N)                                             INDEPENDENT                                         FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                     ANALYZED                    VARIABLE          DEPENDENT                                      VARIABLES
                                      ESSAY)                                                                             VARIABLES
                                                                                                    VARIABLE

Kemmerling    2002    Germany     Redistributive    1980,      cities (83) grant allocation   Statistical         congruence between        private capital stock,       "The main findings of our analysis are (i)
and Stephan                       consequences      1986, 1988                                Yearbook of         the local city            gross value added of a       the contribution of local public capital to
(2002)                                                                                        German Cities and   government and the        city's manufacturing         private production in cities is posit- ive
                                                                                              Municipalities      state ('Bundesland')      sector, infrastructure       and significant (ii) political affiliation,
                                                                                                                  government,               capital stock, Annual        measured by the coincidence of party
                                                                                                                  percentage                investment in                colour between state and local
                                                                                                                  difference of the two     infrastructure, labour       government, is decisive in explaining the
                                                                                                                  major parties, which      input, number of             distribution of investment grants across
                                                                                                                  are the parties SPD       working hours in the         cities (iii) cities with a prevalence of
                                                                                                                  and C여                    manufacturing sector,        'marginal voters' neither spend more on
                                                                                                                                            special grant-in-aid for     public infrastructure nor receive more
                                                                                                                                            investments, financial       investment grants from higher-tier
                                                                                                                                            situation of a city, trade   governments (iv) redistributive concerns
                                                                                                                                            tax revenues, number of      of higher-tier governments matter for the
                                                                                                                                            (four-wheel) motor           allocation of grants, whereas efficiency
                                                                                                                                            vehicles, number of          considerations appear to be less
                                                                                                                                            manufacturing firms,         important." (421)


Khemani       2004a   India       Political         1960-1992 major          total spending on Reserve Bank of    electoral cycle                                        "we find that politicians manipulate fiscal
(2004a)                           business cycles             states         public services India Bulletin                                                              policies before elections to provide
                                                              (14)                                                                                                       targeted favors to specific interest groups,
                                                                                                                                                                         possibly in exchange for campaign
                                                                                                                                                                         support. In election years in the Indian
                                                                                                                                                                         states, there is evidence of small
                                                                                                                                                                         manipulations of fiscal instruments to
                                                                                                                                                                         target benefits to narrow interest groups,
                                                                                                                                                                         such that there is no net effect on the
                                                                                                                                                                         overall fiscal deficit." (151-2)


Khemani       2007    India       Redistributive    1972-1995 major          disaggregated    Reserve Bank of     Political affiliation     real state income per        "Plan grants are directed toward affiliated
(2007)                            consequences                states         intergovernmenta India Bulletin      dummy (=1 if center       capita, total population,    states" especially where the proportion of
                                                              (15)           l transfers: (1)                     and state                 state election cycle,        seats is small -- "greater transfers are
                                                                             statutory                            governments belong        state government             directed towards those states where the
                                                                             transfers                            to the same party),       coalition dummy,             national ruling party has more seats to
                                                                             (Finance                             affiliation*seats         interaction of affiliation   gain, rather than to the core support states
                                                                             Commission),                         (proportion of seats      after 1990 (national         where it already controls a high
                                                                             (2) plan grants                      controlled by             coalition)                   proportion of seats to the national
                                                                             (Planning                            national ruling                                        legislature" (21, 22). "On the other hand,
                                                                             Commission),                         party), affiliation*(1-                                statutory transfers...are [lower]" for
                                                                             (3) central                          seats), Congress                                       affiliated states, perhaps to correct for
                                                                             schemes                              Party dummy                                            politically-motivated plan transfers (22).
                                                                             (discretionary)                                                                             Central scheme transfer results are
                                                                                                                                                                         insignificant except for a "very restricted
                                                                                                                                                                         political variable" (25). Suggests district-
                                                                                                                                                                         level is more appropriate for central
                                                                                                                                                                         scheme transfers


Khemani       2003    India       Swing versus      1972-1995 states         transfers per      Reserve Bank of   affiliation, seat         state population,            "We find that plan grants are directed
(2003)                            core                        (15)           capita (statutory, India Bulletin                              income                       towards affiliated states as the model
                                                                             plan grants, and                                                                            predicts, and furthermore towards those
                                                                             central themes)                                                                             affiliated states that contribute a smaller
                                                                                                                                                                         share of the total number of state
                                                                                                                                                                         legislators in the national assembly for the
                                                                                                                                                                         national ruling party." (21) National
                                                                                                                                                                         political parties target greater resources to
                                                                                                                                                                         those states where they have more seats to
                                                                                                                                                                         gain in the national legislature, rather than
                                                                                                                                                                         to states that are their core support bases,
                                                                                                                                                                         and hence pivotal for winning a majority."
                                                                                                                                                                         (26)
                                                                                                         COLLECTION
                                    CLASSIFICATION                                                                               POLITICAL
                                                         YEAR(S)                     DEPENDENT         INSTRUMENT FOR                                 OTHER INDEPENDENT
  AUTHOR      YEAR     COUNTRY      (FROM REVIEW                     UNITS (N)                                                 INDEPENDENT                                          FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                        ANALYZED                      VARIABLE            DEPENDENT                                       VARIABLES
                                       ESSAY)                                                                                    VARIABLES
                                                                                                           VARIABLE

Kneebone     2001    Canada        Political           1966-1997 province change in         CANSIM (from                   change in provincial     change in                     "One of our key results is the finding of a
and                                business cycles               (290)    provincial        matrices 2782                  expenditure lagged,      unemployment rate,            clear electoral cycle in provincial
McKenzie                                                                  expenditure (as a through to 2791)               dummy for year with      change in Canada-US           revenues. In election years, the
(2001)                                                                    fraction of GDP)                                 provincial election,     nominal exchange rate,        predilection of provincial governments of
                                                                                                                           dummy for Liberal        change in short-term          all political stripes to increase taxes is
                                                                                                                           Party, New               real interest rate,           temporarily halted…. Opportunistic
                                                                                                                           Democratic Party,        change in ratio of debt       responses with respect to provincial
                                                                                                                           Progressive              to GDP over previous 2        spending are equally widespread but the
                                                                                                                           Conservative Party       fiscal years                  direction of the response is less
                                                                                                                           and Social Credit                                      homogenous across political parties. The
                                                                                                                           Party                                                  prevailing tendency is to decrease
                                                                                                                                                                                  spending in Health, Social Services and
                                                                                                                                                                                  Industrial Development in election years
                                                                                                                                                                                  vis-a-vis non-election years, while
                                                                                                                                                                                  spending in Education, transportation, and
                                                                                                                                                                                  Communication, and Recreation and
                                                                                                                                                                                  Culture increases in election years versus
                                                                                                                                                                                  non-election years." (771)


Knoesen      2009    South Africa Political returns          2000 municipal amount of                                      vote share of African                                  Electoral support for ANC in 2000
(2009)                                                            ities     transfer                                       National Congress                                      elections may help explain variation in
                                                                            (regression                                    (ANC)                                                  ANC's stated objective of poverty
                                                                            analysis in                                                                                           reduction and its actual allocation.
                                                                            progress)
Kramon       2009    Kenya         Political returns         2002 individual whether          Afrobarometer                whether individual       individual                    Estimates suggest that a vote-buying
(2009)                                                            (1278)     individual voted Round 3 survey               was approached by a      characteristics such as       attempt increases the probability of voting
                                                                             or not                                        political party          education, cash income,       by about 10 %-tage points.One's
                                                                                                                           representative and       gender, urban, age            assessment of the intrinsic value of
                                                                                                                           been offered a bribe                                   democracy plays a role in determining an
                                                                                                                           or gift in exchange                                    individual's decision to vote.(11) Those
                                                                                                                           for a vote, vote                                       approached by party are about 14
                                                                                                                           margin, democratic                                     percentage points more likely to vote than
                                                                                                                           belief, political                                      those who do not. In the Kenyan case,
                                                                                                                           efficacy                                               vote buying view potentially swing voters
                                                                                                                                                                                  as acceptable targets. Also, those with
                                                                                                                                                                                  low level of education is likely to be most
                                                                                                                                                                                  influenced. (25)

Kramon and   2011    Ghana,        Political           2003          househol childhood               Demographic and      indicator for whether    household head's age,         Findings consistent with our that that the
Posner               Kenya,        favoritism          Ghana,        d        vaccinations,           Health Surveys       household in             literacy, religious, rural,   answer to the "are ruling party supporters
(2011)               Malawi,                           Kenya;                 educational                                  question is located in   household assets,             favored?" question depends on the
                     Zambia                            2004                   attainment,                                  a core district of the   material used to make         outcome one happens to be investigating.
                                                       Malawi;                access to                                    ruling party (60% of     household floor,              i.e. in Kenya, those who had voted for
                                                       2007                   improved water                               voters)                  distance from the             KANU received more childhood
                                                       Zambia                 sources,                                                              country's capital of the      vaccinations but less access to water. (5)
                                                                              household                                                             sampling cluster              Also results show "who is favored?"
                                                                              electrification                                                                                     (targeting) depends on "favored relative to
                                                                                                                                                                                  whom?." (7)
Kudamatsu            28 African    Redistributive      latest        individual   individual-level Demographic and         dummy for                mother fixed effects,         "Democratization since 1990 is followed
(2012)               countries     consequences        survey per    baby by      infant mortatiliy Health Surveys         democratic year          mother's birth cohort by      by a reduction in the infant mortality rate
                                                       country       mother       dummy for                                                         child's birth year fixed      by 1.2 percentage points, which is as
                                                       since 1996-   (643846)     whether a child                                                   effects, dummy for            much as 12% of the sample mean" (16).
                                                       2006                       dies before age                                                   girls, multiple births,       "The probability that a mother sees her
                                                                                  one.                                                              and birth order               baby die within the first month of life falls
                                                                                                                                                                                  by 0.6 percentage points after
                                                                                                                                                                                  democratization" (16).

Kunicová and 2003    Russia        Programmatic        1994-2003 individual       dissent of a        electronically       electoral mandate        bill and deputy               "We found evidence that particularly in
Remington                          versus pork                   voters in        deputy with his     recorded roll-call   (SMD/PL), parties,       characteristics               the second and third Dumas, SMD
(2003)                                                           Russian          faction leader on   votes produced by    faction, faction                                       deputies were more likely to break with
                                                                 State            distributive        Moscow-based         switching                                              faction discipline than were their list-
                                                                 Dumas            issues              firm INDEM                                                                  based comrades on distributive issues, and
                                                                                                                                                                                  that this effect grew stronger after the first
                                                                                                                                                                                  Dumas." Results for voting on issues that
                                                                                                                                                                                  can be characterized as non-distributive
                                                                                                                                                                                  were at variance with expectations, but
                                                                                                                                                                                  confined to the first Dumas. (29)
                                                                                                  COLLECTION
                                  CLASSIFICATION                                                                          POLITICAL
                                                     YEAR(S)                   DEPENDENT        INSTRUMENT FOR                                OTHER INDEPENDENT
   AUTHOR    YEAR     COUNTRY     (FROM REVIEW                  UNITS (N)                                               INDEPENDENT                                      FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                    ANALYZED                    VARIABLE           DEPENDENT                                      VARIABLES
                                     ESSAY)                                                                               VARIABLES
                                                                                                    VARIABLE

Kwon        2005    South Korea Political          1988-97     Election     Election timing:   National Statistical Election timing: pre-   Election timing: GDP       "Korean government boosted government
(2005)                          business cycles                timing:      government         Office               election period         (log), unemployment,       expenditure in accordance with the ruling
                                                               quarters     expenditure (as                         distribution of         quarter dummy              party's electoral concerns. While the
                                                               (40)         percentage of                           national subsidies:     distribution of national   results of the analysis clearly show
                                                               distributi   GDP, log)                               "Pro-votes" -- vote     subsidies: total revenue   substantively significant effects of pre-
                                                               on of        distribution of                         share of ruling party   minus national             electoral manipulation of government
                                                               national     national                                candidate in each       subsidies, gross           expenditure, the consequences of election-
                                                               subsidies:   subsidies                               province, difference    regional domestic          induced public expenditure cycles may be
                                                               provinces                                            between the             product (grdp), number     even greater." (333) "The results of the
                                                               (13)                                                 proportion of votes     of farm households,        electoral impact show that Cox-
                                                                                                                    received by the         population (logged)        McCubbins' prediction of resource
                                                                                                                    leading two                                        distribution to core supporter groups did
                                                                                                                    candidates (both for                               not fare well with the Korean data. The
                                                                                                                    the presidential                                   coefficient for pro-ruling party vote shares
                                                                                                                    election due to data                               did not reach statistical significance. In
                                                                                                                    limitations)                                       contrast, the Lindbeck-Weibull/Dixit-
                                                                                                                                                                       Londregan prediction is consistent with
                                                                                                                                                                       the empirical pattern of the allocation of
                                                                                                                                                                       national subsidies in Korea. The point
                                                                                                                                                                       estimates of electoral margin shows a
                                                                                                                                                                       negative sign with statistical significance
                                                                                                                                                                       at the p<0.05 level" (337-338)

Lake and            multi-       Redistributive    various     countries education             World Bank          level of democracy       per capita GNP, land       Level of democracy is positively
Baum                country      consequences      years 1975, (approx. (illiteracy, pupil-                        as measured by           area, population,          significant with the level of public goods
(2001)                                             1985,       55 to 85) teacher ratio,                            contestability of the    percent of urban           provision at the 0.05 level or better in
                                                   1987,                 school                                    political market and     population, OECD           84% of the 38 OLS regressions reported
                                                   1990, 1992            enrollment, etc.)                         lags                     dummy                      (609). In general, impact of regime type
                                                                         and public health                                                                             on public services is strongest within 2
                                                                         (health care                                                                                  years or less.
                                                                         access, clean
                                                                         water access, no.
                                                                         of physicians,
                                                                         unintended
                                                                         births, infant
                                                                         mortality etc.)
Lancaster   1986    multi-       Programmatic                  countries indicators
                                                                         level of pork         interviews and    electoral systems                                     "inverse relationship exists between
(1986)              country      versus pork                             activity (no          secondary sources                                                       district size and pork activity - the larger a
                                                                         regression            i.e. Aberbach                                                           nation's ratio of representatives to
                                                                         analysis)             (1981),                                                                 electoral districts, the less pork barrel
                                                                                               Loewenberg and                                                          activity."(70) "less pork allocation occurs
                                                                                               Patterson (1979)                                                        in countries with multimember districts
                                                                                                                                                                       can be logically extended to countries
                                                                                                                                                                       with a single at-large electoral district."
                                                                                                                                                                       (71) U.S., Canada, New Zealand, India,
                                                                                                                                                                       and Britain with single-member district
                                                                                                                                                                       system, have the greatest likelihood for
                                                                                                                                                                       pork barrel activity. Countries with multi
                                                                                                                                                                       member district such as Columbia,
                                                                                                                                                                       Greece, Norway, and Spain have lesser
                                                                                                                                                                       propensity. Number of representatives per
                                                                                                                                                                       district in multi-member districts affects
                                                                                                                                                                       the accountability linkage and influences
                                                                                                                                                                       pork barrel politics. Multi-member
                                                                                                                                                                       systems with fewer members per district
                                                                                                                                                                       are more likely to have pork barrel activity
                                                                                                                                                                       than those with larger districts. (71)
                                                                                                  COLLECTION
                                 CLASSIFICATION                                                                         POLITICAL
                                                   YEAR(S)                    DEPENDENT         INSTRUMENT FOR                              OTHER INDEPENDENT
   AUTHOR    YEAR     COUNTRY    (FROM REVIEW                 UNITS (N)                                               INDEPENDENT                                     FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                  ANALYZED                     VARIABLE            DEPENDENT                                    VARIABLES
                                    ESSAY)                                                                              VARIABLES
                                                                                                    VARIABLE

Leigh       2008    Australia   Political returns 2001-2004 seats          total funding per Department of     National party,            population density,       Distribution of funding under 4 programs
(2008)                                                      (150)          electorate ($m) Transport and       Liberal party,             quadratic in income       strongly skewed towards electorates that
                                                                                             Regional Services income                                               were held by the Coalition government in
                                                                                             and Australian                                                         2001. Compared with non-government
                                                                                             Local Government                                                       seats, seats held by National Party or
                                                                                             Association 2004                                                       Liberal Party received more.(295)
                                                                                                                                                                    "Compared with non-government seats,
                                                                                                                                                                    National Party electorates tend to have an
                                                                                                                                                                    additional 5.2 Stronger Families &
                                                                                                                                                                    Communities grants, and an additional 3.4
                                                                                                                                                                    Regional Partnership grants. For Liberal
                                                                                                                                                                    Party electorates, figures are additional
                                                                                                                                                                    3.4 and 1.7.


Letha Kannan 2009   India       Political returns 1980-2008 states         incumbent vote Election                Fiscal policy           literacy, urbanization,   "Our results indicate that voters in the
(2009)                                                      (24)           share obtained by Commission           variables including     population, schedule      Indian states seem to be fiscal
                                                                           the party of the                       various taxes, non-     class population and      conservatives as when they do react to
                                                                           incumbent chief                        tax revenues and        real SD, vote share       fiscal policy; they punish the incumbents
                                                                           minister at the                        expenditures of state   obtained by incumbent     for increases in expenditures or
                                                                           state in sub-                          governments             in previous election      taxes....Voters in general do not seem to
                                                                           national elections                                                                       respond to government revenue
                                                                                                                                                                    expenditures." (87)


Limosani and 2001   Italy       Political         1969-1994 municipal      public spending     ISAT database      revenues deriving                                 "in pre-election periods, national
Navarra                         business cycles             ities          in goods and                           from grants by the                                representative try to obtain investment
(2001)                                                      (75=3regi      services and                           central government,                               expenditures for specific projects to be
                                                            ons*25ye       public spending                        revenues from                                     spent in their electoral districts with the
                                                            ars)           in capital goods                       creating a loan,                                  help of "friendly" local governments in
                                                                           and investment                         binary variable for                               order increase their popular consent."
                                                                           in 3 Italian                           elections dates                                   (323)"This result shows the existence of a
                                                                           regions                                assuming the value                                positive dependent relationship between
                                                                           (Lombardia,                            of 1 in case of                                   the amount of grants from the central
                                                                           Campania,                              election                                          government and both consumption and
                                                                           Sicillia)                                                                                investment spending and reveals the
                                                                                                                                                                    scarce autonomy of a financially assisted
                                                                                                                                                                    local institutions… heterogeneity in
                                                                                                                                                                    regional backgrounds does not play a
                                                                                                                                                                    crucial role in influencing local public
                                                                                                                                                                    spending patterns. Loans seem to
                                                                                                                                                                    influence only expenditures in goods and
                                                                                                                                                                    services and do not have any impact on
                                                                                                                                                                    investments." (321-2)


Lindberg and 2008   Ghana       Political returns 1996 and    individual   voting rationale    survey conducted                           demographics              "Our survey results suggest that voting
Morrison                                          2000        (690         (reason for         by authors from                                                      behavior is not so much ethnic in essence
(2008)                                            elections   interview    voting for a        June and July,                                                       as it is a rational response to the classic
                                                              ees from     political           2003                                                                 information problem for the voter using
                                                              6            representative                                                                           the parties' historical affiliations as a key
                                                              constitue    asked in an open-                                                                        source of evaluation."
                                                              ncies)       ended question                                                                           "clientelistic and ethnically predisposed
                                                                           in survey; non-                                                                          voting is a minor feature of Ghanaian
                                                                           regression                                                                               elections and that when present, it seems
                                                                           analysis)                                                                                to be nurtured by intense partisan
                                                                                                                                                                    competition." (121-122)
                                                                                                   COLLECTION
                                   CLASSIFICATION                                                                          POLITICAL
                                                     YEAR(S)                   DEPENDENT         INSTRUMENT FOR                                  OTHER INDEPENDENT
   AUTHOR      YEAR     COUNTRY    (FROM REVIEW                 UNITS (N)                                                INDEPENDENT                                    FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                    ANALYZED                    VARIABLE            DEPENDENT                                        VARIABLES
                                      ESSAY)                                                                               VARIABLES
                                                                                                     VARIABLE

Lizzeri and   2004    Britain     Political         19th       local    franchise in 19th                            changes in English                               "We have presented a model of voluntary
Persico                           favoritism        century    governme century England                              public finances                                  franchise extension that is based on
(2004)                                                         nt       that was                                     during age of reform                             divisions within the elite. In our model
                                                                        extended                                                                                      these divisions arise endogenously
                                                                        progressively at                                                                              through the political process: groups
                                                                        the parliamentary                                                                             within the elite benefit differently from the
                                                                        level with three                                                                              status quo. We have shown that it is
                                                                        Reform Acts of                                                                                possible that a majority within the elite
                                                                        1832, 1867,                                                                                   may favor expanding the franchise.... In
                                                                        1884 (historical                                                                              England, substantial subgroups within the
                                                                        analysis)                                                                                     elite championed franchise expansion.
                                                                                                                                                                      Our analysis can account for some of the
                                                                                                                                                                      changes in the scope of public sector
                                                                                                                                                                      intervention during the British age of
                                                                                                                                                                      reform, notably the increased provision of
                                                                                                                                                                      public goods and the decline of special
                                                                                                                                                                      interest politics. In our model, these
                                                                                                                                                                      changes are a consequence of the reform
                                                                                                                                                                      process; reform takes place when
                                                                                                                                                                      increased needs for public goods lead to a
                                                                                                                                                                      majority of the elite to demand a
                                                                                                                                                                      redirection of the role of government
                                                                                                                                                                      away from special interest politics toward
                                                                                                                                                                      increased provision of the public
                                                                                                                                                                      goods."(755)



Luna and      2010    Chile       Redistributive    2000-2008 municipal 8 types of social National System of         electoral results for 5   municipality's         Overall programs effectively target poor
Mardones                          consequences                ity (1911) transfers and     Municipal                 bigger parties for 3      dependence on          communities with higher social (education
(2010)                                                                   targeted          Indicators                elections at              Municipal Common       and health) needs. (11) "even in context of
                                                                         subsidies (i.e.                             municipal level           Fund, municipality's   efficient socioeconomic targeting and
                                                                         family, drinkable                           (1996, 2001, 2004),       health and education   fiscal oversight, political targeting is also
                                                                         water and                                   dummy for reelected       needs, bureaucratic    present; and, political investment
                                                                         sewerage,                                   mayors, participation     capacity, year         rationales in Chile diverge from the
                                                                         pension,                                    rate, competitiveness                            standard predictions that have been
                                                                         education,                                  of the race                                      derived from machine party ridden and/or
                                                                         health)                                                                                      weak states cases. (2)


Manacorda,    2010    Uruguay     Political returns 2005-2008 househol      political support   two surveys          PANES (Plan de       age, gender, years of       Beneficiaries of a large government anti-
Miguel, and                                                   ds            for FA (Frente      conducted by         Atencion Nacional a education                    poverty program, PANES, showed 11-
Vigorito                                                      (approx.      Amplio)             authors in           la Emergencia                                    14% increase in support for the current
(2010)                                                        3000)         government          collaboration with   Social, anti-poverty                             government than non-beneficiaries. (13)
                                                                                                Veronica             program) program                                 "These effects last into the post-program
                                                                                                Amarante and         effects                                          period." (27)
                                                                                                Economics and
                                                                                                Sociology
                                                                                                Departments at the
                                                                                                Univ. of the
                                                                                                Republic, Ministry
                                                                                                of Social
                                                                                                Development staff
                                                                                                (conducted in
                                                                                                2006-2007 and
                                                                                                2008)
                                                                                                     COLLECTION
                                    CLASSIFICATION                                                                            POLITICAL
                                                       YEAR(S)                   DEPENDENT         INSTRUMENT FOR                             OTHER INDEPENDENT
   AUTHOR      YEAR     COUNTRY     (FROM REVIEW                  UNITS (N)                                                 INDEPENDENT                                FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                      ANALYZED                    VARIABLE            DEPENDENT                                   VARIABLES
                                       ESSAY)                                                                                 VARIABLES
                                                                                                       VARIABLE

McGillivray   1997    United       Swing versus      U.S. 1979, district      amount of           based on 1963         district comparative District Unemployment   "marginal electorates do appear to secure
(1997)                States and   core              Canada     (Electoral    protection          MUSIC                 advantage, district                          more favorable levels of protection in
                      Canada                         1970       district      measured by         classification        marginality, number                          Canada's high party discipline system. In
                                                                model),       tariff protection   scheme, a             of districts in State,                       the U.S., a low party discipline system,
                                                                industry                          Standard Industrial   district party                               safe districts with senior representatives
                                                                (industry                         Classification        representative,                              receive slightly higher levels of
                                                                district                          (SIC) based import    district represented                         protection." (603)
                                                                model)                            code and SIC 1972     by a minister,
                                                                                                  code.                 marginal districts
                                                                                                                        represented by a
                                                                                                                        Minister, seniority of
                                                                                                                        representative,
                                                                                                                        member of Ways and
                                                                                                                        Means Committee,
                                                                                                                        Member of Foreign
                                                                                                                        and Interstate
                                                                                                                        Commerce
                                                                                                                        Committee


McIntosh and 2009     Uganda       Swing versus            2001 district      loans and           FINCA/Uganda,surprise vote share,         linear time trend      "we find evidence that investment jumps
Allen                              core                                       spending            the country’sborrower loan cycle,                                immediately following the presidential
(2009)                                                                                            largest      violence, threats                                   election in a manner correlated with the
                                                                                                               against
                                                                                                  microfinance lend                                                size of the surprise support for the victor,
                                                                                                               opposition/movemen                                  Yoweri Museveni." The surprise vote
                                                                                                               t/citizen, election                                 share has a significant effect on both
                                                                                                               results contested,                                  borrowing and savings behavior. The
                                                                                                               close vote,                                         implication is that counties that backed
                                                                                                               movement promises                                   Museveni to a surprising extent saw
                                                                                                               patronage,                                          significant increases in both investment
Miguel        2004    Kenya and    Redistributive    Kenya:     villages various local      Busia, Kenya:      local ethnic diversity       average years of       "The school funds result implies that the
(2004)                Tanzania     consequences      1995-1996 (66 in     school funding: 1996 ICS School (ethnolinguistic                  education, proportion change from ethnic homogeneity to
                                                     (well data Meatu,    annual school     and Pupil          fractionalization,           formal sector          average levels of diversity in Busia, Kenya
                                                     from 2000- Tanzania) spending per      Questionnaires,    ELF), interaction            employment, proportion is associated with a drop of approximately
                                                     1)         and       pupil, desks per 1996 Government term for difference              homes with iron roofs, 25 percent in average funding - a large
                                                     Tanzania: primary pupil, latrines per Examinations        between the impact           proportion households effect - while the estimated effect for
                                                     1997-2002 school     pupil, classrooms Namelists, Global of ethnic diversity on        grow cash crops,       Meatu, Tanzania, is positive and
                                                                communi per pupil,          Positioning        the public goods             proportion households statistically significant." (353) "local
                                                                ty which proportion wells Systems (GPS)        outcome in Kenya             own cattle, proportion ethnic diversity has a significantly more
                                                                includes with normal        readings taken by versus Tanzania               Catholic               negative effect on local public goods
                                                                area      flows             NGO field-                                                             provision in Busia, Kenya, than in Meatu,
                                                                within 5                    workers, survey on                                                     Tanzania." (357)
                                                                km of a                     well conditions by
                                                                primary                     NGO field-workers
                                                                school                      for this study.
                                                                (84 in                      Meatu, Tanzania:
                                                                Busia,                      2001-2 household
                                                                Kenya)                      village council
                                                                                            surveys


Miguel and    2003    Ghana        Swing versus      1996 and     schools     total school        World Bank's pilot dummy if NCD           household expenditure At the administrative district level:
Zaidi                              core              2000         (192)       funding, and        Public Expenditure (national ruling       per capita, proportion "Schools in administrative districts where
(2003)                                               elections,               logged total        Tracking Survey    party) swept all       urban, years of          the NDC swept the parliamentary
                                                     school                   school funding                         parliamentary seats    education for head of    elections received 27 percent higher
                                                     funding in                                                      in an administrative   household, proportion funding per pupil than other schools",
                                                     1998 -                                                          district in 1996       school-aged (5-15)       which is significant at the 99 percent level
                                                     1999                                                            election, average      children who have        (16) Negative and significant coefficient
                                                                                                                     margin of victory in   attended school in last on quadratic of average margin of victory
                                                                                                                     1996, vote share of    year, household          (17) Patronage and possibly public choice
                                                                                                                     NCD                    educational              effects. At parliamentary seat level : null
                                                                                                                                            expenditures per school- results for political variables
                                                                                                                                            aged child, proportion
                                                                                                                                            individuals who have
                                                                                                                                            received vaccinations,
                                                                                                                                            religion
                                                                                                                                            fractionalization index,
                                                                                                                                            ethno-linguistic
                                                                                                                                            fractionalization index
                                                                                                    COLLECTION
                                      CLASSIFICATION                                                                          POLITICAL
                                                         YEAR(S)                  DEPENDENT       INSTRUMENT FOR                                  OTHER INDEPENDENT
   AUTHOR       YEAR     COUNTRY      (FROM REVIEW                 UNITS (N)                                                INDEPENDENT                                        FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                        ANALYZED                   VARIABLE          DEPENDENT                                        VARIABLES
                                         ESSAY)                                                                               VARIABLES
                                                                                                      VARIABLE

Milesi-        2002    OECD and      Programmatic      1991-94     countries total primary       OECD Economic          average district      population share over          Empirical results consistent with their
Ferretti,              Latin         versus pork                   (40)      spending/GDP        Outlook, IMF           magnitude (AM),       65, GDP per capita,            model where a more proportional system
Perotti, and           American                                              ratio               Government             standardized district OECD dummy                     can have an ambiguous impact on total
Rostagno               countries                                                                 Financial Statistics   margin (SM),                                         primary government spending, depending
(2002)                                                                                                                  average deviation                                    on relative strength of its effects on
                                                                                                                        from proportionality                                 transfer and public good spending. For
                                                                                                                                                                             OECD countries, positive relationship
                                                                                                                                                                             between LogSM and transfers survives,
                                                                                                                                                                             but in Latin America there is no bivariate
                                                                                                                                                                             relation between proportionality and
                                                                                                                                                                             transfers. (634)
Milligan and 2005      Canada        Swing versus      1988-2001 electoral spending through      Atlantic Canada        government seat,     economic controls               That political factors explain some portion
Smart                                core                        districts agencies              Opportunities          vote margin,                                         of the variation in regional development
(2005)                                                           in 5                            Agency, Economic       government seat *                                    assistance that economic factors cannot.
                                                                 regions/y                       Development            vote margin, cabinet                                 Positive and significant effects for
                                                                 ear                             Agency of Canada       minister, years of                                   government seat and cabinet minister
                                                                 (1344)                          for Quebec             experience by MP,                                    variable.
                                                                                                 Regions


Min            2009    India         Redistributive    1992-2003 villages electrification of     satellite images       indicator for village   village population,          "When it comes to village electrification,
(2009)                               consequences                in Uttar village                from Defense           represented by a BSP    presence of school or        the differences across parties are
                                                                 Pradesh                         Meteorological         (Bharatiya Samaj        medical facility, literacy   substantial, and the most positive effects
                                                                 (approx.                        Satellite Program's    Party) legislator or    rate, distance to nearest    are in villages located in BSP
                                                                 1.2                             Operational            BJP (Bharatiya          town, paved approach         constituencies." (24) "Using matching
                                                                 million =                       Linescan System        Janata Party) in the    road, income index,          technique to evaluate similar villages that
                                                                 approx.                                                State Assembly,         nighttime light output       differ only on whether they switched to
                                                                 98,000                                                 population classified   in constituency (1992)       BSP representation in the critical 2002
                                                                 villages                                               as Scheduled Caste                                   elections, I also show a positive BSP
                                                                 *14                                                    according to 2001                                    treatment effect." (34)
                                                                 years)                                                 Indian Census,
                                                                                                                        interaction term of
                                                                                                                        the two, other party
                                                                                                                        indicators
                                                                                                                        Samajwadi Party,
                                                                                                                        Congress


Nakayama       2007    France, Italy, Political returns 1950s-                 credit allocation archives (non-         power balance           one-party dominance          "In Japan, as well as Italy, the MPs of the
(2007)                 Japan                            1970s                  patterns from     regression             between the             over long period in all      dominant parties seized every chance to
                                                                               central to local analysis)               bureaucracy and the     three countries used as      get more credit for their constituencies'
                                                                               government                               parties, size and       a control                    investment programs, and all these
                                                                                                                        viability of local                                   clientelist practices helped them develop
                                                                                                                        governments                                          and consolidate their local electoral bases.
                                                                                                                                                                             In contrast the [French] Gaullist regime
                                                                                                                                                                             was based on the exclusion of any kind of
                                                                                                                                                                             pork barrels, including those for the
                                                                                                                                                                             Gaullist party's own MPs." (3)
                                                                                                                                                                             "Gaullists' hegemony essentially relied on
                                                                                                                                                                             the capacity of the bureaucracy to
                                                                                                                                                                             neutralize them and integrate local
                                                                                                                                                                             political forces irrespective of their party
                                                                                                                                                                             labels.... to perpetuate Gaullist
                                                                                                                                                                             dominance, the government needed to
                                                                                                                                                                             keep allocation patterns from their own
                                                                                                                                                                             Majority MPs demanding pork barrels --
                                                                                                                                                                             but failed to do this in the late 1960s and
                                                                                                                                                                             1970s" (35)
                                                                                                 COLLECTION
                                  CLASSIFICATION                                                                       POLITICAL
                                                    YEAR(S)                    DEPENDENT       INSTRUMENT FOR                              OTHER INDEPENDENT
   AUTHOR     YEAR     COUNTRY    (FROM REVIEW                 UNITS (N)                                             INDEPENDENT                                       FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                   ANALYZED                     VARIABLE          DEPENDENT                                    VARIABLES
                                     ESSAY)                                                                            VARIABLES
                                                                                                   VARIABLE

Neto and     2003    Brazil      Political         1985 to     bills (269   who proposes       PRODASEN          who proposes what :                                  "our statistical findings suggest that
Santos                           favoritism        1999 and    bills        what : probability                   concentration or                                    dominant deputies, owing to their close
(2003)                                             1995        proposed     of a deputy                          dominance of a                                      links to voters, are less concerned with
                                                               from         proposing at                         deputy's pattern of                                 national policy-making than are deputies
                                                               1985-99      least one                            vote distribution,                                  who tend to share their constituencies
                                                               and 1291     parochial or                         affiliation with                                    with other politicians. "Furthermore, the
                                                               bills in     national-level                       National Renewal                                    incentive to delegate national policy-
                                                               1995)        bill in 1995; who                    Alliance Party/Social                               making to the executive, as expected by
                                                                            gets bill passed:                    Democratic Party                                    the ISM, is, on the one hand, stronger
                                                                            probability of a                     (AREANA/PDS)                                        among supporters of the president." (467)
                                                                            deputy getting at                    who gets bill passed:                               "The fact that the variable tapping
                                                                            least one law                        affiliation with                                    whether or not deputies have held an
                                                                            passed per                           ARENA/PSD, career                                   office in the Chamber's power structure
                                                                            legislature                          length, electoral                                   has a significant effect on their likelihood
                                                                                                                 power, held                                         of getting a law passed is evidence that
                                                                                                                 legislative office                                  there is an important distributive
                                                                                                                 (members of Mesa                                    component in lawmaking, as far as
                                                                                                                 da Camara dos                                       deputies' legislative output is concerned."
                                                                                                                 Deputados), held                                    (473)
                                                                                                                 executive office




Nichter      2008    Argentina   Swing versus      survey      voters       received rewards survey conducted    Peronist                Income, education,          Contrary to conventional wisdom about
(2008)                           core              conducted   (1920)       by a party in the by Stokes (2005)   sympathizer, opinion    housing quality, log        vote buying, parties can activate their own
                                                   in Dec                   last campaign                        of Peronists,           population, ballot, age,    passive constituencies by rewarding
                                                   2001-Jan                                                      Peronists rating,       gender                      unmobilized supporters for turnout.
                                                   2002                                                          1995/9 Peronist                                     Turnout buying involves a less stringent
                                                                                                                 voter                                               monitoring requirement than vote buying -
                                                                                                                                                                     the ability to observe turnout instead of
                                                                                                                                                                     voting decisions. Formal models show
                                                                                                                                                                     turnout buying is incentive-compatible.
                                                                                                                                                                     (29)
Ortega and   2008    Venezuela   Political returns 2000 and    municipal total number of     National Electoral total expenditure on     Human development           "Mercal does not appear to have had
Penfold-                                           2004        ity (331) "NO" votes in       Council            "Mision Ribas"           index, population,          significantly different electoral effects
Becerra                                                                  2004                                   considered an            unemployment rate,          across municipalities of different political
(2008)                                                                   referendum,                            excludable good          fraction of dwellings       leanings." (21) "the yields for electoral
                                                                         change in                              measuring                without potable water,      returns for excludable goods under
                                                                         Chavez votes                           distribution of          fraction of dwellings       Chavez social programs suggest that [the
                                                                         between 2000                           scholarships in the      classified to be in "poor   relationship between the yields for
                                                                         and 2004                               form of cash             condition", fraction of     excludable goods and political
                                                                         elections                              transfers to             the adult population        competition] can take a "U" form. The
                                                                                                                individuals at the       with a high school          electoral return for "Mision Ribas" was
                                                                                                                local level, total       degree or more, average     higher than "Mercal" in hyper-competitive
                                                                                                                expenditure on           number of persons per       arenas, that is, in districts in which both
                                                                                                                "Mision Mercal"          dwelling, fraction of the   the mayor and the governor were
                                                                                                                considered public        population aged 14 or       controlled by opposition forces.... under
                                                                                                                good because any         younger, fraction of the    hyper-competitive conditions loyal voters
                                                                                                                citizen had access to    labor force employed in     are more easily identifiable because
                                                                                                                these discount stores    communications and          probably some sort of self-selection
                                                                                                                to purchase              fraction of the labor       process applies." (25) "Regarding the
                                                                                                                subsidized goods,        force employed in           electoral returns of non-excludable
                                                                                                                total expenditure on     agriculture                 goods,...statistical evidence from Chavez's
                                                                                                                "Mision Identidad"                                   "Misiones" programs does not suggest that
                                                                                                                to study how process                                 these types of goods have higher electoral
                                                                                                                of enfranchisement                                   yields under more competitive political
                                                                                                                (obtaining an ID                                     conditions." (25-26)
                                                                                                                card) interacted with
                                                                                                                the distribution of
                                                                                                                these public and
                                                                                                                excludable goods ,
                                                                                                                dummy for whether
                                                                                                 COLLECTION
                                  CLASSIFICATION                                                                           POLITICAL
                                                    YEAR(S)                   DEPENDENT        INSTRUMENT FOR                                  OTHER INDEPENDENT
   AUTHOR     YEAR     COUNTRY    (FROM REVIEW                 UNITS (N)                                                 INDEPENDENT                                       FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                   ANALYZED                    VARIABLE           DEPENDENT                                        VARIABLES
                                     ESSAY)                                                                                VARIABLES
                                                                                                   VARIABLE

Palaniswamy 2008     India       Redistributive    2005-2006 gram          income and         gram panchayat,        proportion of         district dummies,             SGRY expenditure at GP level is not
and Krishnan                     consequences                panchayat     expenditure from   survey by IF           reserved seats for    population                    significantly influenced by composition of
(2008)                                                       in            gram panchayat's   Panchayati Raj         various social                                      GP or by elite bargaining power variables,
                                                             Karnatak      Sampoorna          Institutions and       groups, minorities,                                 but they are large in magnitude. GP
                                                             a (79)        Grameen Rozgar     Institute for Social   women, total number                                 population and district dummies, which
                                                                           Yojana (SGRY),     and Economic           of seats in GP, elite                               are used in the formulas, are most
                                                                           a program for      Change                 capture measured by                                 significant in determining SGRY income
                                                                           wage                                      land ownership and                                  (27-28).
                                                                           employment and                            political competition
                                                                           infrastructure                            of councilors



Park         2002    South Korea Swing versus      1990-1999 electoral     intergovernmenta Financial                assembly member's       fiscal independence         Before decentralization (1991-95), none
(2002)                           core                        district      l transfers      Yearbook of Local        alignment with the      index (share of local tax   of the political variables are significant.
                                                                                            Governments by           incumbent               and non-tax revenue to      After decentralization, only degree of
                                                                                            Ministry of              administration, each    total revenue), per         political competition becomes significant.
                                                                                            Government               member's seniority,     capita inhabitant tax,      (16) "before decentralization, highly
                                                                                            Administration and       political               ratios of population        competitive districts (swing) received
                                                                                            Home Affairs             competitiveness         receiving running           favorable distribution, and highly
                                                                                                                                             water, ratios of people     competitive districts with their assembly
                                                                                                                                             engaged in agriculture,     members from the incumbent party
                                                                                                                                             population density          received negative treatment in the
                                                                                                                                                                         distribution of National Subsidy." (18)


Peisakhin    2009    India       Redistributive          2008 individual   right/register to field experiment        field experiment                                    Recourse to the RTI substantially speeds
(2009)                           consequences                 (61 from     vote (placement conducted by              with 3 groups:                                      up the voter registration process by
                                                              slums, 61    on electoral roll) authors in 2008        information                                         comparison to standard application
                                                              middle-      measured by                               treatment (Right to                                 procedure. Urban poor are 24 times more
                                                              class)       length of time                            Information Act),                                   likely to be placed on the roll than the
                                                                           before an                                 bribe treatment,                                    control, and middle class 14 more times.
                                                                           applicant is                              control                                             Threat of information disclosure forces
                                                                           added to the draft                                                                            public officials to operate more efficiently,
                                                                           district electoral                                                                            and can be a legal substitute to bribery
                                                                           roll                                                                                          (23)
Pereira      1996    Portugal    Redistributive          1989 communi intergovernmenta Direccao-Geral da             "the proportion of      structure of local          "The 'political' variable P under this
(1996)                           consequences                 ties (186) l lump-sum         Administracao            votes in local          government lobby            alternative political hypothesis is not
                                                                         grants from a      Autarquica               elections for the       organization, per capita    statistically significant even at a 90%
                                                                         fund for financial (DGAA)                   political party which   tax base, economies of      degree of confidence, and there are at
                                                                         imbalance                                   form the                sharing the                 least two reasons to explain this result."
                                                                         "which                                      government" (196)       consumption of capital      One is that rewarding core supporters and
                                                                         consolidates                                                        intensive public goods      buying off opposing voters might cancel
                                                                         almost all                                                          (economies of scale)        each other out. The other is that parties
                                                                         transfers from                                                                                  care only about the total number of votes,
                                                                         central to local                                                                                not about the vote share
                                                                         governments"
                                                                         (194)


Persico,      2007   Mexico      Political         1944-2001 municipal annual change in National Water  state governors'                     time-varying                On average across municipalities,
Rodríguez-                       business cycles             ities     number of        Commission (CNA election                             demographics                economically and statistically significant
Pueblita, and                                                (463)     potable water    reports)                                                                         cycle in spending associated with state
Silverman                                                              (inlet) and                                                                                       governor elections -- the years in which a
(2007)                                                                 sewage (outlet)                                                                                             s
                                                                                                                                                                         governor’ election is held, the change in
                                                                       connections to                                                                                    water inlets per capita is, on ,average
                                                                       public networks                                                                                   0.0072 higher. Political budget cycle in
                                                                       (per capita)                                                                                      this form of spending appears only in
                                                                       (also has formal                                                                                  those states where the incumbent
                                                                       model on                                                                                          governor is from the PRI. In PRI-
                                                                       factions)                                                                                         governed state the election is associated
                                                                                                                                                                         with an additional 0.0077 inlets per
                                                                                                                                                                         capita. (39)
                                                                                                   COLLECTION
                                   CLASSIFICATION                                                                          POLITICAL
                                                     YEAR(S)                   DEPENDENT         INSTRUMENT FOR                                 OTHER INDEPENDENT
   AUTHOR      YEAR     COUNTRY    (FROM REVIEW                 UNITS (N)                                                INDEPENDENT                                       FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                    ANALYZED                    VARIABLE            DEPENDENT                                       VARIABLES
                                      ESSAY)                                                                               VARIABLES
                                                                                                     VARIABLE

Porto and     2001    Argentina   Redistributive    1960-1999 provinces transfers per           Porto (1990),        deputies per capita,     population share,          "The results of the estimation strongly
Sanguinetti                       consequences                (88=22    capita                  population share     senators per capita      income per capita and      support the hypothesis that the political
(2001)                                                        provinces                         from national                                 density                    representation at the National Congress
                                                              *4                                census (INDEC                                                            has been a key variable that has
                                                              decades)                          1993)                                                                    determined the distribution of transfers
                                                                                                                                                                         across provinces in Argentina in the last
                                                                                                                                                                         decades. The estimated coefficients for
                                                                                                                                                                         the variables deputies and senators, both
                                                                                                                                                                         expressed in per capita terms, are positive
                                                                                                                                                                         and significant in all the regressions."
                                                                                                                                                                         (252)
Rao and       2001    India       Political returns 1983-1993 major         statutory           National Institute   share of a ruling        state domestic product,    "some evidence for the importance of
Singh                                                         states        transfers (shared   for Public Finance   party or ruling          per capita state           variables that may proxy bargaining
(2001)                                                        (14)          taxes plus non-     and Policy, New      coalition's Members      domestic product,          power" of Indian states, such as the
                                                                            plan grants),       Delhi                of Parliament who        population                 positive estimate of "lagged effect of a
                                                                            grants for state                         come from a                                         match between the state and central ruling
                                                                            plan schemes,                            particular state,                                   parties on grants for state plan schemes"
                                                                            discretionary                            dummy if party                                      and in the loglinear specification, "the
                                                                            transfers (grants                        controlling the center                              positive effect of the proportion of ruling
                                                                            for central plan                         is same as the party                                party/coalition MPs on per capita statutory
                                                                            schemes and                              controlling the state                               transfers, again with a lag"
                                                                            centrally                                (both lagged three
                                                                            sponsored                                years)
                                                                            schemes), total
                                                                            transfers (sum)



Remmer        2007    Argentina   Political         1983-2003 provinces patronage         Ministerio de              vote share of            unemployment, capital      "Party dominance, on the other hand,
(2007)                            favoritism                            spending: (1)     Economia y                 winning                  spending, tax revenue,     makes a difference, with two of the three
                                                                        total personnel   Produccion                 gubernatorial            discretionary federal      coefficients suggesting that the narrower
                                                                        spending,                                    candidate                transfers, nontax          the electoral base of the incumbent, the
                                                                        personnel                                                             provincial revenues,       greater the incentives to spend on
                                                                        spending per                                                          dummy if Peronist          patronage rather than public goods.
                                                                        capita, personnel                                                     president, dummy if        Replacing the indicator of party
                                                                        spending as a                                                         Peronist provincial        competition with a measure of the
                                                                        percentage of                                                         governor, dummy for        percentage difference between the vote for
                                                                        total provincial                                                      gubernatorial election     winning parties and their closest
                                                                        spending net of                                                       year, one-year lag, one-   competitors yields similar but weaker
                                                                        interest payments                                                     year lead                  results" (369-370)
                                                                        on the provincial
                                                                        debt


Rodden        2002    European    Redistributive    1977-1999 countries real per capita                              votes per capita         agricultural share of      Overrepresented states are rather
(2002)                Union       consequences                (14)      net transfers paid                                                    workforce, GDP per         dramatically favored in the distribution of
                                                                        or received, real                                                     capita, population,        both agricultural and structural funds.
                                                                        per capita
                                                                        transfers
                                                                        received by each
                                                                        member state
                                                                                                 COLLECTION
                                  CLASSIFICATION                                                                      POLITICAL
                                                    YEAR(S)                   DEPENDENT        INSTRUMENT FOR                                OTHER INDEPENDENT
  AUTHOR      YEAR     COUNTRY    (FROM REVIEW                 UNITS (N)                                            INDEPENDENT                                        FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                   ANALYZED                    VARIABLE           DEPENDENT                                      VARIABLES
                                     ESSAY)                                                                           VARIABLES
                                                                                                   VARIABLE

Rodden and   2004    Brazil      Swing versus      1991-2000 states        constitutional     Ministry of       total number of votes      real state domestic      "The margin separating the top two vote
Arretche                         core                        (26)          transfers, total   Finance           received by the            product per capita, staterecipients in the state in the last
(2004)                                                                     (current +                           president in the most      population, dummy for    presidential election has a negative effect
                                                                           capital)                             recent election,           presidential election    on discretionary transfers... states that
                                                                           discretionary                        distance between           years                    provided more votes for the president in
                                                                           transfers, total                     vote shares of top                                  the most recent election received larger
                                                                           transfers                            two presidential                                    per capita transfers" (22) "Inserting an
                                                                           (constitutional                      candidates in most                                  interaction term (margin x votes for
                                                                           transfers +                          recent election,                                    president) in model 4 helps solve the
                                                                           current                              effective turnout                                   puzzle. The positive conditional
                                                                           discretionary +                      proxied by unspoiled                                coefficient for president’s vote is
                                                                           capital                              ballots cast as a                                   significant and increases slowly with the
                                                                           discretionary)                       share of voting age                                 margin of victory. However, the
                                                                                                                population, average                                 coefficient for margin is only negative for
                                                                                                                of number of seats                                  states where the president failed to garner
                                                                                                                per million                                         more than one million votes, and even in
                                                                                                                inhabitants for                                     those it is not statistically significant over
                                                                                                                Senate and for                                      much of the sample range. When the
                                                                                                                Chamber of                                          president receives over 1 million votes,
                                                                                                                Deputies, number of                                 the conditional coefficient for margin
                                                                                                                the state's deputies                                actually becomes positive, and it is
                                                                                                                that belong to                                      significant when the number surpasses 2
                                                                                                                president's legislative                             million. Thus for the large states, the
                                                                                                                coalition as share of                               effect of margin is positive, and the
                                                                                                                total legislative seats,                            president won these states in 90 percent of
                                                                                                                dummy if governor                                   the cases (the large margin of victory was
                                                                                                                belongs to                                          his own). The results are consistent only
                                                                                                                president's legislative                             with a majoritarian or “winner takes the
                                                                                                                coalition, dummy if                                 spoils” perspective. Cardoso appears to
Rodden and   2004    India       Swing versus      1957-2003, states       sum of transfers   Reserve Bank of   for eras of Congress       logged real state        for periods of Congress dominance:
Wilkinson                        core              1972-2003 (28)          for central plan   India Bulletin    dominance: Congres         domestic product per     "Discretionary resources are apparently
(2004)                                                                     schemes and                          s's share of state's       capita, percent of       targeted at states where Congress has the
                                                                           centrally                            parliament                 population living in     strongest support, but also at states that
                                                                           sponsored                            delegation (proxy for      urban areas, dummy if are relatively evenly divided between
                                                                           schemes, real                        core support),             under "president's rule" Congress and opposition parties…the only
                                                                           loans, total                         absolute difference        (emergency central       clear losers are states that are dominated
                                                                           grants                               between Congress's         rule), dummy if state is by opposition parties...we also find that
                                                                                                                share of state's           a "special category"     states controlled by Congress Chief
                                                                                                                parliament                 state (special           Ministers receive more discretionary
                                                                                                                delegation and 50%         reservations for         grants than states controlled by the
                                                                                                                (proxy for swing),         minorities)              opposition" (21-22). "Swing states are
                                                                                                                dummy if state chief                                favored regardless of the partisanship of
                                                                                                                minister is same                                    the state chief minister" contra DDD and
                                                                                                                party as central                                    Khemani (22). for coalition
                                                                                                                government. for                                     governments: pooled results: no
                                                                                                                coalition                                           significant political variables (24).
                                                                                                                governments: share                                  Dropping earlier coalitions and looking
                                                                                                                of state parliament                                 post-1995: swing proxy not significant,
                                                                                                                that is chief                                       states with chief ministers of the same
                                                                                                                minister's party,                                   party as prime minister favored, states
                                                                                                                share of state                                      with governments sharing at least one
                                                                                                                parliament that is                                  party with national coalition are
                                                                                                                junior coalition                                    disfavored (authors are "flummoxed" by
                                                                                                                partners, share of                                  this), the states with high numbers of
                                                                                                                state parliament that                               junior partners and outside supporters of
                                                                                                                is outside-coalition                                coalition are favored (25-26)
                                                                                                                support parties,
                                                                                                                absolute difference
                                                                                                 COLLECTION
                                 CLASSIFICATION                                                                          POLITICAL
                                                     YEAR(S)                   DEPENDENT       INSTRUMENT FOR                                 OTHER INDEPENDENT
  AUTHOR     YEAR     COUNTRY    (FROM REVIEW                   UNITS (N)                                              INDEPENDENT                                     FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                    ANALYZED                    VARIABLE          DEPENDENT                                       VARIABLES
                                    ESSAY)                                                                               VARIABLES
                                                                                                   VARIABLE

Rozevitch   1993    Israel      Political           1966-1983, municipal Government           "Local Authorities   dummy if mayor is population, number of           "The political dummy variable gave
and Weiss                       favoritism          1971-1983 ities (26) participation in     in Israel -          member of ruling   cars, area of city, size of    interesting results. As hypothesized,
(1993)                                                                   municipal            Financial Data"'     party in Knesset   cumulative deficit,            politics plays a larger role in the general
                                                                         spending             and "Local           (national level),                                 grant than in the earmarked grant. In fact,
                                                                         (grants),divided     Authorities in       dummy if mayor and                                this variable is always significant for the
                                                                         between              Israel - Physical    prime minister are                                general grant and never significant (and in
                                                                         "general" and        Data"                Maarach (Labor),                                  three cases even negative) for the
                                                                         "earmarked"                               dummy f mayor and                                 earmarked grant...politicians do show
                                                                                                                   prime minister are                                favoritism, however, a glance at the size
                                                                                                                   both Likud                                        of the coefficients shows that the benefits
                                                                                                                                                                     are not large" (340) "the Likud seems to
                                                                                                                                                                     favor its members in most regressions
                                                                                                                                                                     while the Maarach shows far less
                                                                                                                                                                     tendency to do so: (342)


Samuels     2002    Brazil      Political returns         1994 incumben whether who ran                            pork (% of all pork-                              Money affects vote totals, but pork
(2002)                                                         ts (244) for reelection                             barrel amendments                                 barreling has no direct effect on votes.
                                                                        won or not in                              released in his or her                            (857) "Deputies thus do not deliver pork
                                                                        1994                                       district), dominance                              because it provides a direct electoral
                                                                                                                   (% of all candidate                               payoff, they do so in order to gain the
                                                                                                                   votes that deputy                                 financial support of powerful economic
                                                                                                                   received where he                                 interests. The money helps win votes."
                                                                                                                   received the largest                              (860)
                                                                                                                   share of his own
                                                                                                                   votes in previous
                                                                                                                   election),
                                                                                                                   pork*dominance,
                                                                                                                   campaign spending
                                                                                                                   (also squared), rank
                                                                                                                   in previous election,
                                                                                                                   # of terms, leader,
                                                                                                                   switched parties,
                                                                                                                   change in % of votes
                                                                                                                   that deputy's list won
                                                                                                                   in his state from one
                                                                                                                   election to next,
                                                                                                                   leader, vote share of
                                                                                                                   1990, turnout, party
                                                                                                                   dummies




Sapienza    2004    Italy       Redistributive      1991-1995 firms         interest rate     Centrale dei         local political          size of bank, percentage "The political strength of the party has a
(2004)                          consequences                                charged to firm I Bilanci              strength -- measured     of non-performing        negative and significant effect on the
                                                                            by bank k at time                      by the percentage of     loans, market            interest rate charged to borrowers"
                                                                            t minus the prime                      votes received by the    concentration, size of
                                                                            rate at time t                         party to which the       firm
                                                                                                                   chairperson of the
                                                                                                                   state-owned bank is
                                                                                                                   affiliated in the area
                                                                                                                   where the firm is
                                                                                                                   borrowing
                                                                                                    COLLECTION
                                   CLASSIFICATION                                                                            POLITICAL
                                                     YEAR(S)                  DEPENDENT           INSTRUMENT FOR                                 OTHER INDEPENDENT
   AUTHOR     YEAR     COUNTRY     (FROM REVIEW                UNITS (N)                                                   INDEPENDENT                                        FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                    ANALYZED                   VARIABLE              DEPENDENT                                       VARIABLES
                                      ESSAY)                                                                                 VARIABLES
                                                                                                      VARIABLE

Schady       2000    Peru         Political         1991-93,   provinces   spending by the                             Vote share for          logged FONCODES              "The coefficient on nonmarginality is
(2000)                            favoritism        1993-95    (194)       FONCODES for                                Fujimori, absolute      poverty index, logged        always negative and is significant at the
                                                               plus        community-                                  difference of vote      population, percentage       5% level when all the controls are added;
                                                               districts   based projects                              share for Fujimori      rural. Also, a Fourier       the coefficient on the measure of core
                                                               of Lima                                                 minus 50%, change       series expansion of the      support is positive and significant at the
                                                                                                                       in Fujimori vote        FONCODES poverty             5% level in every specification" (299)
                                                                                                                       between 1990 and        index, average age of        When all controls are added, in 1991-
                                                                                                                       1993                    household head,              1993, the highest level of spending went
                                                                                                                                               average number of            to provinces where 50-60% of the
                                                                                                                                               years of schooling,          population voted for Fujimori (see Fig 4).
                                                                                                                                               occupation, proportion       In 1993-95, though, "the coefficients on
                                                                                                                                               owning a car, a              the measures of marginal vote and core
                                                                                                                                               refrigerator, color TV,      support are jointly insignificant at the 5%
                                                                                                                                               and black-and-white          level in all the specifications" (301).
                                                                                                                                               TV, proportion that is       However the coefficient on change in
                                                                                                                                               indigenous, infant           Fujimori vote is negative and significant,
                                                                                                                                               mortality rate, dummy if     meaning that provinces that turned against
                                                                                                                                               border province,             Fujimori got more FONCODES spending.
                                                                                                                                               department dummies


Shi and              85 countries Political         1975-1995 country-     govt. fiscal          Database on           election dummy          GDP per capita, GDP          Fiscal deficit increases by 0.9 percent of
Svensson                          business cycles             year         budget balance        Political                                     growth rate                  GDP in election years. Political budget
(2006)                                                        (1204)       as a share of         Institutions from                                                          cycles are large at 1.3 pp in developing
                                                                           GDP                   the World Bank                                                             countries but almost nonexistent in
                                                                                                                                                                            developed countries. (1372-1373)

Solé-Ollé and 2006   Spain        Political         1993-2003 local        per capita            capital grants from   dummy if                population, land area,       "partisan alignment has a sizeable effect
Sorribas-                         favoritism        (last two governme     Central,              survey on budget      municipality is         property value per           on the amount of grants received by
Navarro                                             years of  nt (869)     Regional, or          outlays conducted     aligned with Central,   capita, property tax rate,   municipalities. The effect is much
(2006)                                              each term              Upper-local           yearly by Ministry    Regional, or Upper-     debt charges/revenue         stronger when the aligned governments
                                                    1994-95,               grants to             of Economics and      local grantor for the                                are single-party governments at both
                                                    1998-99,               respective            Finance               respective term of                                   layers." "in the single party case, aligned
                                                    and 2002-              municipality                                office; cut point                                    municipalities receive over 40% more
                                                    03)                    (located in a                               density, or                                          grants than unaligned ones." (28)
                                                                           particular                                  proportion of swing                                  "There is also a significant effect when the
                                                                           electoral district,                         voters, measured by                                  party at one or both layers is the leader of
                                                                           for a particular                            difference in                                        a coalition, and when a single party or a
                                                                           municipal term                              absolute value                                       party leading a coalition at the municipal
                                                                           of office)                                  between the vote                                     level is also partner of a coalition at the
                                                                                                                       share of the party in                                upper level. However, parties which are
                                                                                                                       government and the                                   mere partners at the municipal level do
                                                                                                                       vote share of the                                    not seem to get more grants from upper
                                                                                                                       next party, having                                   tiers of governments controlled by the
                                                                                                                       either more or fewer                                 same party. " (28)
                                                                                                                       votes


Stasavage            44 African   Redistributive    1980-1996 country-     public spending UNESCO, World               multiparty            GDP per capita (log),          Multipary political competition results in
(2005)               countries    consequences                year         on                Development               competition, election aid (% GDP), rural             1.1% of GDP increase of total education
                                                              (365)        eduction/primary Indicators                 year                  population (%),                and 4.4% of total expenditure increase to
                                                                           education/univer                                                  population under 15            education. Multiparty competition would
                                                                           sity a(% GDP                                                      (%)                            increase primary education spending by
                                                                           and % total govt.                                                                                0.45% of GDP and 1.5% in share of total
                                                                           spending)                                                                                        govt. spending (351). As hypothesized,
                                                                                                                                                                            spending on unversities remains
                                                                                                                                                                            unchanged with electoral competition.
                                                                                                COLLECTION
                                  CLASSIFICATION                                                                      POLITICAL
                                                     YEAR(S)                DEPENDENT         INSTRUMENT FOR                            OTHER INDEPENDENT
  AUTHOR      YEAR    COUNTRY     (FROM REVIEW                  UNITS (N)                                           INDEPENDENT                                    FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                    ANALYZED                 VARIABLE            DEPENDENT                                  VARIABLES
                                     ESSAY)                                                                           VARIABLES
                                                                                                  VARIABLE

Stokes (2005) 2005   Argentina   Swing versus            2001 individual patron (received Survey conducted      Peronist             income, education,      Based on income, education, housing
                                 core                         s (1920) politician's help), in 2001-02 in 3      sympathizer, radical housing quality,        quality evidence, results show that
                                                                         job (would turn provinces              sympathizer, ballot population (log), gender political machine buys votes of poor
                                                                         to party operative                     given by party                               people. In Argentina, the more a party is
                                                                         for job), reward                       operative                                    able to monitor its voters, the more
                                                                         (received goods                                                                     effective its vote buying efforts will be.
                                                                         in last campaign)                                                                   Parties with the most decentralized
                                                                         (formal model on                                                                    organizational structure are best able to
                                                                         machine politics)                                                                   monitor. Community structure voter is in
                                                                                                                                                             also matters because those in smaller
                                                                                                                                                             communities are easier to vote and
                                                                                                                                                             receiving the ballot directly from a party
                                                                                                                                                             operative is more likely to cause effective
                                                                                                                                                             targeting.

Stratmann    2002    Germany     Programmatic      1990-94,  legislator which committee      Amliches           FPTP legislator       party-list rank, percent   "The coefficients on the FPTP legislator
and Baur                         versus pork       1994-98,  s          legislators are      Handbuch des       dummy                 in election, district      variable in all district committee
(2002)                                             1998-2002            on: those that are   Deutschen                                service employment,        regressions are statistically significant,
                                                                        District             Bundestages,                             district industry          indicating that FPTP legislators are
                                                                        Committees           bundestag.de,                            employment, district       members on committees that help their
                                                                        (Construction,       Bevolkerung und                          farm employment,           district constituency" (511)
                                                                        Traffic,             Erwerbstatigkeit                         district homogeneity,
                                                                        Agriculture),                                                 female, seniority, age
                                                                        Defense and
                                                                        Development,
                                                                        and (Women and
                                                                        Youth, Family
                                                                        and Elderly,
                                                                        Health) Some
                                                                        are neutral
                                                                                                     COLLECTION
                                  CLASSIFICATION                                                                               POLITICAL
                                                      YEAR(S)                    DEPENDENT         INSTRUMENT FOR                                   OTHER INDEPENDENT
   AUTHOR     YEAR     COUNTRY    (FROM REVIEW                    UNITS (N)                                                  INDEPENDENT                                       FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                     ANALYZED                     VARIABLE            DEPENDENT                                         VARIABLES
                                     ESSAY)                                                                                    VARIABLES
                                                                                                       VARIABLE

Su and Yang 2000     China       Non-democracy 1978-1994 province/ provincial shares Database                            Central Committee        provincial income per      Central Committee Index positive and
(2000)                                                   year      of state sector                                       Index (CCI) which        capita, population size.   significant at 0.01 level, implying a
                                                                   capital                                               shows provincial         Geographic location        province that enjoys greater representation
                                                                   construction                                          representation           indicator (central,        on Party Central Committee is likely to
                                                                   investment                                                                     coastal)                   have more resources allocated to that
                                                                                                                                                                             province from the central government.
                                                                                                                                                                             (223) As far as central cities, Beijing,
                                                                                                                                                                             Shanghai, Tianjin, greater representation
                                                                                                                                                                             on the Central Committee is associated
                                                                                                                                                                             with less state sector capital construction
                                                                                                                                                                             investment, and estimate is insignificant at
                                                                                                                                                                             0.01 level. Thus supports autonomous
                                                                                                                                                                             center model. (225)


Szwarcberg   2008    Argentina   Political returns 1995-2005 municipal candidate’s        political parties’          monitor clients,                                       "Candidates who employ viewers to
(2008)                                                       ities (8) position on the closed-ballots                 gender, and the                                        monitor client turnout are likely to run in a
                                                                       party ballot,                                  probability of getting                                 position 7 times higher than those that do
                                                                       broker's electoral                             elected. Also                                          not employ viewers. Also, candidates that
                                                                       performance                                    qualitative results on                                 combine viewers’ monitoring with their
                                                                                                                      programmatic                                           own surveillance practices are prone to
Szwarcberg   2010    Argentina   Political returns         2005 local         indicator for       semi-closed and in- unpaid party activist municipality                     "The negative and significant coefficients
(2010)                                                          elected       whether             depth interviews, (bottom-up),             population, gender,             on unpaid party activist, outsiders, and
                                                                represent     candidates took     participant         outsider(top-down), Buesnos Aires                      minority party show that these
                                                                atives        attendance in       observations at     Union Civica                                           independent variables predict monitoring.
                                                                (144)         rallies             political meetings, Radical, minority                                      Likewise, the positive and significant
                                                                                                  rallies and on      parties, National and                                  coefficient on size shows that candidates
                                                                                                  Election Day by     provincial incumbent                                   competing in bigger municipalities are
                                                                                                  author                                                                     more prone to monitor voters than
                                                                                                                                                                             candidates competing in smaller districts.
                                                                                                                                                                             " (20)
Tandon       2007    India       Political returns 1991-2004 regions          vote share of       Election               vote change arising employment,                     Robust relationship between the regional
(2007)                                                       (472)            parliamentary       Commission             from reform, wage population                        exposure to tariff decreases and Left
                                                                              members (Left,                             change from reform,                                 gains. Results driven by BJP and far left
                                                                              Con, Far Left,                             regional exposure to                                rather than Congress and unaffiliated
                                                                              BJP, Reg)                                  tariff, industry                                    regional parties. "The results verify that
                                                                                                                         opened up to FDI                                    voters hurt by reform changed their
                                                                                                                         without a license.,                                 preferences based on party lines and
                                                                                                                         industry delicensed,                                increased their support for parties more
                                                                                                                         share of Assembly                                   likely to implement complementary
                                                                                                                         Constituencies held                                 policies to alleviate the distress caused by
                                                                                                                         by the party of the                                 reform." (11)
                                                                                                                         state government


Tavits       2009    Denmark,    Programmatic        2000,        municipal   amount of           Statistical offices    share of seats held in   per person tax income,     "the incumbent government seat share in a
(2009)               Finland,    versus pork         2001, 2002   ity         national            of each country,       municipal council by     unemployment rate,         municipal council is a significant
                     Norway,                         depending    (ranging    government          Ministry of Interior   those parties that       population size (log),     predictor of the change in the amount of
                     Sweden                          on country   from 273-   grants to           and Health             form the cabinet at      district magnitude per     central government grant allocated to that
                                                                  432)        municipality, per   (Denmark,              the central level (for   1000 population            municipality: the higher the seat share, the
                                                                              person, and         )Ministry of Justice   coalition                                           higher the increase in the amount of
                                                                              changes in the      (Finland), Gateway     government, total                                   grants." (113) "there is some evidence
                                                                              per-person sum 1    to Local Finland,      seat share for all                                  that pork-barreling is indeed electorally
                                                                              year before         The Commune            coalition partners),                                beneficial. Voters in Sweden and Finland
                                                                              national election   Database,              voter turnout                                       seem to be persuaded by partisan pork-
                                                                                                  Norwegian Social                                                           barreling and willing to support the
                                                                                                  Science Data                                                               incumbents for such particularized
                                                                                                  Services (Norway),                                                         benefits; however, the results for the latter
                                                                                                  Swedish Election                                                           are not very robust... In the case of
                                                                                                  Authority                                                                  Norway I was not able to confirm that
                                                                                                  (Sweden)                                                                   grant allocation serves its desired purpose
                                                                                                                                                                             of winning votes for the incumbent
                                                                                                                                                                             parties." (116) "strong evidence that
                                                                                                                                                                             incumbent governments in advanced
                                                                                                                                                                             parliamentary and proportional systems
                                                                                                                                                                             engage in geographically targeted
                                                                                                                                                                             spending in the manner of pork-barrel
                                                                                                                                                                             politics." (118)
                                                                                                     COLLECTION
                                   CLASSIFICATION                                                                            POLITICAL
                                                       YEAR(S)                   DEPENDENT         INSTRUMENT FOR                                OTHER INDEPENDENT
   AUTHOR      YEAR     COUNTRY    (FROM REVIEW                   UNITS (N)                                                INDEPENDENT                                       FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                      ANALYZED                    VARIABLE            DEPENDENT                                      VARIABLES
                                      ESSAY)                                                                                 VARIABLES
                                                                                                       VARIABLE

Teitelbaum    2010    India       Political           1973-2006 states        social services    Indian budget         literacy and party      per capita state            "…breakdown of the traditional patronage
and Thachil                       favoritism                    (16)          vs. economic                             fragmentation, % of     domestic product (log),     system, and the emergence of lower caste
(2010)                                                                        services                                 seats won by 'Other     dummy for election          voters have given rise to new demands for
                                                                              (development                             Backward Classes'       years, population log       spending on social services.... party
                                                                              expenditures)                            (OBC) candidates, %                                 fragmentation is positively associated with
                                                                                                                       of seats by three                                   higher levels of spending on social
                                                                                                                       most dominant                                       services and lower spending on economic
                                                                                                                       ideological centers                                 services that can be more easily targeted
                                                                                                                       of Indian politics                                  at specific groups of voters... emergence
                                                                                                                       (Indian National                                    of lower caste parties has given rise to
                                                                                                                       Congress, Bharatiya                                 greater demand for spending on
                                                                                                                       Janata Pary, and left                               education." (29)
                                                                                                                       (communist) party)



Thachil       2011    India       Political returns         2008 voters in    vote for Hindu    survey conducted membership in party           religiosity,         participation in service activities delivered
(2011)                                                           24           nationalist party by author        auxilliary                    communalism, income, by Hindu nationalist party's auxilliary
                                                                 villages                                        organization,                 caste                organization increases likelihood of
                                                                 in 4                                            participant in party                               support for that party even on the party of
                                                                 districts                                       auxilliary                                         low-caste voters
                                                                 in                                              organization
                                                                 Chhattisg                                       activities,
Thames        2001    Russia      Programmatic        1994-1998 individual    rate at which     Panorama,        mandate (SMD or               control for final faction   "SMD-PR divide in party cohesion did
(2001)                            versus pork                    deputies     individual        Federal'noe      PR), factional                or deputy group             exist across all contested votes and
                                                                              deputies voted    Sobranie:        switching, indicator          membership, party           contested budgetary policy votes between
                                                                              with the          Spravochnik      for PR deputies in            affiliation                 1994 and 1998. Thus, a bifurcation
                                                                              positions of the (Moscow, 000      the top 11 positions                                      between how SMD deputies voted and PR
                                                                              majority of their Panorama, 1994); of their parties' lists,                                  deputies voted existed in the Russian State
                                                                              faction or deputy Federal'noe      type of partisan                                          Duma" (879) Type of policy being
                                                                              group's member Sobranie: Sovet     organization to                                           decided is less important for deputies to
                                                                              (contested votes Federatsii,       which deputies                                            degree of support of their faction or
                                                                              and budget        Gosudarstvennaya belonged, margin,                                         deputy group majorities.
                                                                              policy votes)     Duma,
                                                                                                Spravochnik
                                                                                                (Moscow, 000
                                                                                                Panorama, 1996).

Thames        2000    Russia      Political returns 1994-6       individual   individual         Center of Applied     deputy’s faction                                    Yeltsin has used the budget process to
(2000)                                                           deputies     deputy’s vote on   Political Research,   type, cadre,                                        build a coalition of supporters that
                                                                 (less than   the final budget   INDEM Statistics      corporatist or mass-                                reduced the need for legislation by
                                                                 450)         vote for each      2.0                   based, ideology                                     presidential decree. Deputies from
                                                                              year                                                                                         corporatist factions has largest impact on
                                                                                                                                                                           Yeltsin's budgets.
Toha          2009    Indonesia   Political           1999-2005 districts     local public     Dalam Angkas.           vote margin, ruling GDP per capita, area            Evidence that electoral incentives drives
(2009)                            favoritism                    (440)         goods measured                           party supporter, won size, population               distribution of public goods is weak.
                                                                              by a block grant                         by Golkar, effective                                Highly populated district and the more a
                                                                              (Dana Alokasi                            number of parties,                                  district could generate revenues from its
                                                                              Umum, or DAU                             ruling party                                        natural resources as a percentage of its
                                                                              logged)                                  supporter,                                          GDP it tend to receive less money,
                                                                                                                       interaction terms,                                  whereas district area is correlated tend to
                                                                                                                       separatism measures                                 receive more transfers. In contrast,
                                                                                                                                                                           relationship between electoral competition
                                                                                                                                                                           and size of transfer per capita current
                                                                                                                                                                           separatist dummy variable is statistically
                                                                                                                                                                           significant.


Treisman      1996    Russia      Political                 1992 administrtransfers of           newspaper             support for Yeltsin     underdevelopment of        Transfers go to districts that hold credible
(1996)                            favoritism                     ative    financial              Segodnya ,            in 1991/margin of       social infrastructure,     "bargaining power", i.e. those that
                                                                 region   resources to the       Council of            vote (voting record),   profits per capita, urban- threatened by declaring sovereignty or
                                                                 (89)     regions net of tax     Ministers (using      deputy in powerful      rural mix (regional        striking. Also go to regions that voted
Vaishnav and 2010     India       Swing versus        1977-2007 constitue payments to
                                                                          public school          info from Ministry
                                                                                                 District              committee,
                                                                                                                       ruling coalition,       need), local               againstMembers No Legislative Assembly
                                                                                                                                                                          Ruling Yeltsin. of evidence of money
Sircar                            core                          ncies     construction in        Information           margin of victory,                                  in Tamil Nadu have preference to target
(2010)                                                          (210)/yea Tamil Nadu             System for            interaction between                                 club (public) goods to core voter group.
                                                                r                                Education             the two                                             MLAs immediately reward compliant
                                                                                                 initiative of the                                                         voters post-election with pork in the form
                                                                                                 National                                                                  of new schools. "We find that there are
                                                                                                     COLLECTION
                                    CLASSIFICATION                                                                          POLITICAL
                                                       YEAR(S)                   DEPENDENT         INSTRUMENT FOR                                  OTHER INDEPENDENT
   AUTHOR      YEAR     COUNTRY     (FROM REVIEW                  UNITS (N)                                               INDEPENDENT                                          FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                      ANALYZED                    VARIABLE            DEPENDENT                                        VARIABLES
                                       ESSAY)                                                                               VARIABLES
                                                                                                       VARIABLE

Vicente       2007    Sao Tome     Redistributive          2006 househol      change in vote      survey conducted    anti vote-buying           demographic controls,       "vote buying is effective in changing
(2007)                and Principe consequences                 d (1275       buying              by author           campaign, dummy            location controls,          voting preferences/behavior. In
                                                                in            (perceptions on     (randomized field   for urban location,        dummy variable of 1 in      particular, when compared with the
                                                                preelectio    parliamentary vs.   experiment)         dummy variable for         treated areas,              incumbent, the challenger seems to
                                                                n survey,     presidential                            swing zones                information gathered        benefit more from vote buying. In
                                                                1034 in       elections);                                                        about candidates,           addition, we show that vote buying seems
                                                                post-         change in voting                                                   electoral observation       to energize the electorate." (32)
                                                                election      behavior for                                                       mission of CPLP             "Regarding the prices of votes, ... higher
                                                                survey)       presidential                                                       (minutes spent at each      in urban areas and swing zones." (27)
                                                                              election                                                           location, survey-based      Effect of anti-vote buying campaign on
                                                                              (intended vs.                                                      measures of                 change in voter turnout is not significant
                                                                              actual reported)                                                   psychological               (30).
                                                                                                                                                 consistency of
                                                                                                                                                 respondents

Wantchekon    2003    Benin        Political               2001 district      vote for a type of survey conducted     past voting behavior, gender, age                      "The empirical results show that
(2003)                             favoritism                   (10           candidate          in all 10            types of treatment                                     clientelism works for all types of
                                                                experime                         experimental         (clientelist, public                                   candidates but particularly well for
                                                                ntal                             districts. In each   policy)                                                regional and incumbent candidates." (421)
                                                                districts,                       district a                                                                  "(1) negative public policy treatment
                                                                approx                           representative                                                              effect in the North and district controlled
                                                                180                              sample of voters                                                            by regional candidates, (2) positive public
                                                                samples                          interviewed in the                                                          policy treatment effect in the South, (3)
Ward and      1999    England      Swing versus      FY 1994- larger          authority's                             need variables             population, square of       "government allocated around £500
John                               core              1995       English       standard                                assessed by factor         population, dummy for       million more to local authorities
(1999)                                                          local         spending                                analysis derived           unitary authorities,        containing marginal constituencies and
                                                                authoritie    assessment, a                           from various               dummy for London            around £155 million more to 'flagship'
                                                                s             figure calculated                       indicators such as         boroughs, dummy for         local authorities than they could have
                                                                including     by a known                              age-structure of           shire counties              been expected to get on the criteria of
                                                                London        formula                                 population, dummy                                      social need and population." (32)
Weitz-        2009    Argentina    Political         around     individual    implementation survey conducted         poverty level, non-        population (log), mean      "increasing levels of opposition decrease
Shapiro                            favoritism        2006       s in 126      of Programa        by author            vote related costs,        education, radio            the probability of observing clientelism
(2009)                                                          cities        Nacional de                             size of opposition,        stations                    where poverty is low. As poverty
                                                                              Seguridad                               PJ (Peronist party),                                   increases, increases in the size of
                                                                              Alimentaria                             Governor party,                                        opposition are associated with an
                                                                              (National Food                          reelection, mayor                                      increased probability of mayoral
                                                                              Security                                party dominates,                                       intervention into the PNSA." (23) "high
Wilkinson     2006    India        Redistributive    1960-1992 major          capital road       Khemani (2004),      political competition      ethnic fractionalization,   Political competition has significant effect
(2006)                             consequences                states         expenditure        Reserve Bank of      (state electoral           religious                   on capital spending on roads. "an
                                                               (14)                              India and State      volatility), state party   fractionalization, total    increase of one standard deviation from
                                                                                                 data                 fractionalization          state population,           the mean level of electoral volatility in a
                                                                                                                      adjusting for              literacy, police density,   state would increase the level of capital
                                                                                                                      independent,               deaths in communal          road expenditure from $7.349 million to
Winters       2009    Ecuador      Swing versus      2006-2007 canton         change from         2002                vote share of Correa       percentage of               "a percentage point increase in the vote
(2009)                             core                        (209)          2006 to 2007 in     Identificacion y    in first and second        individuals living in       for Correa during the first round
                                                                              the number of       Eleccion de         round                      poverty in each canton,     corresponds to a greater increase in BDH
                                                                              Bono de             Beneficiarios                                  percentage of               than would the same percentage point
                                                                              Desarrollo          (SELBEN) survey                                population that is of       increase in the poverty rate." (7) "I have
                                                                              Humano (BDH)        conducted by                                   indigenous origin, total    shown a repeated association between the
                                                                              recipients          government                                     number of recipients in     vote share that Correa obtained in a
                                                                                                                                                 2006, population            canton during the elections and the
Worthington   1998    Australia    Political         1981-82 to states (6) (1) real per           Commonwealth        number of seats a          Education expenditure       "The most crucial elements of this process
and Dollery                        favoritism        1991-92               capita education       grants commission   state has in               disability ratio, heath     of purchasing political capital would
(1998)                                                                     special purpose        report on general   parliament, the            expenditure disability      appear to be a state's number of federal
                                                                           payments SPPs          revenue grant       proportion of seats        ratio, social security      seats, the share of those seats held by the
                                                                           (2) real per           relativities        held by the                and welfare expenditure     federal government, and the incidence of
                                                                           capita health                              Commonwealth               disability ratio            federal elections. Similarity of political
                                                                           SPPs, (3) real per                         (national)                                             persuasion [state-center party alignment]
                                                                           capita social                              government party,                                      would also appear to be instrumental, at
                                                                           security and                               dummy if state and                                     least in regards to education payments.
                                                                           welfare SPPs                               national                                               Moreover, the proxies for political safety
                                                                                                                      governments share                                      perform well for education and welfare,
                                                                                                                      same ruling party,                                     though inconsistently for health" (310)
                                                                                             COLLECTION
                                 CLASSIFICATION                                                                POLITICAL
                                                   YEAR(S)                 DEPENDENT       INSTRUMENT FOR                       OTHER INDEPENDENT
   AUTHOR    YEAR     COUNTRY    (FROM REVIEW                 UNITS (N)                                      INDEPENDENT                                FINDINGS ACCORDING TO PAPER (PAGE NO.)
                                                  ANALYZED                  VARIABLE          DEPENDENT                             VARIABLES
                                    ESSAY)                                                                     VARIABLES
                                                                                               VARIABLE

Yarahuán    2007    Mexico      Political         2001-2006 municipal per capita                        ENP Laakso-
                                                                                        Ministry of Social                                            "FISM resources seem not to be
(2007)                          favoritism                  ity (2436) expenditure      Development     Taageperta Index                              influenced by electoral competition at the
                                                                       (fund for                        (party competition),                          municipal level, by the concurrence of
                                                                       municipal social                 indicator for same                            party between municipal and federal or by
                                                                       infrastructure                   party in federal and                          the presence of municipal elections.
                                                                       FISM,                            municipal                                     However the estimation shows a positive
                                                                       conditional cash                 governments for 2                             cyclical state election effect and a positive
                                                                       transfer program                 terms or more,                                effect of state governments on distribution
                                                                       Oportunidades,                   indicator for same                            at the municipal level. Oportunidades
                                                                       Milk subsidy                     party in federal and                          shows positive significant effect on party
                                                                       program                          state govt, municipal                         concurrence in state and federal
                                                                       Liconsa)                         or state election,                            government. Liconsa appears as the
                                                                                                        party affiliation of                          program with the greatest political
                                                                                                        municipal                                     influence in the distribution of its
Blaydes       2008 Egypt        Non-democracy 1986-1996 governora change in the      Central Agency for previous level of                             "Governorates with high levels of support
(2008)                                                  tes (26) percentage          Public             infrastructure                                for the Brotherhood-Wafd alliance (i.e.
                                                                  of residences in a Mobilization       coverage, urban-                              vote share around 50 percent) saw 30
                                                                  governorate that and Statistics       rural status,                                 percent less water infrastructure
                                                                  have piped                            Brotherhood-Wafd                              development than governorates that saw
                                                                  public water,                         voteshare in the                              low levels of support for the Brotherhood-
                                                                  change in the                         1984 election,                                Wafd alliance (i.e. voteshare
                                                                  percentage of                         percent of non-NDP                            around 0 percent)" (105). Similar results
                                                                  residential                           parliamentary                                 for sewerage infrastructure.
                                                                  buildings in a                        representatives
                                                                  governorate that                      between 1984-1995
                                                                  are connected to
                                                                  the public
                                                                  sewerage
                                                                  network
Zucco       2010    Brazil      Political returns 2006  municipal Lula's vote share                     Bolsa Familia cash level of socioeconomic     "Bolsa Familia had a considerable impact
(2010)                                                  ity                                             transfer program      development, level of   on the voting decisions of individuals…
                                                                                    COLLECTION
                       CLASSIFICATION                                                                     POLITICAL
  AUTHOR                                 YEAR(S)                  DEPENDENT        INSTRUMENT                               OTHER INDEPENDENT
             COUNTRY   (FROM REVIEW                UNITS (N)                                            INDEPENDENT                                FINDINGS ACCORDING TO PAPER (PAGE NO.)
   (YEAR)                               ANALYZED                   VARIABLE       FOR DEPENDENT                                 VARIABLES
                           ESSAY)                                                                         VARIABLES
                                                                                     VARIABLE


Ames        Brazil     Programmatic 1987-1988 municipalit voting in               Constituent        spatial distribution   constituency          "deputies receiving pork benefits voted to
(1995a)                versus pork            y (403)     Constituent             Assembly,          of electoral           attributes, career    weaken the legislature and strengthen the
                                                          Assembly                Kinzo (1989)       support                paths of politician   executive, and tended to oppose statism-
                                                          (Congressional                             (dominance and         (local, business,     welfarim, and popular democracy….In
                                                          Power, Support                             clustering),           bureaucratic),        sum, pork buys deputies." (339)
                                                          for Executive,                             seniority and                                Deputies with clustered vote bases
                                                          Statism-                                   electoral                                    supported congressional power and statis-
                                                          Welfarism,                                 insecurity, state                            welfarist issues. Wealth and industry
                                                          Support for                                unity and state                              had antiexecutive, economically statist-
                                                          Popular                                    interest, political                          welfarist, supportive of popular
                                                          Democracy) and                             party, pork barrel                           democracy.
                                                          Collar's
                                                          Emergency
                                                          Decrees

Ariga       Australia, Programmatic Australia    district,     electoral          original dataset                                                "electoral cohesiveness is higher under
(2005)      Ireland,   versus pork  (1969-83), candidate       cohesiveness (a                                                                    the electoral rules without intra-party
            Finland,                Ireland                    measurement of                                                                     competition [Norway, Australia, the
            Japan,                  (1973-89),                 last margin that                                                                   U.K.] than under those with it [Ireland,
            Norway,                 Finland                    depends on                                                                         Japan, Finland]" (27). "Time series of the
            the United              (1995-                     electoral rules                                                                    estimated electoral cohesiveness under
            Kingdom                 2003),                     which may                                                                          electoral systems without intra-party
                                    Japan (1960-               include intra-                                                                     competition (Australia and the U.K.)
                                    90),                       party competition                                                                  fluctuates largely around higher places in
                                    Norway                     and magnitude)                                                                     the figure than the estimated electoral
                                    (1993-                     across different                                                                   cohesiveness under systems with intra-
                                    2001), the                 electoral systems,                                                                 party competition (Ireland, Japan)" (29).
                                    United                     across time, and                                                                   "..after estimating electoral cohesiveness,
                                    Kingdom                    across parties                                                                     it will be used as a key explanatory
                                    (1951-92)                                                                                                     variable for the allocation of budget to
                                                                                                                                                  collective-goods-type and particularistic-
                                                                                                                                                  goods-type spending, which is the
                                                                                                                                                  ultimate goal of this project" (32).
Chhibber  India       Programmatic 1967-1997 major       proportion of      Reserve Bank   dummy for party       sources of income,    "multiparty states have higher levels of
and                   versus pork            states (15) state's budget     of India       system, one year      state's per capita    expenditures on club goods or civil
Nooruddin                                                allocated to civil Bulletin       lag, nature of        income, growth        administration spending than their two-
(2004)                                                   administration (as                ideological           rate, revenue from    party counterparts. Center-right
                                                         a club good)                      competition in the    income taxes,         competition generates more bloated state
                                                                                           assembly,             grants from central   apparatuses when compared to a single-
                                                                                           competitiveness of    government, loans     party dominant system than competition
                                                                                           election, dummy       from central          between two centrist parties or a center
                                                                                           for coalition         government            and leftist party." (18) Two party systems
                                                                                           government, voter                           need to win majority of vote forcing
                                                                                           turnout                                     political parties to build coalitions,
                                                                                                                                       whereas in multi-party systems, parties
                                                                                                                                       have an incentive to focus on their voting
                                                                                                                                       blocs since any loss of support can result
                                                                                                                                       in defeat. (24)


Crisp and   Columbia Programmatic 1986-90      senators    vote            bill            prior success of                            "Prior success of senator and increase
Ingall               versus pork  (pre-reform) (300)       concentration   (Hirschman-     senator,                                    experienced is positively associated with
(2002)                            and 1994-                                Herfindahl      experience with                             vote concentration. Negative sign on
                                  1998 (post-                              Index used to   the new rules,                              interaction term of experience and
                                  reform)                                  calculate)      previous victory in                         traditional party interaction indicates that
                                                                                           a subnational                               while patterns of support for candidates
                                                                                           district,                                   from traditional parties are becoming
                                                                                           membership in                               more concentrated over time, they are
                                                                                           traditional party,                          doing so more slowly than those for
                                                                                           experience*traditi                          candidates from new and traditionally
                                                                                           onal party                                  small parties." (738)
Edwards    multi-    Programmatic 1970-2000 democracie government             World Bank's     district magnitude     lagged GDP,           "Increases in district magnitude in
and Thames country   versus pork            s (77)     spending               World            (log), indicator for   lagged GDP per        candidate-centered systems lead to
(2007)                                                 measured by            Development      whether election       capita, lagged GDP    decreases in public goods spending, not
                                                       general                Indicators       occurred in a          growth, lagged        simply total expenditure. In the model,
                                                       government final       (2004), UN       previous time          trade openness as a   the coefficient for the log of district
                                                       consumption            Common           period, dummy          % of GDP, % of        magnitude is positive and statistically
                                                       expenditure and        Database         variable for           the population        significant. This means that there is a
                                                       education                               presidential or        above the age of 65   positive relationship between increasing
                                                       expenditure as a                        parliamentary                                magnitude and education expenditure
                                                       % of GDP                                regime                                       when electoral systems are party-
                                                                                                                                            centered."(343) Negative relationship
                                                                                                                                            between increasing magnitude and
                                                                                                                                            candidate-centered electoral systems.


Gagliarducc Italy    Programmatic 1994-2006 individual   fraction of bills    online archive   Margin of victory,     demographic            "Our empirical results show that
i,                   versus pork            members      targeted to the      of bills for     appointments in        characteristics, self- majoritarian elections are strongly
Nannicini,                                  of Italian   region of election   legislative      parliament and in      declared previous associated with a greater amount of
and                                         House of     over the total       activity         government, party      job                    targeted policies, such as pork-barrel
Naticchioni                                 Representa   number of bills                       affiliation,                                  projects favoring local constituencies...
(2008)                                      tives        presented as main                     political                                     proportional representation, however,
                                            (1699)       sponsor, fraction                     experience, local                             would produce other side effects, such as
                                                         of parliament                         government                                    rent extraction by elected officials who
                                                         votes missed                          experience, vote                              are less accountable to their voters." (3)
                                                         without any                           share, system of                              "Being elected in the majoritarian system
                                                         legitimate reason                     election, electoral                           entails an increase in the share of
                                                         over the total                        district                                      geographically targeted bills of 8.2%,
                                                         number of                                                                           this is more than doubles the share of
                                                         electronic votes                                                                    targeted bills with respect to the
                                                                                                                                             predicted value for 6.4 for proportional
                                                                                                                                             representatives at the threshold." (22)
                                                                                                                                             "...being elected in the majoritarian
                                                                                                                                             system entails a fall in the absenteeism
                                                                                                                                             rate equal to 14.9%, that is, a fall of more
                                                                                                                                             than 30% with respect to the predicted
                                                                                                                                             value of 47.7 for proportional
                                                                                                                                             representatives..." (23)
Hainmuelle Germany   Programmatic 1950s-       district       SMD vote share       Caramani         margin of victory      district-level          For both parties under PR, incumbency
r and Kern           versus pork  2000s (8     (248 up to     of party, PR vote    (2000),          in each district to    covariates              status has positive and significant effect
(2008)                            federal      1990 and       share of party       Statistisches    votes in the next                              on party vote share. Incumbency status
                                  elections)   328 after      (Social              Bundesmat        election, party                                increases PR vote share by 1.7% for SPD
                                               unification)   Democratic                            incumbency,                                    and 1.4 % for CDU on average. Also,
                                                              Party(SPD), and                       legislator                                     incumbency has substantial effect on
                                                              Christian                             incumbency                                     party's SMD vote share, and is slightly
                                                              Democratic                                                                           larger for legislator incumbency.
                                                              Union (CDU))
Healy and   United   Programmatic 1988-2004 counties          presidential vote    General          disaster damage        county-level            "When disaster damage increases by 1%,
Malhotra    States   versus pork            (3141)            share for            Election Data    measure (natural       variables that can      the regression results suggest that relief
(2008)                                                        incumbent party      for the U.S.,    disasters), disaster   change over time        expenditures increase by 0.072%. There
                                                              in county            Congressional    relief expenditure,    (economic               appears to be a significant political
                                                                                   Quarterly's      prevention             conditions), county     component to disaster relief spending
                                                                                   Votings and      spending (also         demographics            too. Counties that more strongly
                                                                                   Elections,       lagged),                                       supported the incumbent party in the
                                                                                   ICPSR            incumbent party's                              previous election received more relief
                                                                                                    previous vote                                  spending than did other counties, an
                                                                                                    share in the county                            effect that would not occur if relief
                                                                                                                                                   spending only reflected response to
                                                                                                                                                   disasters." (17)




Hirano      Japan    Programmatic 1980-2002 municipalit       targeted subsidies   account          minimum distance       % of workforce          Under the MMD/SNTV system subsidies
(2006a)              versus pork            y (approx         measured by per      settlement       to any LDP             engaged in first tier   were lower for municipalities farther
                                            3300*years        capita central-to-   (Chiho Zaisei    member's home          industries              from an LDP incumbent's hometown
                                            )                 municipality         Kenkyu Kai),     office                 (agriculture and        (78). Municipalities far from LDP
                                                              government           various issues                          fisheries), % of        incumbents' home offices received more
                                                              transfers (Kokko                                             population              subsidies after electoral reforms (80).
                                                              shishutsukin)                                                considered
                                                                                                                           dependent (over
                                                                                                                           65, under 15),
                                                                                                                           income
Horiuchi    Japan   Programmatic FY 1991 -   municipalit total transfers       Chiho Zaisei LDP's district seat number of seats        "Our regression estimates suggest that in
and Saito           versus pork  FY 1994,    ies (3300) from central           Chosa Kenkyu share               divided by district    most (but not all) years, the LDP's seat
(2003)                           FY 1995 -               government to         Kai (fn 21)                      population (logged,    share had a negative effect on the subsidy
                                 FY 1998                 municipal                                              is a measure of        allocation…we might interpret the
                                                         government                                             apportionment),        negative estimate as a consequence of the
                                                         (logged)                                               municipality fiscal    DLP's efforts to buy off marginal voters"
                                                                                                                strength index         (677) Yet "after FY1994, the magnitude
                                                                                                                (formula),             of the coefficient dropped and the effect
                                                                                                                population,            became statistically insignificant"
                                                                                                                income, dependent      Authors speculate it could be due to
                                                                                                                population,            SMD, earthquake dummy, or that LDP
                                                                                                                industrial structure   "may have, indeed, lost their power to
                                                                                                                (agricultural share,   allocate public resources to their
                                                                                                                service share),        constituencies after Japan entered into an
                                                                                                                municipality urban,    era of coalition governments after 1993"
                                                                                                                dummy if city          (678)
                                                                                                                designated by
                                                                                                                ordinance, dummy
                                                                                                                for Kobe
                                                                                                                earthquake




Kunicová Russia     Programmatic 1994-2003 individual    dissent of a          electronically   electoral mandate bill and deputy      "We found evidence that particularly in
and                 versus pork            voters in     deputy with his       recorded roll-   (SMD/PL),         characteristics      the second and third Dumas, SMD
Remington                                  Russian       faction leader on     call votes       parties, faction,                      deputies were more likely to break with
(2003)                                     State         distributive issues   produced by      faction switching                      faction discipline than were their list-
                                           Dumas                               Moscow-based                                            based comrades on distributive issues,
                                                                               firm INDEM                                              and that this effect grew stronger after
                                                                                                                                       the first Dumas." Results for voting on
                                                                                                                                       issues that can be characterized as non-
                                                                                                                                       distributive were at variance with
                                                                                                                                       expectations, but confined to the first
                                                                                                                                       Dumas. (29)
Lancaster      multi-    Programmatic          countries   level of pork   interviews and electoral systems                      "inverse relationship exists between
(1986)         country   versus pork                       activity (no    secondary                                             district size and pork activity - the larger
                                                           regression      sources i.e.                                          a nation's ratio of representatives to
                                                           analysis)       Aberbach                                              electoral districts, the less pork barrel
                                                                           (1981),                                               activity."(70) "less pork allocation occurs
                                                                           Loewenberg                                            in countries with multimember districts
                                                                           and Patterson                                         can be logically extended to countries
                                                                           (1979)                                                with a single at-large electoral district."
                                                                                                                                 (71) U.S., Canada, New Zealand, India,
                                                                                                                                 and Britain with single-member district
                                                                                                                                 system, have the greatest likelihood for
                                                                                                                                 pork barrel activity. Countries with
                                                                                                                                 multi member district such as Columbia,
                                                                                                                                 Greece, Norway, and Spain have lesser
                                                                                                                                 propensity. Number of representatives
                                                                                                                                 per district in multi-member districts
                                                                                                                                 affects the accountability linkage and
                                                                                                                                 influences pork barrel politics. Multi-
                                                                                                                                 member systems with fewer members per
                                                                                                                                 district are more likely to have pork
                                                                                                                                 barrel activity than those with larger
Milesi-        OECD and Programmatic 1991-94   countries   total primary   OECD           average district    population share   Empirical results consistent with their
Ferretti,      Latin     versus pork           (40)        spending/GDP    Economic       magnitude (AM),     over 65, GDP per   model where a more proportional system
Perotti, and   American                                    ratio           Outlook, IMF   standardized        capita, OECD       can have an ambiguous impact on total
Rostagno       countries                                                   Government     district margin     dummy              primary government spending, depending
(2002)                                                                     Financial      (SM), average                          on relative strength of its effects on
                                                                           Statistics     deviation from                         transfer and public good spending. For
                                                                                          proportionality                        OECD countries, positive relationship
                                                                                                                                 between LogSM and transfers survives,
                                                                                                                                 but in Latin America there is no bivariate
                                                                                                                                 relation between proportionality and
                                                                                                                                 transfers. (634)
Stratmann   Germany   Programmatic 1990-94,  legislators which committee         Amliches       FPTP legislator            party-list rank,       "The coefficients on the FPTP legislator
and Baur              versus pork  1994-98,              legislators are on:     Handbuch des dummy                        percent in election,   variable in all district committee
(2002)                             1998-2002             those that are          Deutschen                                 district service       regressions are statistically significant,
                                                         District                Bundestages,                              employment,            indicating that FPTP legislators are
                                                         Committees              bundestag.de,                             district industry      members on committees that help their
                                                         (Construction,          Bevolkerung                               employment,            district constituency" (511)
                                                         Traffic,                und                                       district farm
                                                         Agriculture),           Erwerbstatigke                            employment,
                                                         Defense and             it                                        district
                                                         Development, and                                                  homogeneity,
                                                         (Women and                                                        female, seniority,
                                                         Youth, Family                                                     age
                                                         and Elderly,
                                                         Health) Some are
                                                         neutral

Tavits      Denmark, Programmatic 2000, 2001,   municipalit   amount of           Statistical       share of seats held    per person tax         "the incumbent government seat share in
(2009)      Finland, versus pork  2002          y (ranging    national            offices of each   in municipal           income,                a municipal council is a significant
            Norway,               depending     from 273-     government          country,          council by those       unemployment           predictor of the change in the amount of
            Sweden                on country    432)          grants to           Ministry of       parties that form      rate, population       central government grant allocated to that
                                                              municipality, per Interior and        the cabinet at the     size (log), district   municipality: the higher the seat share,
                                                              person, and         Health            central level (for     magnitude per          the higher the increase in the amount of
                                                              changes in the per- (Denmark,         coalition              1000 population        grants." (113) "there is some evidence
                                                              person sum 1 year )Ministry of        government, total                             that pork-barreling is indeed electorally
                                                              before national     Justice           seat share for all                            beneficial. Voters in Sweden and
                                                              election            (Finland),        coalition partners),                          Finland seem to be persuaded by partisan
                                                                                  Gateway to        voter turnout                                 pork-barreling and willing to support the
                                                                                  Local Finland,                                                  incumbents for such particularized
                                                                                  The Commune                                                     benefits; however, the results for the
                                                                                  Database,                                                       latter are not very robust... In the case of
                                                                                  Norwegian                                                       Norway I was not able to confirm that
                                                                                  Social Science                                                  grant allocation serves its desired
                                                                                  Data Services                                                   purpose of winning votes for the
                                                                                  (Norway),                                                       incumbent parties." (116) "strong
                                                                                  Swedish                                                         evidence that incumbent governments in
                                                                                  Election                                                        advanced parliamentary and proportional
                                                                                  Authority                                                       systems engage in geographically
                                                                                  (Sweden)                                                        targeted spending in the manner of pork-
                                                                                                                                                  barrel politics." (118)
Thames    Russia   Programmatic 1994-1998 individual rate at which        Panorama,       mandate (SMD or         control for final       "SMD-PR divide in party cohesion did
(2001)             versus pork            deputies   individual           Federal'noe     PR), factional          faction or deputy       exist across all contested votes and
                                                     deputies voted       Sobranie:       switching,              group membership,       contested budgetary policy votes between
                                                     with the positions   Spravochnik     indicator for PR        party affiliation       1994 and 1998. Thus, a bifurcation
                                                     of the majority of   (Moscow, 000    deputies in the top                             between how SMD deputies voted and
                                                     their faction or     Panorama,       11 positions of                                 PR deputies voted existed in the Russian
                                                     deputy group's       1994);          their parties' lists,                           State Duma" (879) Type of policy being
                                                     member               Federal'noe     type of partisan                                decided is less important for deputies to
                                                     (contested votes     Sobranie:       organization to                                 degree of support of their faction or
                                                     and budget policy    Sovet           which deputies                                  deputy group majorities.
                                                     votes)               Federatsii,     belonged, margin,
                                                                          Gosudarstvenn
                                                                          aya Duma,
                                                                          Spravochnik
                                                                          (Moscow, 000
                                                                          Panorama,
                                                                          1996).

Ames      Brazil   Swing versus   1986-1990 deputies   whether deputy     Joint           municipal               percent migrants,       "The higher the percentage of a
(1995b)            core                                submitted          Commission      dominance (vote         social "match"          municipality's votes a deputy won in
                                                       budgetary          on the Budget   share), distance        variables on            1986, the more likely that deputy was to
                                                       amendment to       (432)           from municipal          income                  pursue more support in the same place in
                                                       particular                         center (vote            distribution,           1990. The negative slope on the squared
                                                       municipality in                    center), distance       government              term means that deputies at some point
                                                       1989 and 1990                      from personal           employees,              regard a municipality as 'locked up', thus
                                                                                          center (where           population;             meriting no additional effort...the actual
                                                                                          candidate got most      interparty              inflection points were beyond nearly all
                                                                                          of his votes),          fragmentation,          cases" (421).
                                                                                          percent of vote to      intraparty
                                                                                          retired deputies        fragmentation, rank
                                                                                                                  in party list, career
                                                                                                                  variables
Arulampala India     Swing versus   1974–75 to major       grants (central     Reserve Bank swing dummy,        annual rainfall,       "A state that is aligned and swing in both
m,                   core           1996–97    states (14) plan scheme and     of India       alignment dummy   state population,      the last Vidhan Sabha and the last Lok
Dasgupta,                                                  centrally           Bulletin 1974-                   proportion of state    Sabha elections is estimated to receive
Dhillon,                                                   sponsored           1996, Finance                    population             19.6% (exp(0.179)−1) higher central
and Dutta                                                  scheme)             Commission,                      characterized as       grants than a state that is unaligned and
(2009)                                                                         National                         scheduled caste,       non-swing with respect to both election
                                                                               Accounts                         per capita state of    types. State that is aligned and swing in
                                                                               Statistics                       agriculture in state   last election receives more grants
                                                                               (Government                      domestic product,      regardless of swing characteristics.
                                                                               of India,                        state's per capita
                                                                               Ministry of                      allocation of
                                                                               Planning,                        proceeds of central
                                                                               Department of                    taxes
                                                                               Statistics,
                                                                               1974–1996)


Brusco,    Argentina Swing versus         1999 individual Probability of a     surveys        Peronist handout, income, education,     Only about 7% of whole sample received
Nazareno,            core                                 Peronist Vote in     conducted by   Peronist          housing, gender        goods and 12% of them are low-income
and Stokes                                                Presidential and     authors in     sympathizer,      age, population        class, majority of respondents said they
(2004)                                                    Gubernatorial        2001-2002      radical                                  pay attention to programmatic appeals
                                                          Elections and                       sympathizer                              more than in the past. "The greater vote-
                                                          types of targeting                                                           buying effort by the Peronist Party, and
                                                          behavior (i.e.                                                               the greater effectiveness of these efforts,
                                                          patron, gift,                                                                reflect, we believe, the party's deeper
                                                          Puntero, job)                                                                penetration of lower- class social
                                                                                                                                       networks and hence its greater ability to
                                                                                                                                       monitor voters." (73)
Costa-i-   Mexico   Swing versus   1990-95    states (32) real per capita   statistical     share of local        income,                   robust and positive effect of public
Font,               core                                  public investment annexes to the municipalities         population,               investment on the support for PRI, the
Rodriguez-                                                allocated         Presidential    affiliated to central percentage rural          more educated population, the lower the
Oreggia,                                                                    Address to the government’s                                     support, negative and significant effect of
and                                                                         Nation          party, dummy for                                the governor of the state being of a
Lunapla                                                                     (various years) states governed by                              different political affiliation to the PRI
(2003)                                                                                      political party
                                                                                            other than PRI,
                                                                                            dummy for
                                                                                            Campeche and
                                                                                            Tabasco (two
                                                                                            major recipients of
                                                                                            public investment
                                                                                            for oil field
                                                                                            exploitation)



Crampton   Canada   Swing versus   1996-2000 electoral    job grants by        Transitional   MP's margin of         unemployment           "I find that the strong hypothesis that
(2004)              core                     district     Human Resources      Jobs Fund,     victory in previous    variables including    districts represented by the governing
                                             (595)        Development          Canada Jobs    election, whether      whether threshold      party will fare better than those
                                                          Canada per           Fund by        MP runs for            level was met, and     represented by the opposition is not
                                                          district during an   Human          reelection in the      by how much            supported. Overall, Liberal districts
                                                          electoral cycle      Resources      subsequent             threshold was          fared more poorly than did the districts of
                                                                               Development    election, whether      exceeded, regional     other parties. Instead distributive
                                                                               Canada         MP has been            dummies, district      politics worked in more subtle ways;
                                                                                              selected into the      population,            channeling monies to the districts of
                                                                                              Cabinet or             indicator for pre or   vulnerable Cabinet Ministers and shoring
                                                                                              Shadow Cabinet         post 1997 election     up Liberal fortunes in the west, to the
                                                                                              of his or her party,                          expense of Liberals in the rest of the
                                                                                              party                                         country and, perhaps, of top members of
                                                                                              identification,                               Opposition Shadow Cabinets. Neither the
                                                                                              interaction                                   swing voter nor loyal voter hypothesis
                                                                                              between party                                 can be supported wholeheartedly; where
                                                                                              identification and                            distributive politics played out most
                                                                                              margin of victory                             strongly, evidence supports the swing
                                                                                                                                            voter hypothesis."
Crisp and   Colombia Swing versus   July 1998-   senators    which             Senate Travel   vote share,             destination appeal    "Senators prefer maintaining their
Desposato            core           July 1999    (100)       "department"      Office          security of seat        (political            existing constituency to developing a
(2004)                                                       (district)                        (margin of              kidnappings, beach    new one" (147) "Where senators do
                                                             incumbent                         victory), political     resort,               branch out, they avoid seeking votes in
                                                             senators traveled                 ambition,               bogota/cundinamar     others' bailiwicks" (148) "Politicians go
                                                             to                                "dominance" (sum        ca)                   where the voters are" (they travel more to
                                                                                               of all candidates'                            more populated districts) (150)
                                                                                               squared vote                                  Candidates running for reelection visit
                                                                                               percentages,                                  their core constituencies more often than
                                                                                               indicating whether                            those who are retiring (150). Candidates
                                                                                               one or a few                                  with a low margin of victory make many
                                                                                               candidates                                    more trips than those in safe seats (150)
                                                                                               dominate the
                                                                                               district), electorate
                                                                                               size


Dahlberg    Sweden   Swing versus         1998 municipalit   grants/transfers Ministry of      To test                 Vacancy rate          "Strong support for the prediction
and                  core                      ies that      for sustainable  Environment      "LW/DL" cutpoint        (vacant jobs          derived from the LW/DL model saying
Johansson                                      applied for   ecological                        density of              divided by            that the incumbent government purchases
(2002)                                         grants        development and                   ideological             unemployed            votes by investing in those municipalities
                                               (115)         for increased                     distribution,           persons), tax base,   where there are a lot of swing voters"
                                                             municipal                         closeness of last       social welfare        (38). Support using both "cutpoint
                                                             employment                        election as given       spending per          density" and distance between blocs. No
                                                                                               absolute value of       capita, cash flow,    support for "the hypothesis that the
                                                                                               difference in vote      population share      incumbent government purchases votes
                                                                                               shares. To test         under 16,             by investing in their own
                                                                                               Cox-McCubbins :         population share      supporters...presented by Cox and
                                                                                               dummy if socialist      under 64,             McCubbins" (38).
                                                                                               majority in             environmental
                                                                                               municipal council,      rating, vote share
                                                                                               vote share for          of environmental
                                                                                               incumbent (Social       party
                                                                                               Democrats)
Diaz-      Mexico    Swing versus   1989-1994 municipal     per capita                            core voter, erosion spatial lag,           total per capita transfers averaged over
Cayeros,             core                                   municipal level                       (decline), census- population (lag)        six years shows weak relationship with
Estevez,                                                    expenditures, per                     based measure of                           the municipal welfare levels (CONAPO),
Magaloni                                                    capita private and                    poverty or social                          larger transfers are strongly associated
(2008)                                                      public goods,                         marginality                                with PRI's traditional strongholds in
                                                            share of private                      (CONAPO), PRI                              municipal elections. "Public goods
                                                            goods within total                    vote margin                                should be assigned to more competitive
                                                            Pronasol                              (1988)                                     municipalities." (156-157) "Discretional
                                                                                                                                             private transfers are preferable to
                                                                                                                                             collective goods to construct electoral
                                                                                                                                             coalitions that are stable over time."
                                                                                                                                             (171)


Dunning    Argentina Swing versus          2003 individuals received items   Survey               usually' vote for                          "Argentina's political machine targets
and Stokes           core                       (500)       (clientelist                          Peronist party,                            supporters when it uses its resources to
(2008)                                                      benefits)                             'usually' vote for                         simulate turnout and opponents when it
                                                            (formal model on                      Radical (UCR)                              seeks to buy the vote of people who it
                                                            certain vs.                           party                                      knows are highly likely to vote." (24)
                                                            potential voters
                                                            and empirical)
Golden and Italy     Swing versus   1953-1994 electoral     total                government-                                                 goods go to districts with influential
Picci                core                     district      infrastructure       administered                                                legislators and not to core districts of
(2008)                                        (32)          investments          quarterly                                                   ruling parties
                                                                                 survey of
                                                                                 subnational
                                                                                 public offices
Horiuchi   South     Swing versus   1993-2003 municipalit total transfers per    Annual Report vote share of           fiscal                "Winning president's vote share has a
and Lee    Korea     core                     y/year      capita from            of Local       winning candidate      independency          negative effect on total transfers, whereas
(2008)                                        (864)       central                Finance (1992-                        index, population     its squared term has a positive effect."
                                                          government to          2002),                                density (logged),     (15) The quadratic relationship between
                                                          municipality           published by                          ratio of population   vote and money is highly significant
                                                                                 Ministry of                           over 65 (logged),     before 1997, but not significant after.
                                                                                 Government                            agricultural          Results consistent with their formal
                                                                                 Administration                        workers per capita    model that shows incumbent president
                                                                                 and Home                              (logged),             allocates a larger amount not only in his
                                                                                 Affairs                               manufacturing         own turf but also in rival's. (17)
                                                                                                                       workers per capita
                                                                                                                       (logged), dummy
                                                                                                                       for Seoul
Idema       OECD        Swing versus   1991-2001 countries    public tertiary   Education at a   percentage of         total number of        "This suggests that the higher the
(2009)      countries   core                     (14)         education         Glance, OECD     higher educated       students enrolled as   education level of swing voters of parties,
                                                              spending per                       individuals           a % of population,     the more likely parties are to increase
                                                              student as a                       amongst core and      government             higher education spending when in
                                                              percentage of per                  swing voters of the   partisanship,          office." (15)
                                                              capita GDP                         parties in            government
                                                                                                 government,           expenditures as a
                                                                                                 divided by the        % of GDP
                                                                                                 average percentage
                                                                                                 in society, average
                                                                                                 level of education
                                                                                                 of core and swing
                                                                                                 voters,


Johansson   Sweden      Swing versus   1981-1995, municipalit grants           1991 and 1994     difference between    per capita taxable     Municipalities with many swing voters
(2003)                  core           1992-1995 ies (255)                     Swedish           two blocs in last     income, share of       get larger transfers when using cutpoint
                                                                               Election Study,   election, cutpoint    young (<19)            density (significant at 10%, tested over
                                                                               Statistics        density               inhabitants, share     1992-1995). Closeness of election has
                                                                               Sweden                                  of old (>64)           no statistically significant effect (whether
                                                                                                                       inhabitants,           1981-1995 or 1992-1995).
                                                                                                                       population density
Khemani     India       Swing versus   1972-1995 states (15) transfers per      Reserve Bank     affiliation, seat     state population,      "We find that plan grants are directed
(2003)                  core                                 capita (statutory, of India                               income                 towards affiliated states as the model
                                                             plan grants, and Bulletin                                                        predicts, and furthermore towards those
                                                             central themes)                                                                  affiliated states that contribute a smaller
                                                                                                                                              share of the total number of state
                                                                                                                                              legislators in the national assembly for
                                                                                                                                              the national ruling party." (21) National
                                                                                                                                              political parties target greater resources
                                                                                                                                              to those states where they have more
                                                                                                                                              seats to gain in the national legislature,
                                                                                                                                              rather than to states that are their core
                                                                                                                                              support bases, and hence pivotal for
                                                                                                                                              winning a majority." (26)
McGillivra United     Swing versus   U.S. 1979,   district     amount of           based on 1963    district            District       "marginal electorates do appear to secure
y          States and core           Canada       (Electoral   protection          MUSIC            comparative         Unemployment   more favorable levels of protection in
(1997)     Canada                    1970         district     measured by         classification   advantage, district                Canada's high party discipline system. In
                                                  model),      tariff protection   scheme, a        marginality,                       the U.S., a low party discipline system,
                                                  industry                         Standard         number of districts                safe districts with senior representatives
                                                  (industry                        Industrial       in State, district                 receive slightly higher levels of
                                                  district                         Classification   party                              protection." (603)
                                                  model)                           (SIC) based      representative,
                                                                                   import code      district
                                                                                   and SIC 1972     represented by a
                                                                                   code.            minister, marginal
                                                                                                    districts
                                                                                                    represented by a
                                                                                                    Minister, seniority
                                                                                                    of representative,
                                                                                                    member of Ways
                                                                                                    and Means
                                                                                                    Committee,
                                                                                                    Member of
                                                                                                    Foreign and
                                                                                                    Interstate
                                                                                                    Commerce
                                                                                                    Committee
McIntosh    Uganda   Swing versus   2001 district   loans and   FINCA/Ugand    surprise vote     linear time trend   "we find evidence that investment jumps
and Allen            core                           spending    a, the         share, borrower                       immediately following the presidential
(2009)                                                          country’s      loan cycle,                           election in a manner correlated with the
                                                                largest        violence, threats                     size of the surprise support for the victor,
                                                                microfinance   against                               Yoweri Museveni." The surprise vote
                                                                lend           opposition/movem                      share has a significant effect on both
                                                                               ent/citizen,                          borrowing and savings behavior. The
                                                                               election results                      implication is that counties that backed
                                                                               contested, close                      Museveni to a surprising extent saw
                                                                               vote, movement                        significant increases in both investment
                                                                               promises                              and illiquid savings, indicating an
                                                                               patronage,                            environment that is both better for
                                                                               opposition                            investment and more
                                                                               promises                              secure....Intriguingly, the effects of
                                                                               patronage                             patronage promises by opposition
                                                                                                                     politicians seem to be larger in both
                                                                                                                     magnitude and significance than those
                                                                                                                     made by the Movement." (19) "We find
                                                                                                                     evidence that the behavior of
                                                                                                                     microentrepreneurs is more sensitive to
                                                                                                                     voting surprises among core rather than
                                                                                                                     swing voters." (31)
Miguel and Ghana     Swing versus   1996 and    schools    total school       World Bank's    dummy if NCD          household            At the administrative district level:
Zaidi                core           2000        (192)      funding, and       pilot Public    (national ruling      expenditure per      "Schools in administrative districts
(2003)                              elections,             logged total       Expenditure     party) swept all      capita, proportion   where the NDC swept the parliamentary
                                    school                 school funding     Tracking        parliamentary         urban, years of      elections received 27 percent higher
                                    funding in                                Survey          seats in an           education for head   funding per pupil than other schools",
                                    1998 - 1999                                               administrative        of household,        which is significant at the 99 percent
                                                                                              district in 1996      proportion school-   level (16) Negative and significant
                                                                                              election, average     aged (5-15)          coefficient on quadratic of average
                                                                                              margin of victory     children who have    margin of victory (17) Patronage and
                                                                                              in 1996, vote share   attended school in   possibly public choice effects. At
                                                                                              of NCD                last year,           parliamentary seat level : null results for
                                                                                                                    household            political variables
                                                                                                                    educational
                                                                                                                    expenditures per
                                                                                                                    school-aged child,
                                                                                                                    proportion
                                                                                                                    individuals who
                                                                                                                    have received
                                                                                                                    vaccinations,
                                                                                                                    religion
                                                                                                                    fractionalization
                                                                                                                    index, ethno-
                                                                                                                    linguistic
                                                                                                                    fractionalization
                                                                                                                    index


Milligan    Canada   Swing versus   1988-2001 electoral    spending through   Atlantic        government seat, economic controls         That political factors explain some
and Smart            core                     districts in agencies           Canada          vote margin,                               portion of the variation in regional
(2005)                                        5                               Opportunities   government seat *                          development assistance that economic
                                              regions/yea                     Agency,         vote margin,                               factors cannot. Positive and significant
                                              r (1344)                        Economic        cabinet minister,                          effects for government seat and cabinet
                                                                              Development     years of                                   minister variable.
                                                                              Agency of       experience by MP,
                                                                              Canada for
                                                                              Quebec
                                                                              Regions
Nichter   Argentina Swing versus   survey       voters   received rewards survey         Peronist                  Income, education,     Contrary to conventional wisdom about
(2008)              core           conducted (1920)      by a party in the conducted by sympathizer,               housing quality,       vote buying, parties can activate their
                                   in Dec 2001-          last campaign     Stokes (2005) opinion of                log population,        own passive constituencies by rewarding
                                   Jan 2002                                              Peronists,                ballot, age, gender    unmobilized supporters for turnout.
                                                                                         Peronists rating,                                Turnout buying involves a less stringent
                                                                                         1995/9 Peronist                                  monitoring requirement than vote buying -
                                                                                         voter                                            the ability to observe turnout instead of
                                                                                                                                          voting decisions. Formal models show
                                                                                                                                          turnout buying is incentive-compatible.
                                                                                                                                          (29)


Park      South     Swing versus   1990-1999 electoral   intergovernmental Financial        assembly               fiscal independence    Before decentralization (1991-95), none
(2002)    Korea     core                     district    transfers         Yearbook of      member's               index (share of        of the political variables are significant.
                                                                           Local            alignment with the     local tax and non-     After decentralization, only degree of
                                                                           Governments      incumbent              tax revenue to total   political competition becomes
                                                                           by Ministry of   administration,        revenue), per capita   significant. (16) "before
                                                                           Government       each member's          inhabitant tax,        decentralization, highly competitive
                                                                           Administration   seniority, political   ratios of population   districts (swing) received favorable
                                                                           and Home         competitiveness        receiving running      distribution, and highly competitive
                                                                           Affairs                                 water, ratios of       districts with their assembly members
                                                                                                                   people engaged in      from the incumbent party received
                                                                                                                   agriculture,           negative treatment in the distribution of
                                                                                                                   population density     National Subsidy." (18)
Rodden and Brazil   Swing versus   1991-2000 states (26) constitutional      Ministry of   total number of        real state domestic   "The margin separating the top two vote
Arretche            core                                 transfers, total    Finance       votes received by      product per capita,   recipients in the state in the last
(2004)                                                   (current + capital)               the president in       state population,     presidential election has a negative effect
                                                         discretionary                     the most recent        dummy for             on discretionary transfers... states that
                                                         transfers, total                  election, distance     presidential          provided more votes for the president in
                                                         transfers                         between vote           election years        the most recent election received larger
                                                         (constitutional                   shares of top two                            per capita transfers" (22) "Inserting an
                                                         transfers +                       presidential                                 interaction term (margin x votes for
                                                         current                           candidates in most                           president) in model 4 helps solve the
                                                         discretionary +                   recent election,                             puzzle. The positive conditional
                                                         capital                           effective turnout                            coefficient for president’s vote is
                                                         discretionary)                    proxied by                                   significant and increases slowly with the
                                                                                           unspoiled ballots                            margin of victory. However, the
                                                                                           cast as a share of                           coefficient for margin is only negative for
                                                                                           voting age                                   states where the president failed to garner
                                                                                           population,                                  more than one million votes, and even in
                                                                                           average of number                            those it is not statistically significant
                                                                                           of seats per                                 over much of the sample range. When the
                                                                                           million inhabitants                          president receives over 1 million votes,
                                                                                           for Senate and for                           the conditional coefficient for margin
                                                                                           Chamber of                                   actually becomes positive, and it is
                                                                                           Deputies, number                             significant when the number surpasses 2
                                                                                           of the state's                               million. Thus for the large states, the
                                                                                           deputies that                                effect of margin is positive, and the
                                                                                           belong to                                    president won these states in 90 percent
                                                                                           president's                                  of the cases (the large margin of victory
                                                                                           legislative                                  was his own). The results are consistent
                                                                                           coalition as share                           only with a majoritarian or “winner takes
                                                                                           of total legislative                         the spoils” perspective. Cardoso appears
Rodden and India     Swing versus   1957-2003, states (28) sum of transfers Reserve Bank        for eras of            logged real state      for periods of Congress dominance:
Wilkinson            core           1972-2003              for central plan    of India         Congress               domestic product       "Discretionary resources are apparently
(2004)                                                     schemes and         Bulletin         dominance: Congr       per capita, percent    targeted at states where Congress has the
                                                           centrally                            ess's share of         of population living   strongest support, but also at states that
                                                           sponsored                            state's parliament     in urban areas,        are relatively evenly divided between
                                                           schemes, real                        delegation (proxy      dummy if under         Congress and opposition parties…the
                                                           loans, total grants                  for core support),     "president's rule"     only clear losers are states that are
                                                                                                absolute difference    (emergency central     dominated by opposition parties...we also
                                                                                                between                rule), dummy if        find that states controlled by Congress
                                                                                                Congress's share       state is a "special    Chief Ministers receive more
                                                                                                of state's             category" state        discretionary grants than states controlled
                                                                                                parliament             (special               by the opposition" (21-22). "Swing
                                                                                                delegation and         reservations for       states are favored regardless of the
                                                                                                50% (proxy for         minorities)            partisanship of the state chief minister"
                                                                                                swing), dummy if                              contra DDD and Khemani (22). for
                                                                                                state chief minister                          coalition governments: pooled results:
                                                                                                is same party as                              no significant political variables (24).
                                                                                                central                                       Dropping earlier coalitions and looking
                                                                                                government. for                               post-1995: swing proxy not significant,
                                                                                                coalition                                     states with chief ministers of the same
                                                                                                governments:                                  party as prime minister favored, states
                                                                                                share of state                                with governments sharing at least one
                                                                                                parliament that is                            party with national coalition are
                                                                                                chief minister's                              disfavored (authors are "flummoxed" by
                                                                                                party, share of                               this), the states with high numbers of
                                                                                                state parliament                              junior partners and outside supporters of
                                                                                                that is junior                                coalition are favored (25-26)
                                                                                                coalition partners,
                                                                                                share of state
Stokes     Argentina Swing versus         2001 individuals patron (received      Survey         Peronist               income, education,     Based on income, education, housing
(2005)               core                      (1920)      politician's help),   conducted in   sympathizer,           housing quality,       quality evidence, results show that
                                                           job (would turn to    2001-02 in 3   radical                population (log),      political machine buys votes of poor
                                                           party operative       provinces      sympathizer,           gender                 people. In Argentina, the more a party is
                                                           for job), reward                     ballot given by                               able to monitor its voters, the more
                                                           (received goods                      party operative                               effective its vote buying efforts will be.
                                                           in last campaign)                                                                  Parties with the most decentralized
                                                           (formal model on                                                                   organizational structure are best able to
                                                           machine politics)                                                                  monitor. Community structure voter is in
                                                                                                                                              also matters because those in smaller
                                                                                                                                              communities are easier to vote and
                                                                                                                                              receiving the ballot directly from a party
                                                                                                                                              operative is more likely to cause effective
                                                                                                                                              targeting.
Vaishnav     India     Swing versus   1977-2007 constituenc public school            District            ruling coalition,                         Ruling Members of Legislative
and Sircar             core                     ies         construction in          Information         margin of victory,                        Assembly in Tamil Nadu have preference
(2009)                                          (210)/year Tamil Nadu                System for          interaction                               to target club (public) goods to core voter
                                                                                     Education           between the two                           group. MLAs immediately reward
                                                                                     initiative of the                                             compliant voters post-election with pork
                                                                                     National                                                      in the form of new schools. "We find
                                                                                     University of                                                 that there are deviations from the "core"
                                                                                     Educational                                                   targeting strategy in those years with
                                                                                     Planning and                                                  exceptionally high levels of electoral
                                                                                     Administration                                                competition. When more than half the
                                                                                     .                                                             ruling coalition's victories come in swing
                                                                                                                                                   constituencies, MLAs alter their post-
                                                                                                                                                   election targeting strategy to reward
                                                                                                                                                   pivotal swing areas." (26)


Ward and     England   Swing versus   FY 1994-    larger         authority's                             need variables       population, square   "government allocated around £500
John                   core           1995        English        standard spending                       assessed by factor   of population,       million more to local authorities
(1999)                                            local          assessment, a                           analysis derived     dummy for unitary    containing marginal constituencies and
                                                  authorities    figure calculated                       from various         authorities, dummy   around £155 million more to 'flagship'
                                                  including      by a known                              indicators such as   for London           local authorities than they could have
                                                  London         formula                                 age-structure of     boroughs, dummy      been expected to get on the criteria of
                                                  boroughs,      aggregating                             population,          for shire counties   social need and population." (32)
                                                  metropolita    measures of                             dummy for
                                                  n districts,   spending need                           whether or not the
                                                  and shire                                              government
                                                  counties                                               controls the local
                                                  (108)                                                  authority based on
                                                                                                         whether
                                                                                                         Conservative
                                                                                                         councillors had a
                                                                                                         majority in
                                                                                                         1994/95, dummy
                                                                                                         for whether local
                                                                                                         authority is
                                                                                                         government
                                                                                                         flagship, marginal
                                                                                                         constituency
                                                                                                         variable
Winters   Ecuador   Swing versus   2006-2007 canton      change from 2006     2002           vote share of         percentage of         "a percentage point increase in the vote
(2009)              core                     (209)       to 2007 in the       Identificacion Correa in first and   individuals living    for Correa during the first round
                                                         number of Bono       y Eleccion de second round           in poverty in each    corresponds to a greater increase in BDH
                                                         de Desarrollo        Beneficiarios                        canton, percentage    than would the same percentage point
                                                         Humano (BDH)         (SELBEN)                             of population that    increase in the poverty rate." (7) "I have
                                                         recipients           survey                               is of indigenous      shown a repeated association between the
                                                                              conducted by                         origin, total         vote share that Correa obtained in a
                                                                              government                           number of             canton during the elections and the
                                                                                                                   recipients in 2006,   increase in the number of BDH transfers
                                                                                                                   population            after the elections...larger rewards for
                                                                                                                                         being a pro-Correa canton during the
                                                                                                                                         runoff election as compared to the first-
                                                                                                                                         round election..."(12)


Baldwin   Mali      Political      1992-1994 electoral   change in            L'Enquete       percentage of        income per capita,    "No support for the hypothesis that
(2005)              favoritism               districts   percentage of the    Malienne        votes obtained by    unemployment,         legislators distribute primarily to their
                                             (42)        population with      Conjoncture     Adema (political     literacy rate,        core supporters…little support for the
                                                         government or        Economique et   party) in the        urbanization level,   theory that legislators redistribute
                                                         public sector jobs   Sociale         second round of      percent of            primarily to swing districts" (20).
                                                                                              elections,           population            "…greater support for the hypothesis that
                                                                                              difference between   speaking various      President Konare targeted swing voters.
                                                                                              Adema's vote         subnational           The coefficient on unemployment is
                                                                                              share and 50%,       languages             positive" and coupled with a positive
                                                                                              turnout (all for                           coefficient on income, "something other
                                                                                              both presidential                          than a benign attempt to reach the most
                                                                                              and legislative                            needy is at work here" (23) "Because
                                                                                              elections)                                 Konare won more than 50 percent of the
                                                                                                                                         second round votes in all of the regions
                                                                                                                                         in my sample, the finding that
                                                                                                                                         government jobs and presidential support
                                                                                                                                         [vote share] are negatively correlated is
                                                                                                                                         equivalent to the finding that more
                                                                                                                                         government jobs went to areas where the
                                                                                                                                         presidential race was closer" (23). Posits
                                                                                                                                         her own alternative explanation, beyond
                                                                                                                                         core/swing, that government jobs went to
                                                                                                                                         potentially destabilizing regions (24)
Banerjee  India   Political    1971, 1991 parliament    1971-1991          Census of      political           land inequality,        "Most striking difference between 1971
and               favoritism              ary           change in the      India          fragmentation,      scheduled tribes,       and 1991 is gains registered by the
Somanatha                                 constituenc   fraction of                       social              scheduled castes,       Scheduled Castes relative to other
n                                         ies           villages with the                 fragmentation       Brahmans,               groups. Higher shares of Scheduled
(2007)                                                  public good                                           Muslims,                Castes in 1991 are associated with more
                                                        (education/health/                                    Christians, Sikhs,      high schools, health centers,
                                                        water)                                                Avg. village            dispensaries, and piped water." (308)
                                                                                                              population, number      Social fragmentation has no systematic
                                                                                                              of villages             effect in 1991 (unlike in 1971 where it
                                                                                                                                      was negative). Estimated effect of land
                                                                                                                                      inequality has systematically switched
                                                                                                                                      sign- coefficient on Gini Index is
                                                                                                                                      negative in 5 cases and positive in 2 in
                                                                                                                                      1991--whereas it was positive in 1971.
                                                                                                                                      Political fragmentation in 1991 and 1971
                                                                                                                                      do not systematically affect the provision
                                                                                                                                      of most goods.


Banerjee, India   Political          1991 districts     share of villages   1991 Census   Non-British         geographic factors      Non-British coefficient positive for 14
Iyer, and         favoritism              (391 in 16    with a particular   of India      dummy, fraction     affecting public        goods out of 26 and negatively
Somanatha                                 major         public good (26                   of land not under   good transport ie.      significant for 4, nonlandlord effect
n                                         states of     categories)                       landlord control,   Average annual          positive for 15 and negative for 3, and
(2005)                                    India)                                          fragmentation       rainfall, maximum       caste and religious fragmentation
                                                                                          index               and minimum             coefficient negative for 10 and positive
                                                                                                              temperatures,           for 4. "British areas do significantly
                                                                                                              dummy for coastal       worse than non-British areas." (646)
                                                                                                              districts, proportion
                                                                                                              of wasteland area
                                                                                                              barren, rocky,
                                                                                                              sandy, or
                                                                                                              mountainous,
                                                                                                              average village
                                                                                                              population, number
                                                                                                              of villages in the
                                                                                                              district,
                                                                                                              demographic
                                                                                                              controls, land
                                                                                                              ownership
Banful   Ghana   Political    1994 to   district   per capita District   internal          binary variable        population,           "DACF formula rules are followed in
(2008)           favoritism   2005      (approx    Assemblies            documents         which is 1 if          population density,   calculating the districts' allocation,
                                        1000-1300) Common Fund           from the          ruling government      ethnic                however, the fund that the districts
                                                   (DACF)                headquarters of   won the district in    fragmentation,        receive, their disbursements, is generally
                                                   disbursement          the District      last presidential      literacy rates        different from this amount…
                                                   logged, per capita    Assembly          election; political                          Counterfactual allocations calculated
                                                   DACF allocation       Common Fund       cycle variables                              suggest that the DACF formula was
                                                   logged,               in Accra,         including                                    purposefully manipulated as most of the
                                                   disbursement/allo     Ghana             presidential                                 major changes occurred in election
                                                   cation;                                 election year                                years.... Growth in DACF disbursement
                                                   performance of                          dummy, dummy                                 is highest in an election year and lowest
                                                   the District                            for one year after                           two years after an election."(22)
                                                   Assembly                                presidential                                 "There is apparently no political
                                                   (percentage                             election;                                    motivation in the first and as yet only re-
                                                   utilization of                          counterfactual                               demarcation of districts in Ghana after
                                                   disbursement,                           allocation variable                          the implementation of the DACF. " (23)
                                                   administration                          calculating fund
                                                   costs, total                            using formula
                                                   expenditure,                            from previous
                                                   proportion of                           year; political
                                                   expenditure spent                       motivation
                                                   on administration                       variables
                                                   or proportion of                        including dummy
                                                   disbursement                            whether a district
                                                   spent on                                was split, political
                                                   administration)                         party that won the
                                                                                           district
Besley,    India   Political    survey     households which households "Independent           Household            Household              "Living in a reserved GP increases a
Rahman,            favoritism   taken in   , villages received "low-      audit of village    regressions:         regressions:           SC/ST household's likelihood of getting"
Pande, and                      September- (396)      spillover" public facilities",          SC/ST dummy          dummies for            a "low-spillover" public good (423).
Rao                             November              goods (i.e., had a surveys of           (see comments),      Christian or           "Household public goods have low
(2004)                          2002                  house or toilet     households          SC/ST reservation    Muslim household,      spillovers and are targeted toward
                                                      built), which                           dummy, Pradhan's     household size,        SC/ST" (425). "…The Pradhan enjoys
                                                      villages received                       village dummy,       age, literacy,         agenda setting power in resource
                                                      "village-level"                         government           occupation of          allocation" (425)
                                                      public goods (i.e.                      headquarter          household head,
                                                      construction or                         dummy. Village-      household of
                                                      improvement on                          level regressions:   elected official,
                                                      within-village                          proportion SC/ST     village fixed
                                                      roads, drains,                          households,          effects, Village-
                                                      streetlights, water                     Pradhan village      level: proportion of
                                                      sources)                                dummy, Pradhan       landless
                                                                                              reserved for         households, log
                                                                                              SC/ST                total village
                                                                                                                   population, log
                                                                                                                   village area,
                                                                                                                   proportion of
                                                                                                                   irrigated land,
                                                                                                                   village literacy
                                                                                                                   rate, distance from
                                                                                                                   nearest town, daily
                                                                                                                   male agricultural
                                                                                                                   wage rate, block
                                                                                                                   fixed effects


Betancourt India   Political    1977-78     districts    allocation of        Indian          % of eligible        % of scheduled         "features of each state common to all
and                favoritism               (325)        educational and      Development     population voting    caste, % Muslim,       districts within a state are important
Gleason                                                  medical services     District Data   in 1977-8 state      agricultural           determinants of the provision of these
(2000)                                                   (doctors/nurses/te   compiled from   assembly             productivity,          two public inputs to rural households. All
                                                         achers per 10        1981 Indian     elections,           proportion             17 state dummies are significantly
                                                         population)          Census and      female/male voter    landless, rural land   different from zero at the 1% level."
                                                                              other sources   turnout ratio,       per person,            (2175) "Existence of statistical
                                                                                              number of voting     proportion urban       discrimination in the outcomes of the
                                                                                              constituencies                              allocation process for medical services
                                                                                                                                          on the basis of caste and
                                                                                                                                          religion....Bureaucratic aspects of
                                                                                                                                          behavior play an important role in
                                                                                                                                          determining the outcomes of the
                                                                                                                                          allocation process for medical services."
                                                                                                                                          (2177)
Bettcher   Japan and Political    Japan: 1947 factions (2)   experience of    (non-              organizational                               "clientels and factions of interest are
(2005)     Italy     favoritism   to 1993                    factions in      regression         dimension                                    based primarily on patronage, while
                                  Italy: 1964                Japan's Liberal  analysis)          (strongly                                    tendencies and factions of principle are
                                  to 1982                    Democratic Party                    organized                                    based primarily on policy beliefs. Both
                                                             and Italy's                         (clientel) or                                factions of interest and faction of
                                                             Christian                           weakly organized                             principle are more highly developed
                                                             Democracy                           (tendency)), and                             organizationally than clienteles or
                                                             (non-regression                     motivational                                 tendencies." (344)
                                                             analysis)                           dimension                                    "Overall these groups [LDP factions and
                                                                                                 (factions of                                 DC factions] are best classified as
                                                                                                 interest or factions                         factions of interest, especially in the
                                                                                                 of principle)                                1980s. The LDP factions approximated
                                                                                                                                              the ideal type more closely than did the
                                                                                                                                              DC factions." (354)


Cadot,      France   Political    1985-1992 regions          infrastructure   Federation         Number of large,       region GDP, region    "Public goods, even if imperfectly
Röller, and          favoritism             (21)             investments      Nationale des      established firms      employment, non-      'targetable'…are used by politicians as
Stephan                                                                       Travaux            in a region (as        residential private   redistribution instruments" (1147). More
(2006)                                                                        Publics, SNCF      proxy for              capital stock,        extremist voters (Le Pen, hunter party) in
                                                                              (national          lobbying activity).    transportation        a region means the region gets more
                                                                              railway            Proxies for swing      infrastructure        investment per worker -- significant in all
                                                                              company) for       voters : (1)           stock,                specifications. Sharing a party with the
                                                                              railroads,         Absolute               transportation        national government gets a region more
                                                                              Laboratoire        difference between     infrastructure net    investment, but insignificant most of the
                                                                              d'Observation      vote for right and     investments           time Absolute difference between right
                                                                              Economique et      left parties (which,                         and left vote totals is significant and
                                                                              des Institutions   they admit, is not                           negative.
                                                                              Locales            a "swing voter"
                                                                              (OEIL)-            variable but is "a
                                                                              gathered data      measure of how
                                                                              from SNCF for      heated the
                                                                              roads              electoral race
                                                                                                 was"(1135)) and
                                                                                                 (2) Vote for Le
                                                                                                 Pen and another
                                                                                                 extreme party
                                                                                                 (disaffected voters
                                                                                                 are ripe for
                                                                                                 buying). Also a
                                                                                                 dummy where
                                                                                                 local and national
                                                                                                 political
                                                                                                 congruence (for
                                                                                                 targeting "local
Calvo and   Argentina Political    2007 individual voters'        survey         size and structure    socio-demographic   "Our results provide ample support for
Murillo     and Chile favoritism        (5600=280 distributive    conducted by   of voters' personal   status of           the importance that voters attach to local
(2009)                                  0*2        expectations   authors        and political         respondents         benefits in Argentina -- given more
                                        countries)                               networks, political   (measured through   clientelistic linkages and broader
                                                                                 behavior of voters    questionnaire)      networks of the PJ and UCR -- and the
                                                                                 (measured through                         importance of global benefits in shaping
                                                                                 questionnaire),                           the distributive expectations of Chilean
                                                                                 likelihood of                             voters -- under more programmatic party-
                                                                                 receiving                                 voter linkages. Proximity to party
                                                                                 handouts, being                           networks, therefore, is a stronger
                                                                                 offered a job in                          predictor of voters' distributive
                                                                                 public sector                             expectations in Argentina, whereas
                                                                                 (patronage), or                           Chilean voters giver more importance to
                                                                                 witnessing                                ideology in perceived access to benefits."
                                                                                 increased public                          (25)
                                                                                 investment in
                                                                                 community if a
                                                                                 given party wins
                                                                                 the election
                                                                                 (pork),
                                                                                 global and local
                                                                                 benefits perceived
                                                                                 by voters :
                                                                                 proximity of
                                                                                 voters to each
                                                                                 party's network,
                                                                                 ideological
                                                                                 proximity between
                                                                                 voters are parties
Case     Albania   Political    1992-1996 rural    block grants to     Ministry of   Voting records in    number of families   "..more assistance allocated to swing
(2001)             favoritism             communes communes of         Labor         1992 local           requesting           communes…" (408), also more money to
                                          (191)    social assistance                 government           assistance,          "pivotal" communes (districts that voted
                                                   funds                             elections and 1994   commune              for the winning party, which she
                                                                                     constitutional       population, family   considers to be a measure of pivotal, got
                                                                                     referendum.          land holding         larger block grants). "Our predictions
                                                                                     Absolute value       patterns             are consistent with those that would
                                                                                     between                                   obtain in…a Dixit and Londregan
                                                                                     Democratic vote                           framework." (409)
                                                                                     and 50% as
                                                                                     measure of swing.
                                                                                     Support of 1994
                                                                                     constitutional
                                                                                     referendum taken
                                                                                     as measure of
                                                                                     Democratic
                                                                                     support
Finan       Brazil   Political    1996-2000 municipalit federal outlays,     Federal         municipal vote   deputy               "A deputy's share of votes in a
(2004)               favoritism             ies         and amount of        Chamber of      share, dummy if  characteristics:     municipality influences both the choice
                                                        federal              Deputies and    mayor and federalgender, education, to reward the municipality and the size of
                                                        expenditure          the Senate      deputy are in theoccupation, age,     the public work" (13). "Being of the
                                                        conditional on                       same political   party affiliation,   same party is positively associated with
                                                        receiving an                         party            civil status,        more and larger public works." (16) "we
                                                        outlay                                                incumbency status. document a strong association between
                                                                                                              Municipality         the allocation of public works during the
                                                                                                              characteristics:     1996-9 electoral cycle and the degree of
                                                                                                              gender, education, electoral support received in the 1994
                                                                                                              party affiliation,   elections. A mere 10 percentage point
                                                                                                              winning              increase in political support implies an
                                                                                                              percentage,          increase of R$75,174 in public works for
                                                                                                              electoral            a municipality during the electoral
                                                                                                              competition for      cycle."(32) Because these public goods
                                                                                                              mayor, electoral     vary in their degree of exclusion,
                                                                                                              competition for      politicians prefer to allocate goods that
                                                                                                              deputies, whether are less excludable to municipalities
                                                                                                              municipality has     where they received more political
                                                                                                              paved roads, shanty- support." (32)
                                                                                                              town, incentives to
                                                                                                              foster economic
                                                                                                              activity, a job
                                                                                                              program, a
                                                                                                              constitution,
                                                                                                              zoning laws,
                                                                                                              budget plan,
                                                                                                              property laws,
                                                                                                              work laws,
Fukui and   Japan    Political    1993-1994 prefectures local grants-in-   personal        influence of Diet                       "Of the four factors discussed above that
Fukai                favoritism                         aid funds and      interviews with members and                             appear to affect significantly, if not
(1996)                                                  public works-      local           national                                determine a Diet member's role as an
                                                        related funds      politicians and bureaucrats, party                      intermediary between a local
                                                        (descriptive, non- aides to        organization,                           constituency and the national treasury,
                                                        regression         members of      impact of 1994                          personal reputation and "tribal"
                                                        analysis)          Parliament      electoral reform                        connections are likely to remain
                                                                           conducted in                                            important assets under any electoral
                                                                           Okayama and                                             system.... The electoral Keiretsu or line
                                                                           Toyama in                                               and the noenkai, on the other hand, are
                                                                           1994 summer                                             likely to undergo a significant change,
                                                                                                                                   since they are devices geared primarily to
                                                                                                                                   electoral competition among candidates
                                                                                                                                   of the same party." (284)
Giraudy    Argentina Political    1993-2002 23             percentage of      Instituto        dummy for              roadblocks,          Partisanship at both the federal and
(2007)               favoritism             provinces      total population   Nacional de      Peronist President     unemployment         provincial level, together with
                                            and Ciudad     that received      Estadisticas y   and                    (lagged), Economic   institutional features of Argentine federal
                                            de Buenos      employment         Censos           Radical/Alianzist      crisis (2001-02)     system strongly effect cross-provincial
                                            Aires / year   programs in        (INDEC)          Governor in                                 allocation of employment programs.
                                                           province                            office, presidential                        Positive and significant relationship
                                                                                               election, Senate                            between the number of roadblocks, level
                                                                                               overrepresentation                          of unemployment, 2001-2 economic
                                                                                               (log), House                                crisis, and % of people getting
                                                                                               overrepresentation                          emergency employment programs.
                                                                                               (log)                                       Presidential elections are statistically
                                                                                                                                           insignificant. (46)


Gonzalez- Nicaragua Political           2008 individual incidents of      survey               political        income, education,         Although we have no conclusive answer
Ocantos,            favoritism               (approx    individual gift,  conducted by         participation,   age, gender,               in relation to the partisan determinants of
Kiewiet de                                   3000)      neighborhood gift authors              support FSLN or urban/rural                 clientelism, the results suggest that there
Jonge,                                                                                         PLC, welfare and                            are two central mechanisms for
Meléndez,                                                                                      public works,                               clientelistic linkages in the Nicaraguan
Osorio, and                                                                                    attended Citizen                            context: 1) voter's perception that parties
Nickerson                                                                                      Power Councils,                             can monitor their vote and 2) voter's
(2010)                                                                                         monitor                                     participation or proximity to state
                                                                                                                                           sponsored community like the Consejos
                                                                                                                                           de Poder Ciudadano." (27)
Horiuchi   Japan     Political    FY1994-5   districts   per capita grants   account            representative in      per capita income    Under Japan's parliamentary system, and
(2003)               favoritism   (FY runs   (126)       from central to     settlement         ruling coalition       (log), total         the single non-transferable vote electoral
                                  Apr to                 municipal           (Chiho Zaisei      measured by            population (log),    rule, strong evidence supporting
                                  March)                 government for      Kenkyu Kai),       whether he voted       age over 65          "universalism within districts" or
                                                         specific projects   various issues     for the leader of      population ration,   members of the opposition share
                                                         (Kokko                                 the ruling coalition   agriculture sector   distributive benefits within each district.
                                                         shishutsukin,                          in the election for    workers ratio,       Also, districts with more "powerful"
                                                         differenced, log)                      the Prime                                   members of the ruling coalition receive a
                                                                                                Minister,                                   larger amount of grants than other
                                                                                                seniority, num. of                          districts, namely the number of
                                                                                                former                                      attendances to Budget Committee,
                                                                                                bureaucrats, num                            seniority, and whether a representative is
                                                                                                of former local                             a former high-ranking official in local
                                                                                                officials, num. of                          government are particularly important.
                                                                                                former local                                (17)
                                                                                                politicians, num.
                                                                                                of former
                                                                                                secretaries



Indridason Iceland   Political    1945-2000 cabinet      coalition           Morgunblaoio       prevalence of                               "It can be hypothesized that in the
(2005)               favoritism                          formation and       (1945-2000         clientelistic                               presence of clientelism, coalition
                                                         termination         various issues),   politics, political                         agreements will avoid restricting
                                                                             pjooviljinn,       institutions,                               politicians' ability to serve their clients.
                                                                             Grimsson           minority status                             This is consistent with the freedom in
                                                                             (1977),                                                        appointment of non-cabinet positions, the
                                                                             Gislason                                                       use of the inner cabinet as a primary
                                                                             (1993),                                                        solution mechanism, the general
                                                                             Jonsson                                                        vagueness of policy agreements, little
                                                                             (1969),                                                        emphasis on procedural rules, and the
                                                                             Eggertsson                                                     use of negative election rule." (458) "The
                                                                             (1999),                                                        presence of clientelism creates a demand
                                                                             Alpingi                                                        for access to the discretionary
                                                                             website                                                        distribution of public resources that is
                                                                             (www.althingi.                                                 concentrated in the hands of the
                                                                             is)                                                            executive. Politicians therefore behave
                                                                                                                                            as if they were officeseekers rather than
                                                                                                                                            policy-seekers, regardless of their 'true'
                                                                                                                                            preferences." (462)
Kasara     African   Political    1970-1987, crop-      tax rate on a crop Jaeger (1992)   dummy if               variant of black       "When a president comes from a crop-
(2007)     nations   favoritism   1966-1995 country                        and McMillan    president's ethnic     market premium,        growing region, farmers face taxes that
                                             combinatio                    (2001)          group comes from       dummy if in CFA        are higher by 23 points….changing our
                                             ns (50)                                       a major crop           franc zone, real per   proxy for ethnic identification of
                                                                                           growing region,        capita PPP GDP,        farmers…slightly lowers the substantive
                                                                                           dummy if crop is       dummy if country       size of the effect, but does not alter the
                                                                                           crown in the           under IMF              basic finding that farmers associated with
                                                                                           region that            program, world         the current regime face higher taxes"
                                                                                           dominates the          price for crop         (167) Rulers are better at co-opting local
                                                                                           governing              (logged),              allies in their home, which means there
                                                                                           coalition, scale for   percentage of total    are fewer alternative candidates to run
                                                                                           democracy,             agricultural land      against the rulers and the rulers can tax at
                                                                                           dummy if political     area devoted to        a higher rate
                                                                                           competition,           crop, percentage of
                                                                                           average duration       total value of
                                                                                           of tenure of           exports of crop,
                                                                                           leaders,               public investment
                                                                                                                  as percent of GDP,
                                                                                                                  distance from
                                                                                                                  capital to a major
                                                                                                                  town in the region




Kramon     Ghana,    Political    2003        household childhood         Demographic      indicator for          household head's       Findings consistent with our that that the
and Posner Kenya,    favoritism   Ghana,                vaccinations,     and Health       whether household      age, literacy,         answer to the "are ruling party supporters
(2011)     Malawi,                Kenya;                educational       Surveys          in question is         religious, rural,      favored?" question depends on the
           Zambia                 2004                  attainment,                        located in a core      household assets,      outcome one happens to be investigating.
                                  Malawi;               access to                          district of the        material used to       i.e. in Kenya, those who had voted for
                                  2007                  improved water                     ruling party (60%      make household         KANU received more childhood
                                  Zambia                sources,                           of voters)             floor, distance from   vaccinations but less access to water. (5)
                                                        household                                                 the country's          Also results show "who is favored?"
                                                        electrification                                           capital of the         (targeting) depends on "favored relative
                                                                                                                  sampling cluster       to whom?." (7)
Lizzeri and Britain   Political    19th        local     franchise in 19th   changes in English                       "We have presented a model of voluntary
Persico               favoritism   century     governmen century England     public finances                          franchise extension that is based on
(2004)                                         t         that was extended   during age of                            divisions within the elite. In our model
                                                         progressively at    reform                                   these divisions arise endogenously
                                                         the parliamentary                                            through the political process: groups
                                                         level with three                                             within the elite benefit differently from
                                                         Reform Acts of                                               the status quo. We have shown that it is
                                                         1832, 1867, 1884                                             possible that a majority within the elite
                                                         (historical                                                  may favor expanding the franchise.... In
                                                         analysis)                                                    England, substantial subgroups within
                                                                                                                      the elite championed franchise
                                                                                                                      expansion. Our analysis can account for
                                                                                                                      some of the changes in the scope of
                                                                                                                      public sector intervention during the
                                                                                                                      British age of reform, notably the
                                                                                                                      increased provision of public goods and
                                                                                                                      the decline of special interest politics. In
                                                                                                                      our model, these changes are a
                                                                                                                      consequence of the reform process;
                                                                                                                      reform takes place when increased needs
                                                                                                                      for public goods lead to a majority of the
                                                                                                                      elite to demand a redirection of the role
                                                                                                                      of government away from special interest
                                                                                                                      politics toward increased provision of the
                                                                                                                      public goods."(755)




Magaloni,   Mexico    Political    1989-1994 municipalit share of private    level of            time trend, lag of   Clientelism reduces risk for politicians
Diaz-                 favoritism             y (2400)    good provision      development using dependent variable     because they can use preexisting
Cayeros,                                                 (Pronasol)          deprivation index                        networks to target core constituencies.
Estévez,                                                                     from CONAPO,                             Clientelism in poorer areas, and degree
(2006)                                                                       number of parties,                       of clientelism declines as municipality
                                                                             margin of victory,                       grows. Against expectations of
                                                                             risk that shows                          modernization theory, greater recourse to
                                                                             how fast PRI is                          clientelism in municipalities at middle
                                                                             losing votes in any                      levels of development. Clientelism less
                                                                             municipality                             prevalent in politically heterogeneous
                                                                             relative to the                          municipalities. (201)
                                                                             national trend for
                                                                             that party
Neto and   Brazil   Political    1985 to    bills (269   who proposes        PRODASEN   who proposes           "our statistical findings suggest that
Santos              favoritism   1999 and   bills        what : probability             what :                dominant deputies, owing to their close
(2003)                           1995       proposed     of a deputy                    concentration or      links to voters, are less concerned with
                                            from 1985-   proposing at least             dominance of a        national policy-making than are deputies
                                            99 and       one parochial or               deputy's pattern of   who tend to share their constituencies
                                            1291 bills   national-level bill            vote distribution,    with other politicians. "Furthermore, the
                                            in 1995)     in 1995; who                   affiliation with      incentive to delegate national policy-
                                                         gets bill passed:              National Renewal      making to the executive, as expected by
                                                         probability of a               Alliance              the ISM, is, on the one hand, stronger
                                                         deputy getting at              Party/Social          among supporters of the president." (467)
                                                         least one law                  Democratic Party      "The fact that the variable tapping
                                                         passed per                     (AREANA/PDS)          whether or not deputies have held an
                                                         legislature                    who gets bill         office in the Chamber's power structure
                                                                                        passed: affiliation   has a significant effect on their likelihood
                                                                                        with                  of getting a law passed is evidence that
                                                                                        ARENA/PSD,            there is an important distributive
                                                                                        career length,        component in lawmaking, as far as
                                                                                        electoral power,      deputies' legislative output is concerned."
                                                                                        held legislative      (473)
                                                                                        office (members of
                                                                                        Mesa da Camara
                                                                                        dos Deputados),
                                                                                        held executive
                                                                                        office
Remmer      Argentina Political    1983-2003 provinces    patronage         Ministerio de      vote share of        unemployment,         "Party dominance, on the other hand,
(2007)                favoritism                          spending: (1)     Economia y         winning              capital spending,     makes a difference, with two of the three
                                                          total personnel   Produccion         gubernatorial        tax revenue,          coefficients suggesting that the narrower
                                                          spending,                            candidate            discretionary         the electoral base of the incumbent, the
                                                          personnel                                                 federal transfers,    greater the incentives to spend on
                                                          spending per                                              nontax provincial     patronage rather than public goods.
                                                          capita, personnel                                         revenues, dummy if    Replacing the indicator of party
                                                          spending as a                                             Peronist president,   competition with a measure of the
                                                          percentage of                                             dummy if Peronist     percentage difference between the vote
                                                          total provincial                                          provincial            for winning parties and their closest
                                                          spending net of                                           governor, dummy       competitors yields similar but weaker
                                                          interest payments                                         for gubernatorial     results" (369-370)
                                                          on the provincial                                         election year, one-
                                                          debt                                                      year lag, one-year
                                                                                                                    lead


Rozevitch   Israel    Political    1966-1983, municipalit Government          "Local           dummy if mayor is    population, number    "The political dummy variable gave
and Weiss             favoritism   1971-1983 ies (26)     participation in    Authorities in   member of ruling     of cars, area of      interesting results. As hypothesized,
(1993)                                                    municipal           Israel -         party in Knesset     city, size of         politics plays a larger role in the general
                                                          spending            Financial        (national level),    cumulative deficit,   grant than in the earmarked grant. In
                                                          (grants),divided    Data"' and       dummy if mayor                             fact, this variable is always significant
                                                          between "general"   "Local           and prime minister                         for the general grant and never
                                                          and "earmarked"     Authorities in   are Maarach                                significant (and in three cases even
                                                                              Israel -         (Labor), dummy f                           negative) for the earmarked
                                                                              Physical Data"   mayor and prime                            grant...politicians do show favoritism,
                                                                                               minister are both                          however, a glance at the size of the
                                                                                               Likud                                      coefficients shows that the benefits are
                                                                                                                                          not large" (340). "the Likud seems to
                                                                                                                                          favor its members in most regressions
                                                                                                                                          while the Maarach shows far less
                                                                                                                                          tendency to do so" (342).
Schady   Peru   Political    1991-93,   provinces      spending by the   Vote share for       logged                "The coefficient on nonmarginality is
(2000)          favoritism   1993-95    (194) plus     FONCODES for      Fujimori, absolute   FONCODES              always negative and is significant at the
                                        districts of   community-based   difference of vote   poverty index,        5% level when all the controls are added;
                                        Lima           projects          share for Fujimori   logged population,    the coefficient on the measure of core
                                                                         minus 50%,           percentage rural.     support is positive and significant at the
                                                                         change in Fujimori   Also, a Fourier       5% level in every specification" (299)
                                                                         vote between 1990    series expansion of   When all controls are added, in 1991-
                                                                         and 1993             the FONCODES          1993, the highest level of spending went
                                                                                              poverty index,        to provinces where 50-60% of the
                                                                                              average age of        population voted for Fujimori (see Fig 4).
                                                                                              household head,       In 1993-95, though, "the coefficients on
                                                                                              average number of     the measures of marginal vote and core
                                                                                              years of schooling,   support are jointly insignificant at the 5%
                                                                                              occupation,           level in all the specifications" (301).
                                                                                              proportion owning     However the coefficient on change in
                                                                                              a car, a              Fujimori vote is negative and significant,
                                                                                              refrigerator, color   meaning that provinces that turned
                                                                                              TV, and black-and-    against Fujimori got more FONCODES
                                                                                              white TV,             spending.
                                                                                              proportion that is
                                                                                              indigenous, infant
                                                                                              mortality rate,
                                                                                              dummy if border
                                                                                              province,
                                                                                              department
                                                                                              dummies
Solé-Ollé   Spain   Political    1993-2003 local         per capita            capital grants   dummy if              population, land     "partisan alignment has a sizeable effect
and                 favoritism   (last two governmen     Central,              from survey on   municipality is       area, property       on the amount of grants received by
Sorribas-                        years of  t (869)       Regional, or          budget outlays   aligned with          value per capita,    municipalities. The effect is much
Navarro                          each term               Upper-local           conducted        Central, Regional,    property tax rate,   stronger when the aligned governments
(2008)                           1994-95,                grants to             yearly by        or Upper-local        debt                 are single-party governments at both
                                 1998-99,                respective            Ministry of      grantor for the       charges/revenue      layers." "in the single party case, aligned
                                 and 2002-               municipality          Economics and    respective term of                         municipalities receive over 40% more
                                 03)                     (located in a         Finance          office; cut point                          grants than unaligned ones." (28)
                                                         particular                             density, or                                "There is also a significant effect when
                                                         electoral district,                    proportion of                              the party at one or both layers is the
                                                         for a particular                       swing voters,                              leader of a coalition, and when a single
                                                         municipal term of                      measured by                                party or a party leading a coalition at the
                                                         office)                                difference in                              municipal level is also partner of a
                                                                                                absolute value                             coalition at the upper level. However,
                                                                                                between the vote                           parties which are mere partners at the
                                                                                                share of the party                         municipal level do not seem to get more
                                                                                                in government and                          grants from upper tiers of governments
                                                                                                the vote share of                          controlled by the same party. " (28)
                                                                                                the next party,
                                                                                                having either more
                                                                                                or fewer votes




Teitelbaum India    Political    1973-2006 states (16) social services vs. Indian budget        literacy and party    per capita state     "…breakdown of the traditional
and Thachil         favoritism                         economic                                 fragmentation, %      domestic product     patronage system, and the emergence of
(2010)                                                 services                                 of seats won by       (log), dummy for     lower caste voters have given rise to new
                                                       (development                             'Other Backward       election years,      demands for spending on social
                                                       expenditures)                            Classes' (OBC)        population log       services.... party fragmentation is
                                                                                                candidates, % of                           positively associated with higher levels
                                                                                                seats by three most                        of spending on social services and lower
                                                                                                dominant                                   spending on economic services that can
                                                                                                ideological centers                        be more easily targeted at specific groups
                                                                                                of Indian politics                         of voters... emergence of lower caste
                                                                                                (Indian National                           parties has given rise to greater demand
                                                                                                Congress,                                  for spending on education." (29)
                                                                                                Bharatiya Janata
                                                                                                Pary, and left
                                                                                                (communist)
                                                                                                party)
Toha       Indonesia Political    1999-2005 districts     local public        Dalam           vote margin,       GDP per capita,          Evidence that electoral incentives drives
(2009)               favoritism             (440)         goods measured      Angkas.         ruling party       area size,               distribution of public goods is weak.
                                                          by a block grant                    supporter, won by population                Highly populated district and the more a
                                                          (Dana Alokasi                       Golkar, effective                           district could generate revenues from its
                                                          Umum, or DAU                        number of parties,                          natural resources as a percentage of its
                                                          logged)                             ruling party                                GDP it tend to receive less money,
                                                                                              supporter,                                  whereas district area is correlated tend to
                                                                                              interaction terms,                          receive more transfers. In contrast,
                                                                                              separatism                                  relationship between electoral
                                                                                              measures                                    competition and size of transfer per
                                                                                                                                          capita current separatist dummy variable
                                                                                                                                          is statistically significant.


Treisman   Russia    Political          1992 administrat transfers of         newspaper       support for Yeltsin   underdevelopment      Transfers go to districts that hold
(1996)               favoritism              ive region financial             Segodnya ,      in 1991/margin of     of social             credible "bargaining power", i.e. those
                                             (89)        resources to the     Council of      vote (voting          infrastructure,       that threatened by declaring sovereignty
                                                         regions net of tax   Ministers       record), deputy in    profits per capita,   or striking. Also go to regions that voted
                                                         payments to          (using info     powerful              urban-rural mix       against Yeltsin. No evidence of money
                                                         center               from Ministry   committee,            (regional need),      going to needy districts, to districts that
                                                                              of Finance),    parliamentary         local                 rapidly reformed, or to districts with
                                                                              various         overrepresentation    agricultural/energy   influential policymakers
                                                                              statistical     , local support for   production,
                                                                              handbooks       economic reform,
                                                                              (published by   ability to raise
                                                                              Russian         taxes locally,
                                                                              Committee on    visits from central
                                                                              Statistics)     officials, having
                                                                                              permanent
                                                                                              representative in
                                                                                              Moscow, ability to
                                                                                              strike or separate
                                                                                              (declare
                                                                                              sovereignty) from
                                                                                              Russia
Wantcheko Benin     Political             2001 district (10 vote for a type of   survey            past voting          gender, age        "The empirical results show that
n                   favoritism                 experiment candidate              conducted in      behavior, types of                      clientelism works for all types of
(2003)                                         al districts,                     all 10            treatment                               candidates but particularly well for
                                               approx 180                        experimental      (clientelist, public                    regional and incumbent candidates."
                                               samples                           districts. In     policy)                                 (421) "(1) negative public policy
                                               per district)                     each district a                                           treatment effect in the North and district
                                                                                 representative                                            controlled by regional candidates, (2)
                                                                                 sample of                                                 positive public policy treatment effect in
                                                                                 voters                                                    the South, (3) positive public policy
                                                                                 interviewed in                                            treatment effect on women in the North
                                                                                 the clientelist                                           and in opposition-controlled districts
                                                                                 village, public                                           (which are mostly in the South), (4)
                                                                                 policy village,                                           positive clientelist treatment effect for all
                                                                                 and control                                               types of candidates, though not
                                                                                 villages.                                                 significant in the probit analysis for the
                                                                                                                                           regional candidates and northern
                                                                                                                                           candidates." (417-8)


Weitz-    Argentina Political    around        individuals implementation survey                   poverty level, non- population (log),   "increasing levels of opposition decrease
Shapiro             favoritism   2006          in 126      of Programa     conducted by            vote related costs, mean education,     the probability of observing clientelism
(2009)                                         cities      Nacional de     author                  size of opposition, radio stations      where poverty is low. As poverty
                                                           Seguridad                               PJ (Peronist                            increases, increases in the size of
                                                           Alimentaria                             party), Governor                        opposition are associated with an
                                                           (National Food                          party, reelection,                      increased probability of mayoral
                                                           Security                                mayor party                             intervention into the PNSA." (23) "high
                                                           Program), mayor                         dominates, turnout                      competition should be compatible with
                                                           list                                                                            clientelism when poverty is high, but
                                                                                                                                           should lessen the incentives for
                                                                                                                                           clientelism as poverty decreases." (29)
Worthingto Australia   Political    1981-82 to states (6)   (1) real per capita   Commonwealt     number of seats a      Education             "The most crucial elements of this
n and                  favoritism   1991-92                 education special     h grants        state has in           expenditure           process of purchasing political capital
Dollery                                                     purpose payments      commission      parliament, the        disability ratio,     would appear to be a state's number of
(1998)                                                      SPPs (2) real per     report on       proportion of seats    heath expenditure     federal seats, the share of those seats held
                                                            capita health         general         held by the            disability ratio,     by the federal government, and the
                                                            SPPs, (3) real per    revenue grant   Commonwealth           social security and   incidence of federal elections. Similarity
                                                            capita social         relativities    (national)             welfare               of political persuasion [state-center party
                                                            security and                          government party,      expenditure           alignment] would also appear to be
                                                            welfare SPPs                          dummy if state         disability ratio      instrumental, at least in regards to
                                                                                                  and national                                 education payments. Moreover, the
                                                                                                  governments share                            proxies for political safety perform well
                                                                                                  same ruling party,                           for education and welfare, though
                                                                                                  dummy for                                    inconsistently for health" (310)
                                                                                                  national election-
                                                                                                  year, dummy for
                                                                                                  state election-year,
                                                                                                  proportion of seats
                                                                                                  held with less than
                                                                                                  a 5% two party
                                                                                                  preferred vote
                                                                                                  ("marginal seats"
                                                                                                  or unsafe seats),
                                                                                                  proportion of
                                                                                                  federal seats
                                                                                                  decided on
                                                                                                  preferences
                                                                                                  ("marginal seats"
                                                                                                  or unsafe seats)
Yarahuán   Mexico   Political     2000-2006 municipalit per capita        Ministry of        ENP Laakso-           "FISM resources seem not to be
(2010)              favoritism              y (2436)    expenditure (fund Social             Taageperta Index      influenced by electoral competition at the
                                                        for municipal     Development        (party                municipal level, by the concurrence of
                                                        social                               competition),         party between municipal and federal or
                                                        infrastructure                       indicator for same    by the presence of municipal elections.
                                                        FISM,                                party in federal      However the estimation shows a positive
                                                        conditional cash                     and municipal         cyclical state election effect and a
                                                        transfer program                     governments for 2     positive effect of state governments on
                                                        Oportunidades,                       terms or more,        distribution at the municipal level.
                                                        Milk subsidy                         indicator for same    Oportunidades shows positive significant
                                                        program Liconsa)                     party in federal      effect on party concurrence in state and
                                                                                             and state govt,       federal government. Liconsa appears as
                                                                                             municipal or state    the program with the greatest political
                                                                                             election, party       influence in the distribution of its
                                                                                             affiliation of        resources." (22)
                                                                                             municipal
                                                                                             government


Akhmedov Russia     Political       1995-2003 region-   budgetary           Goskomstat       months to election,   Overall, strong evidence of sizable
and                 business cycles           month     expenditure,        (State           incumbent term,       opportunistic cycles in fiscal policies and
Zhuravskay                                    (6767)    budget revenue,     Committee of     dummy for             no evidence of cycles in economic
a (2004)                                                growth inflation,   Statistics),     incumbent             growth, in particular high spending occur
                                                        regional            Ministry of      governor is           a 1-2 months before elections. Shift of
                                                        budgetary wage      Finance of the   supported by the      public expenditures to what is most
                                                        arrears, wage       Russian          Communist             visible to voters (1318) "Total budget
                                                        level, and income   Federation       coalition             expenditures experience the first
                                                                                                                   significant jump up of 7 percent nine
                                                                                                                   months before elections. After that there
                                                                                                                   are no significant changes until one
                                                                                                                   month before elections when
                                                                                                                   expenditures rise significantly by 13
                                                                                                                   percent." (1316) "Magnitude of the cycle
                                                                                                                   decreases with education, urbanization,
                                                                                                                   level of democracy, transparency of the
                                                                                                                   government, and freedom of media."
                                                                                                                   (1334)
Canes-      U.S.     Political             2008 individuals % homes sold         zillow.com,      gubernatorial          real income         On average, gubernatorial election
Wrone and            business cycles            (1000 per                        Conventional     election year and      growth, change in depresses home sales by two and three-
Park (2013)                                                                      Mortgage         its interaction with   unemployment,       tenths of a pp.
                                                                                 Home Price       electoral              demographic
                                                                                 Index            competitiveness        change, lagged
                                                                                 published by                            foreclosure , years
                                                                                 the Federal                             indicators
                                                                                 Home Loan
                                                                                 Mortgage
                                                                                 Corporation
Canes-    10 OECD Political         1975-2006 country by   nongovernment OECD.Stat                polarization           interest rates,    Nongoverment investment declines in the
Wrone and countries business cycles           quarter      gross fixed                            between major          country and year   preelection period if polarization between
Park                                          (936)        capital formation                      coalitions,            fixed effects      major parties is high or the race is
(2012)                                                     (GFCF)                                 electoral                                 competitive (103). When polarization is
                                                                                                  competitiveness                           high, 2-3pp decrease of private fixed
                                                                                                                                            investment growth in the quarter before
                                                                                                                                            the election. Private fixed investment
                                                                                                                                            decline in preelection period occurs when
                                                                                                                                            parties are within 15pp of each other
                                                                                                                                            (113).
Cole       India     Political       1992-1999 districts   all credit,           "Banking         years until next   loan characteristics   Credit is targeted toward districts in
(2009)               business cycles           (412)       agriculture credit,   Statistics" by   scheduled election (loan size, number     which the majority party just won or just
                                                           nonagricultural       Reserve Bank     (1-4 yrs)          of agricultural        lost the election, only in election years.
                                                           credit                of India                            loans, interest        Loan write-offs are greatest in districts in
                                                                                                                     rates),                which winning party enjoyed the most
                                                                                                                     nonperforming loan     electoral success, only following an
                                                                                                                     variables, district    election. (248) Agriculture credit issued
                                                                                                                     characteristics        by banks is lower in off election years (2-
                                                                                                                     (population,           4 yrs prior to election) than the year prior
                                                                                                                     literacy etc)          to an election year. The difference is
                                                                                                                                            about 8 % points. (230) "Electoral
                                                                                                                                            cycles serve as an instrument for
                                                                                                                                            identifying the effect of marginal loans
                                                                                                                                            on output" (230)
Drazen and Columbia Political       1987-2000 municipalit votes                  Contraloria       vote share in       per capita GDP,         Decrease in some expenditures such as
Eslava              business cycles           ies (1100)                         General, a        previous election, time trend, social       transfers to retirees, payments to
(2005)                                                                           public agency     per capita          indicators              temporary workers, payments of debt
                                                                                 monitoring        investment                                  service. Pre-election increases in
                                                                                 public finances   spending (log), per                         development projects such as
                                                                                                   capita current                              infrastructure, power, and road
                                                                                                   spending (log), per                         construction. " Columbian voters
                                                                                                   capita government                           penalize the incumbent party for running
                                                                                                   deficit fiscal debt                         high deficits. Furthermore, high capital
                                                                                                                                               expenditures increase the share of votes
                                                                                                                                               obtained by the incumbent party relative
                                                                                                                                               to the challenger." (25)


Ferraz     Brazil    Political       1996-2003 municipalit   cost of enforcing   Sao Paulo         election dummy      municipal,              "environmental regulation is
(2006)               business cycles           ies in Sao    environmental       Environmental     for mayor, years to economic controls       implemented less stringently whenever
                                               Paulo         regulation or       Agency and        election,                                   incumbent mayors have electoral
                                               (645)         number of           election data     reelection,                                 incentives to distort such regulations in
                                                             environmental       from Tribunal     election*reelection                         order to reduce the costs of regulation to
                                                             licenses            Superior                                                      local industries."
                                                             distributed         Electoral

Gimpelson, Russia    Political       1992-8    region (78) public                Goskomstat        electoral cycle       Per capita gross      "a negative correlation between
Treisman,            business cycles                       employment            RF                variables for         regional product,     development (measured either by per
and                                                        (share of total                         gubernatorial and     the urban share of    capita gross regional product or
Monusova                                                   employment in                           governor elections,   the population, the   urbanization) and public employment.
(2000)                                                     education,                              governor affiliated   shares of the         public employment growth was faster in
                                                           health care and                         with communists       population under      ethnically-defined subunits of the
                                                           sport, social                                                 16 and over 55,       Russian state.... The evidence on
                                                           protection,                                                   ethnic minorities,    electoral effects is mixed.... Political
                                                           culture, art and                                              previous year         ideology or affiliations of the governor
                                                           science, and                                                  unemployment,         do not appear to influence his public
                                                           public                                                        change in             employment strategy." (13-15)
                                                           administration)                                               unemployment,
                                                                                                                         previous
                                                                                                                         year/change in
                                                                                                                         federal transfers
                                                                                                                         and loans, Year
                                                                                                                         dummies
Golden and India    Political       2000-2009 districts in electricity use      Uttar Pradesh    state assembly     division fixed         Line loss higher in election years than
Min                 business cycles           Uttar                             Power            election year,     effects, H야            non-election years in UP. No evidence
(2012)                                        Pradesh                           Corporation      candidates'                               that line loss is greater in districts with
                                              (1661)                            Ltd.             criminal record,                          criminal candidates.
                                                                                                 voter turnout, win
                                                                                                 margin of last
                                                                                                 election
Khemani    India    Political       1960-1992 major       total spending on Reserve Bank         electoral cycle                           "we find that politicians manipulate
(2004a)             business cycles           states (14) public services   of India                                                       fiscal policies before elections to provide
                                                                            Bulletin                                                       targeted favors to specific interest
                                                                                                                                           groups, possibly in exchange for
                                                                                                                                           campaign support. In election years in
                                                                                                                                           the Indian states, there is evidence of
                                                                                                                                           small manipulations of fiscal instruments
                                                                                                                                           to target benefits to narrow interest
                                                                                                                                           groups, such that there is no net effect on
                                                                                                                                           the overall fiscal deficit." (151-2)


Kneebone   Canada   Political       1966-1997 province      change in           CANSIM           change in            change in            "One of our key results is the finding of a
and                 business cycles           (290)         provincial          (from matrices   provincial           unemployment         clear electoral cycle in provincial
McKenzie                                                    expenditure (as a   2782 through     expenditure          rate, change in      revenues. In election years, the
(2001)                                                      fraction of GDP)    to 2791)         lagged, dummy for    Canada-US            predilection of provincial governments of
                                                                                                 year with            nominal exchange     all political stripes to increase taxes is
                                                                                                 provincial           rate, change in      temporarily halted…. Opportunistic
                                                                                                 election, dummy      short-term real      responses with respect to provincial
                                                                                                 for Liberal Party,   interest rate,       spending are equally widespread but the
                                                                                                 New Democratic       change in ratio of   direction of the response is less
                                                                                                 Party, Progressive   debt to GDP over     homogenous across political parties. The
                                                                                                 Conservative Party   previous 2 fiscal    prevailing tendency is to decrease
                                                                                                 and Social Credit    years                spending in Health, Social Services and
                                                                                                 Party                                     Industrial Development in election years
                                                                                                                                           vis-a-vis non-election years, while
                                                                                                                                           spending in Education, transportation,
                                                                                                                                           and Communication, and Recreation and
                                                                                                                                           Culture increases in election years versus
                                                                                                                                           non-election years." (771)
Kwon     South   Political       1988-97   Election       Election timing: National        Election timing:       Election timing:      "Korean government boosted government
(2005)   Korea   business cycles           timing:        government         Statistical   pre-election period    GDP (log),            expenditure in accordance with the ruling
                                           quarters       expenditure (as    Office        distribution of        unemployment,         party's electoral concerns. While the
                                           (40)           percentage of                    national subsidies:    quarter dummy         results of the analysis clearly show
                                           distribution   GDP, log)                        "Pro-votes" -- vote    distribution of       substantively significant effects of pre-
                                           of national    distribution of                  share of ruling        national subsidies:   electoral manipulation of government
                                           subsidies:     national subsidies               party candidate in     total revenue minus   expenditure, the consequences of election-
                                           provinces                                       each province,         national subsidies,   induced public expenditure cycles may
                                           (13)                                            difference between     gross regional        be even greater." (333) "The results of
                                                                                           the proportion of      domestic product      the electoral impact show that Cox-
                                                                                           votes received by      (grdp), number of     McCubbins' prediction of resource
                                                                                           the leading two        farm households,      distribution to core supporter groups did
                                                                                           candidates (both       population (logged)   not fare well with the Korean data. The
                                                                                           for the presidential                         coefficient for pro-ruling party vote
                                                                                           election due to                              shares did not reach statistical
                                                                                           data limitations)                            significance. In contrast, the Lindbeck-
                                                                                                                                        Weibull/Dixit-Londregan prediction is
                                                                                                                                        consistent with the empirical pattern of
                                                                                                                                        the allocation of national subsidies in
                                                                                                                                        Korea. The point estimates of electoral
                                                                                                                                        margin shows a negative sign with
                                                                                                                                        statistical significance at the p<0.05
                                                                                                                                        level" (337-338)
Limosani   Italy       Political       1969-1994 municipalit   public spending ISAT database   revenues deriving                       "in pre-election periods, national
and                    business cycles           ies           in goods and                    from grants by the                      representative try to obtain investment
Navarra                                          (75=3regio    services and                    central                                 expenditures for specific projects to be
(2001)                                           ns*25years    public spending                 government,                             spent in their electoral districts with the
                                                 )             in capital goods                revenues from                           help of "friendly" local governments in
                                                               and investment in               creating a loan,                        order increase their popular consent."
                                                               3 Italian regions               binary variable for                     (323)"This result shows the existence of
                                                               (Lombardia,                     elections dates                         a positive dependent relationship
                                                               Campania,                       assuming the                            between the amount of grants from the
                                                               Sicillia)                       value of 1 in case                      central government and both
                                                                                               of election                             consumption and investment spending
                                                                                                                                       and reveals the scarce autonomy of a
                                                                                                                                       financially assisted local institutions…
                                                                                                                                       heterogeneity in regional backgrounds
                                                                                                                                       does not play a crucial role in influencing
                                                                                                                                       local public spending patterns. Loans
                                                                                                                                       seem to influence only expenditures in
                                                                                                                                       goods and services and do not have any
                                                                                                                                       impact on investments." (321-2)



Persico,   Mexico      Political       1944-2001 municipalit annual change in National Water state governors'        time-varying      On average across municipalities,
Rodríguez-             business cycles           ies (463)   number of        Commission     election                demographics      economically and statistically significant
Pueblita,                                                    potable water    (CNA reports)                                            cycle in spending associated with state
and                                                          (inlet) and                                                               governor elections -- the years in which a
Silverman                                                    sewage (outlet)                                                                     s
                                                                                                                                       governor’ election is held, the change in
(2007)                                                       connections to                                                            water inlets per capita is, on ,average
                                                             public networks                                                           0.0072 higher. Political budget cycle in
                                                             (per capita)                                                              this form of spending appears only in
                                                             (also has formal                                                          those states where the incumbent
                                                             model on                                                                  governor is from the PRI. In PRI-
                                                             factions)                                                                 governed state the election is associated
                                                                                                                                       with an additional 0.0077 inlets per
                                                                                                                                       capita. (39)


Shi and    85          Political       1975-1995 country-    govt. fiscal      Database on    election dummy         GDP per capita,   Fiscal deficit increases by 0.9 percent of
Svensson   countries   business cycles           year (1204) budget balance as Political                             GDP growth rate   GDP in election years. Political budget
(2006)                                                       a share of GDP    Institutions                                            cycles are large at 1.3 pp in developing
                                                                               from the World                                          countries but almost nonexistent in
                                                                               Bank                                                    developed countries. (1372-1373)
Alperovich Israel   Political   1976 and   localities   general            "Local           percent of          population, ratio of   "As expected, the coefficient of POL is
(1984)              returns     1978       (52)         government         Authorities in   residents who       the number of          positive, highly significant, and
                                                        participation in   Israel -         voted for parties   people 65 years old    extremely similar for both governments.
                                                        budgets of local   Financial        which formed the    or above and 14        A significant coefficient confirms the
                                                        authorities        Data"' and       government          years old or below     basic hypothesis that in forming their
                                                                           "Local                               to the number of       policy towards local authorities,
                                                                           Authorities in                       people aged 19 to      governments are responding, among
                                                                           Israel -                             65, per capita         other factors, to the political orientation
                                                                           Physical Data"                       annual deficits        of the population...governments in Israel
                                                                                                                                       acted on the assumption that better
                                                                                                                                       results can be achieved by pursuing
                                                                                                                                       policies which reward their traditional
                                                                                                                                       supporters" (293) (After entering a
                                                                                                                                       squared term for vote share) "per-capita
                                                                                                                                       grants in 1976 increased as a function of
                                                                                                                                       POL up to 61% of government
                                                                                                                                       supporters and decreased thereafter...for
                                                                                                                                       1978....55% [was the cutoff]" (294)



Arriola    Africa   Political   1960-99    40 African time to an         Africa South       cabinet size,       GDP, oil exporter,     "Recruiting more ministers into the
(2009)              returns                states     extraconstitutiona of the Sahara      polity, executive   mineral exporter,      cabinet is an effective coup-inhibiting
                                                      l change in the                       years in power,     ethnic                 strategy for the leader who fears being
                                                      chief executive                       civil war           fractionalization,     overthrown. Each additional cabinet
                                                      (regime duration)                                         French colony,         appointment lowers coup hazard by 23%
                                                                                                                Urbanization, trade    to 25%, with all else being equal,
                                                                                                                (%GDP),                depending on the regime variable used in
                                                                                                                population(log)        the model." (17)
Besley,    India    Political         2002 households holders of BPL     survey           political             village                In targeting public resources politicians
Pande, and          returns                (approx     (Below Poverty    conducted by     opportunism           characteristics such   exhibit group preferences and are
Rao                                        5400) in    Line) cards       authors in                             as village literacy    opportunistic, and individual and village
(2005)                                     259                           2002                                   rate 1991 and          characteristics mediate the extent of
                                           villages in                                                          whether village had    opportunism. (28) "Selection is
                                           4 southern                                                           at least one Gram      manifested in more educated politicians
                                           states                                                               Sabha meeting in       being less opportunistic. Incentives are
                                                                                                                the last year,         shaped by village meetings in which
                                                                                                                individual             villagers ratify beneficiary lists chosen
                                                                                                                characteristics such   by politicians" (25)
                                                                                                                as education, a
                                                                                                                household's
                                                                                                                political history,
                                                                                                                SC/ST household


Bruhn      Mexico   Political   1988-1991 states (31) amount of          Tercer Informe   PRI vote 1988, left   "marginalization       "Only the left vote reaches at least .1
(1996)              returns                           PRONASOL           de Gobierno,     vote 1988 (PRI        index" which is a      significance in models of spending
                                                      spending per       1991: Anexo      opponent), PAN        government             reorientation. Curiously, PRI vote
                                                      capita,                             vote 1988             measure and            reaches .1 significance in explaining
                                                      "reorientation" of                                        includes poverty,      spending per capita" (160)
                                                      funds (change                                             living standards,
                                                      from 1988 levels)                                         and access to
                                                                                                                services; GDP per
                                                                                                                capita, change in
                                                                                                                GDP per capita
Calvo and   Argentina Political   1987, 1990, provinces   share of           Ministry of the   Peronist vote     logged median           "Both positive and significant effects of
Murillo               returns     1995, 2000 (24) and     expenditures       Economy           share, UCR vote   income, over-           the Peronist vote on both the amount of
(2004)                                        Buenos      financed by the                      share (both       representation          federally financed expenditures and on
                                              Aires       federal                              logged)           (share of legislators   revenue sharing…By contrast, the UCR-
                                                          government,                                            over share of           Alianza has no significant effect on
                                                          relative share of                                      population), logged     either federal financing or the relative
                                                          fiscal resources                                       population share        revenue-sharing ratio" (749)
                                                          (revenue sharing                                                               "patronage can serve as a distributive
                                                          ratio) received by                                                             mechanism to benefit partisan
                                                          each province                                                                  constituencies with higher electoral
                                                                                                                                         returns for political parties whose
                                                                                                                                         constituencies are more dependent on
                                                                                                                                         public largesse (demand side). We also
                                                                                                                                         show how institutionally defined
                                                                                                                                         territorial biases can affect access to
                                                                                                                                         patronage if overlapping with geographic
                                                                                                                                         concentration of electoral support for
                                                                                                                                         political parties (supply side)." (754)
Chhibber   India   Political   1967-84    states (15) loan             Reserve Bank   the "political         revenue accruing to    "The coefficient for the political variable
(1995)             returns                            disbursements    of India       situation of the       the center (national   was positive and significant, suggesting
                                                      from centre to   Bulletin       central party in a     revenues), per         that the central government used political
                                                      state, food                     state", determined     capita income          considerations to disburse resources to
                                                      disbursements                   by three criteria:                            the states" (85) Also, "political factors
                                                      from centre to                  "first, whether a                             have a positive and significant influence
                                                      state                           state faced either a                          on the distribution of food from the
                                                                                      national or state                             centre" (87)
                                                                                      election; second,
                                                                                      whether the state
                                                                                      government and
                                                                                      national
                                                                                      government were
                                                                                      controlled by the
                                                                                      same party prior to
                                                                                      the election; third,
                                                                                      whether there was
                                                                                      serious intra-party
                                                                                      factional conflict
                                                                                      or not." (84)




Chong, De Mexico   Political         2009 electoral   turnout          field          treatment,         demographics               "Our results so far indicate a robust and
La O,              returns                precincts                    experiment     interventions with                            negative effect of dissemination of
Karlan, and                               (1298)                       conducted by   information on                                information about extensive corruption
Wantchekro                                                             authors in     Fondo para la                                 on voter turnout. The drop in
n                                                                      2009 in        Infraestructura                               participation translates into a decrease in
(2010)                                                                 Jalisco,       Social Municipal                              the incumbent's vote share, although this
                                                                       Morelos,       (FISM)                                        last result is not statistically significant
                                                                       Tabasco                                                      in all specifications.... Precincts that
                                                                                                                                    were informed that the municipality
                                                                                                                                    exercised all the transfer turned out at
                                                                                                                                    higher rates than precincts without this
                                                                                                                                    information." (19)
de          Brazil   Political   2008-10    individuals number of          survey           (experiment                               "Bolsa Familia has had little or not
Figueiredo           returns                (1000)      politicians from   conducted by     embedded in                               effects on political attitudes beyond
and Hidalgo                                             the PT or PSDB     authors in       survey with all                           generating support for President Lula for
(2009)                                                  that received a    2008 after       voters, non-Bolsa                         voters in survey sample." (19)
                                                        positive           October          voters group,                             "Consistent with our matching estimates,
                                                        evaluation         Municipal        Bolsa voters)                             Bolsa Familia recipients do not appear to
                                                        (feeling           elections        receipt of PT                             attach any positive affection to the PT
                                                        thermometer)                        message                                   party label. However, the partisan
                                                                                            treatment, PSDB                           message is substantial." For PSDB
                                                                                            message                                   politicians, partisan cues are not
                                                                                            treatment, cues                           significant. (19)
                                                                                            treatment
De La O     Mexico   Political   1994 and   electoral    voting turnout    evaluation       share of people      population 1994      "I show that average turnout and vote
(2008)               returns     2000       seccion      (difference in    survey,          treated early by     logged, population   shares in the 1994 election do not predict
                                            (precinct)   turnout 2000-     ENCASEH          Education, Health change                  enrollment in the randomized early
                                            (355)        1994)             1997 ran by      and Nutrition                             treatment but the latter does not influence
                                                                           Progresa's       program                                   future electoral outcomes. The estimates
                                                                           implementers,    (Progresa)                                suggest that a section fully treated twenty-
                                                                           Federal          measured by                               one months before election time
                                                                           Electoral        multiplying a                             increased its turnout in 2000 by five
                                                                           Institute,       variable that is one                      points and its incumbent vote share by
                                                                           Census of        when the precinct                         four points. For an average section with
                                                                           Population and   has localities that                       578 potential voters the estimations
                                                                           Housing          were part of the                          imply a change in in the probability of
                                                                           (2000) and       early treatment,                          turnout from 0.60 in 1994 to 0.65 in
                                                                           Partial          and share of                              2000; and a change in the probability of
                                                                           Cenusus          people living in                          voting for the incumbent from 0.38 to
                                                                           (1995)           localities that were                      0.42. I also show that the conditional
                                                                           produced by      randomized with                           effect of Progresa on the strength of the
                                                                           INEGI (used      respect to the                            PRI in 1994 is smallest among those
                                                                           for              precinct                                  precincts where the PRI had more than
                                                                           experimental     population; vote                          eighty-five percent of the votes in 1994
                                                                           setup)           share by political                        compared to precincts where the PRI
                                                                                            party                                     strength was moderate or low." (43)
Denemark   Australia   Political   1990-1993 constituenc total sports grants Australian      preelection          unemployment,          "…marginal electoral divisions in both
(2000)                 returns               ies         dollars             auditor-general marginality          household income       1990 and 1993 can be seen as receiving
                                                                                             (measured by                                significantly more funds than safer seats"
                                                                                             absolute value of 2                         (905) yet districts with "very safe"
                                                                                             party preferred                             electoral margins (17.5-25%) also
                                                                                             swing required to                           received larger grants (903)
                                                                                             lose the seat), seat
                                                                                             controlled by
                                                                                             ALP, member of
                                                                                             the preelection
                                                                                             Labor Cabinet
Dunning    India       Political   2007-8      individual received a benefit   survey          Scheduled Caste political knowledge "while there is evidence that reservation
(2010)                 returns                 (968)      or job from the      conducted by    or Scheduled        and engagement  increases the propensity of member of
                                                          village council in   authors in      Tribe participation                 the target groups to say that the council
                                                          previous 1-2         2009                                                prioritizes them, reservations does not
                                                          years                                                                    strongly boost perceptions that the
                                                                                                                                   council effectively serves SC and ST
                                                                                                                                   communities." (20) "we fail to reject the
                                                                                                                                   null hypothesis of no effect for most of
                                                                                                                                   the many variables that tap distributive
                                                                                                                                   or policy outcomes... In Karnataka,
                                                                                                                                   reservation of village council
                                                                                                                                   presidencies does not appear to have
                                                                                                                                   strong distributive or policy effects." (40)



Evans      Canada      Political   1958-2000 electoral     share of vote for   Elections       federal spending     avg percent change   "Approximately 8% of all majority party
(2006)                 returns               district      the majority        Canada,         per capita in an     in the GDP per       candidates win or lose their electoral
                                             (990)         candidate, or       Parliament of   electoral district   capita during the    district races by less than 3% of the vote,
                                                           opposition          Canada's        two years prior to   previous electoral   and 26% of the time the margin of
                                                           incumbent           website         a federal election   cycle; average       victory or defeat is less than 10% of the
                                                           candidate                                                change in the        vote. A $100 increase in per capita
                                                                                                                    provincial GDP per   spending in an electoral district will
                                                                                                                    capita during the    increase the share of the vote for the
                                                                                                                    previous electoral   majority party by approximately 2% of
                                                                                                                    cycle, electoral     the total vote cast and at best leave
                                                                                                                    district dummy       unchanged the share of the vote for the
                                                                                                                    variable, year       incumbent opposition party candidate"
                                                                                                                    dummy                (246)
Hirano   Japan   Political   1977-1992       district-     per capita central-   Nikkei NEEDS     legislators who          1st Tier workers/    legislators who pass away in office:
(2006)           returns     (includes 5     level         to-local              database and     pass away in             workers (log),       "There is no statistically significant
                             elections for   (1760) and    government            various issues   office : effect of       population density   evidence that losing any type of Diet
                             Lower           municipalit   transfers             of the           individual               (log), per capita    member affects the overall amount of
                             House of        y-level                             shichosonbets    representative           income               transfers directed to particular electoral
                             the             (49895)                             u kessan jyoko   measured by death                             districts." (15) "Contrary to popular
                             Japanese                                            shirabe          of a legislator,                              perception, there is no evidence that
                             Diet)                                                                whether deceased                              senior LDP incumbents had a
                                                                                                  incumbent was                                 statistically significant effect on the
                                                                                                  elected for several                           distribution of any type of transfer." (16)
                                                                                                  terms, whether                                close elections : "victory of an LDP
                                                                                                  deceased                                      incumbent affects the distribution of
                                                                                                  incumbent did                                 central government transfers to
                                                                                                  poorly in the                                 municipalities which are part of the
                                                                                                  election to                                   incumbent's core electoral support in the
                                                                                                  legislative session,                          first year following the election." (18)
                                                                                                  whether LDP                                   "individual LDP representatives are not a
                                                                                                  representative                                free-flowing "pipeline of pork".... There
                                                                                                  passes away                                   is no evidence that representatives, even
                                                                                                  close elections :                             senior ones, affect the distribution of
                                                                                                  indicator for                                 central government transfers across
                                                                                                  whether LDP                                   districts. Electoral margins, rather than
                                                                                                  candidate won the                             seniority, is the main characteristic
                                                                                                  last seat in district,                        determining which Diet representative
                                                                                                  difference in                                 affects year-to-year central government
                                                                                                  government                                    transfers to Japan. Only marginal
                                                                                                  transfers received                            representatives - the ones with the strong
                                                                                                  from being part of                            incentives and perhaps the ones the LDP
                                                                                                  LDP candidate's                               would like to help most - appear to have
Horiuchi   Japan     Political   FY 1994 - municipalit per capita amount       account           number of             municipality fiscal    "number of ruling-coalition members
(2007)               returns     FY 1995    ies (3342) of total transfers      settlement        incumbents who        strength index,        (divided by the district magnitude has a
                                 (August               (general and            (Chiho Zaisei     belonged to the       municipality           negative (-0.030) and highly significant
                                 1993-March            specific                Chosa Kenkyu      ruling coalition      population (log),      effect in the first-difference regression….
                                 1996)                 subsidies) from         Kai, various      during each           municipality           Coalition leaders allocate, ceteris
                                                       the central to          issues).          budget cycle,         population density     paribus, disproportionately larger
                                                       municipal               Municipality      divided by the        (log), taxable         amounts to districts with more opposition-
                                                       governments (in         population at     district magnitude.   income per capita      party members to secure smooth
                                                       log)                    the end of        Dummy one if          (log), ratio of        operation of their rule." (21-22)
                                                                               previous fiscal   incumbent is a        population aged        "...districts with powerful members of the
                                                                               year from         member of             below 14 and of        ruling coalition (that is, districts with
                                                                               Kokudo Chiri      conferences for       population over 65,    Cabinet ministers, party leaders, and
                                                                               Kyokai            coalition leaders;    ratio of number of     those who attended the Budget
                                                                               (various          number of             persons employed       Committee frequently) tend to receive the
                                                                               issues)           attendances by        in the agricultural    larger amount of per capita transfers, but
                                                                                                 ruling-party          sector against total   these effects are statistically
                                                                                                 members at the        number of              insignificant." "The number of seats per
                                                                                                 Budget                employed persons,      capita, a measure of malapportionment of
                                                                                                 Committees.           ratio of the           seats in Japan’s Lower House, is positive
                                                                                                 number of seats       population in          and significant in all regression
                                                                                                 divided by the        densely inhabited      estimates."(22)
                                                                                                 district population   districts
                                                                                                 (logged)


John and   England   Political   1981/2-     larger        grant related       former             unitary authority,   population,            "We find that there is both targeting of
Ward                 returns     1995/6      English       expenditure         Department for    Westminster           measures of            extra resources to marginal seats and the
(2001)                                       local         assessment and      the               election dummy,       average individual     use of a scattergun approach that spreads
                                             authorities   standard spending   Environment,      pre-Westminster       social need (factory   benefit widely, especially when the
                                             (107)         assessment          Transport, and    election dummy,       analysis of census     government sees its re-elect under
                                                           (GREA/SSA)          the Regions       interaction lead      information)           threat." Targeting more important in the
                                                                                                 and Westminster                              late 1980s. (331)
                                                                                                 election dummy,
                                                                                                 county council
                                                                                                 election dummy,
                                                                                                 London borough
                                                                                                 election dummy,
                                                                                                 control of local
                                                                                                 authority, Tory
                                                                                                 flagship, hung
                                                                                                 local dummy
Keefer and India    Political   1999-2003 constituenc actual spending     Indian         party stronghold      political, caste,      "party stronghold variable is a significant
Khemani             returns               ies (483)   incurred by the     constituency   (1 if party won       religion               determinant of variation in
(2009)                                                MP as a             development    every election in     fragmentation,         spending….MPLADS disbursements in
                                                      percentage of       fund called    the 90s), candidate   constituency           constituencies with dominant party are
                                                      what the MP was     Member of      stronghold, margin    population density,    10 % points lower, more than one-half of
                                                      entitled to spend   Parliament     victory, 1993-99      proportion of          the standard deviation in spending in the
                                                      on public works     Local Area     spending, reserved    villages with          data. " (105) "Party stronghold effects
                                                      in his or her       Development    indicator for         various public         are also robust to the control for a
                                                      constituency        Scheme         whether a             goods                  constituency's average margin of
                                                                          (MPLADS)       constituency is                              victory." (107)
                                                                                         reserved for
                                                                                         candidates
                                                                                         belonging to the
                                                                                         scheduled castes
                                                                                         and tribes,
                                                                                         candidate
                                                                                         switched




Knoesen    South    Political         2000 municipalit amount of                         vote share of                                Electoral support for ANC in 2000
(2009)     Africa   returns                ies         transfer                          African National                             elections may help explain variation in
                                                       (regression                       Congress (ANC)                               ANC's stated objective of poverty
                                                       analysis in                                                                    reduction and its actual allocation.
                                                       progress)
Kramon     Kenya    Political         2002 individual whether             Afrobarometer whether individual     individual             Estimates suggest that a vote-buying
(2009)              returns                (1278)     individual voted    Round 3       was approached         characteristics such   attempt increases the probability of
                                                      or not              survey        by a political party   as education, cash     voting by about 10 %-tage points.One's
                                                                                        representative and     income, gender,        assessment of the intrinsic value of
                                                                                        been offered a         urban, age             democracy plays a role in determining an
                                                                                        bribe or gift in                              individual's decision to vote.(11) Those
                                                                                        exchange for a                                approached by party are about 14
                                                                                        vote, vote margin,                            percentage points more likely to vote
                                                                                        democratic belief,                            than those who do not. In the Kenyan
                                                                                        political efficacy                            case, vote buying view potentially swing
                                                                                                                                      voters as acceptable targets. Also, those
                                                                                                                                      with low level of education is likely to be
                                                                                                                                      most influenced. (25)
Leigh      Australia   Political   2001-2004 seats (150) total funding per Department of National party,               population density, Distribution of funding under 4 programs
(2008)                 returns                           electorate ($m)   Transport and Liberal party,                quadratic in        strongly skewed towards electorates that
                                                                           Regional      income                        income              were held by the Coalition government in
                                                                           Services and                                                    2001. Compared with non-government
                                                                           Australian                                                      seats, seats held by National Party or
                                                                           Local                                                           Liberal Party received more.(295)
                                                                           Government                                                      "Compared with non-government seats,
                                                                           Association                                                     National Party electorates tend to have an
                                                                           2004                                                            additional 5.2 Stronger Families &
                                                                                                                                           Communities grants, and an additional
                                                                                                                                           3.4 Regional Partnership grants. For
                                                                                                                                           Liberal Party electorates, figures are
                                                                                                                                           additional 3.4 and 1.7.


Letha      India       Political   1980-2008 states (24) incumbent vote Election                   Fiscal policy       literacy,             "Our results indicate that voters in the
Kannan                 returns                           share obtained by Commission              variables           urbanization,         Indian states seem to be fiscal
(2009)                                                   the party of the                          including various   population,           conservatives as when they do react to
                                                         incumbent chief                           taxes, non-tax      schedule class        fiscal policy; they punish the incumbents
                                                         minister at the                           revenues and        population and real   for increases in expenditures or
                                                         state in sub-                             expenditures of     SD, vote share        taxes....Voters in general do not seem to
                                                         national elections                        state governments   obtained by           respond to government revenue
                                                                                                                       incumbent in          expenditures." (87)
                                                                                                                       previous election
Lindberg   Ghana       Political   1996 and    individual    voting rationale     survey                               demographics          "Our survey results suggest that voting
and                    returns     2000        (690          (reason for voting   conducted by                                               behavior is not so much ethnic in essence
Morrison                           elections   interviewee   for a political      authors from                                               as it is a rational response to the classic
(2008)                                         s from 6      representative       June and July,                                             information problem for the voter using
                                               constituenc   asked in an open-    2003                                                       the parties' historical affiliations as a key
                                               ies)          ended question in                                                               source of evaluation."
                                                             survey; non-                                                                    "clientelistic and ethnically predisposed
                                                             regression                                                                      voting is a minor feature of Ghanaian
                                                             analysis)                                                                       elections and that when present, it seems
                                                                                                                                             to be nurtured by intense partisan
                                                                                                                                             competition." (121-122)
Manacorda, Uruguay   Political   2005-2008 households political support     two surveys      PANES (Plan de age, gender, years      Beneficiaries of a large government anti-
Miguel, and          returns               (approx.   for FA (Frente        conducted by     Atencion Nacional of education         poverty program, PANES, showed 11-
Vigorito                                   3000)      Amplio)               authors in       a la Emergencia                        14% increase in support for the current
(2010)                                                government            collaboration    Social, anti-                          government than non-beneficiaries. (13)
                                                                            with Veronica    poverty program)                       "These effects last into the post-program
                                                                            Amarante and     program effects                        period." (27)
                                                                            Economics and
                                                                            Sociology
                                                                            Departments at
                                                                            the Univ. of
                                                                            the Republic,
                                                                            Ministry of
                                                                            Social
                                                                            Development
                                                                            staff
                                                                            (conducted in
                                                                            2006-2007 and
                                                                            2008)


Nakayama   France,   Political   1950s-                 credit allocation   archives (non- power balance       one-party            "In Japan, as well as Italy, the MPs of the
(2007)     Italy,    returns     1970s                  patterns from       regression     between the         dominance over       dominant parties seized every chance to
           Japan                                        central to local    analysis)      bureaucracy and     long period in all   get more credit for their constituencies'
                                                        government                         the parties, size   three countries      investment programs, and all these
                                                                                           and viability of    used as a control    clientelist practices helped them develop
                                                                                           local governments                        and consolidate their local electoral
                                                                                                                                    bases. In contrast the [French] Gaullist
                                                                                                                                    regime was based on the exclusion of any
                                                                                                                                    kind of pork barrels, including those for
                                                                                                                                    the Gaullist party's own MPs." (3)
                                                                                                                                    "Gaullists' hegemony essentially relied
                                                                                                                                    on the capacity of the bureaucracy to
                                                                                                                                    neutralize them and integrate local
                                                                                                                                    political forces irrespective of their party
                                                                                                                                    labels.... to perpetuate Gaullist
                                                                                                                                    dominance, the government needed to
                                                                                                                                    keep allocation patterns from their own
                                                                                                                                    Majority MPs demanding pork barrels --
                                                                                                                                    but failed to do this in the late 1960s and
                                                                                                                                    1970s" (35)
Ortega and Venezuela Political    2000 and    municipalit total number of National       total expenditure     Human                  "Mercal does not appear to have had
Penfold-             returns      2004        y (331)     "NO" votes in    Electoral     on "Mision Ribas"     development index,     significantly different electoral effects
Becerra                                                   2004 referendum, Council       considered an         population,            across municipalities of different
(2008)                                                    change in Chavez               excludable good       unemployment           political leanings." (21) "the yields for
                                                          votes between                  measuring             rate, fraction of      electoral returns for excludable goods
                                                          2000 and 2004                  distribution of       dwellings without      under Chavez social programs suggest
                                                          elections                      scholarships in the   potable water,         that [the relationship between the yields
                                                                                         form of cash          fraction of            for excludable goods and political
                                                                                         transfers to          dwellings classified   competition] can take a "U" form. The
                                                                                         individuals at the    to be in "poor         electoral return for "Mision Ribas" was
                                                                                         local level, total    condition", fraction   higher than "Mercal" in hyper-
                                                                                         expenditure on        of the adult           competitive arenas, that is, in districts in
                                                                                         "Mision Mercal"       population with a      which both the mayor and the governor
                                                                                         considered public     high school degree     were controlled by opposition forces....
                                                                                         good because any      or more, average       under hyper-competitive conditions loyal
                                                                                         citizen had access    number of persons      voters are more easily identifiable
                                                                                         to these discount     per dwelling,          because probably some sort of self-
                                                                                         stores to purchase    fraction of the        selection process applies." (25)
                                                                                         subsidized goods,     population aged 14     "Regarding the electoral returns of non-
                                                                                         total expenditure     or younger,            excludable goods,...statistical evidence
                                                                                         on "Mision            fraction of the        from Chavez's "Misiones" programs does
                                                                                         Identidad" to         labor force            not suggest that these types of goods
                                                                                         study how process     employed in            have higher electoral yields under more
                                                                                         of                    communications         competitive political conditions." (25-26)
                                                                                         enfranchisement       and fraction of the
                                                                                         (obtaining an ID      labor force
                                                                                         card) interacted      employed in
                                                                                         with the              agriculture
                                                                                         distribution of
Rao and    India      Political   1983-1993 major       statutory transfers National     share of a ruling     state domestic         "some evidence for the importance of
Singh                 returns               states (14) (shared taxes plus Institute for party or ruling       product, per capita    variables that may proxy bargaining
(2001)                                                  non-plan grants), Public Finance coalition's           state domestic         power" of Indian states, such as the
                                                        grants for state    and Policy,  Members of            product, population    positive estimate of "lagged effect of a
                                                        plan schemes,       New Delhi    Parliament who                               match between the state and central
                                                        discretionary                    come from a                                  ruling parties on grants for state plan
                                                        transfers (grants                particular state,                            schemes" and in the loglinear
                                                        for central plan                 dummy if party                               specification, "the positive effect of the
                                                        schemes and                      controlling the                              proportion of ruling party/coalition MPs
                                                        centrally                        center is same as                            on per capita statutory transfers, again
                                                        sponsored                        the party                                    with a lag"
                                                        schemes), total                  controlling the
                                                        transfers (sum)                  state (both lagged
                                                                                         three years)
Samuels   Brazil   Political   1994 incumbents whether who ran      pork (% of all         Money affects vote totals, but pork
(2002)             returns          (244)      for reelection won   pork-barrel            barreling has no direct effect on votes.
                                               or not in 1994       amendments             (857) "Deputies thus do not deliver pork
                                                                    released in his or     because it provides a direct electoral
                                                                    her district),         payoff, they do so in order to gain the
                                                                    dominance (% of        financial support of powerful economic
                                                                    all candidate votes    interests. The money helps win votes."
                                                                    that deputy            (860)
                                                                    received where he
                                                                    received the
                                                                    largest share of his
                                                                    own votes in
                                                                    previous election),
                                                                    pork*dominance,
                                                                    campaign
                                                                    spending (also
                                                                    squared), rank in
                                                                    previous election,
                                                                    # of terms, leader,
                                                                    switched parties,
                                                                    change in % of
                                                                    votes that deputy's
                                                                    list won in his
                                                                    state from one
                                                                    election to next,
                                                                    leader, vote share
                                                                    of 1990, turnout,
                                                                    party dummies
Szwarcberg Argentina Political   1995-2005 municipalit candidate’s        political             monitor clients,                        "Candidates who employ viewers to
(2008)               returns               ies (8)     position on the    parties’ closed-      gender, and the                         monitor client turnout are likely to run in
                                                       party ballot,      ballots               probability of                          a position 7 times higher than those that
                                                       broker's electoral                       getting elected.                        do not employ viewers. Also, candidates
                                                       performance                              Also qualitative                        that combine viewers’ monitoring with
                                                                                                results on                              their own surveillance practices are prone
                                                                                                programmatic                            to run in a position two times higher than
                                                                                                mobilization,                           those that do not employ this tactic". (34-
                                                                                                Charismatic-                            35) "Brokers that recur to authoritarian
                                                                                                linkages,                               practices have better chances to become
                                                                                                Retrospective                           elected than those who only rely on
                                                                                                support, vote                           democratic
                                                                                                buying with minor                       strategies of mobilization." (39)
                                                                                                consumption
                                                                                                goods, clientelistic
                                                                                                mobilization
                                                                                                (public goods),
                                                                                                threatening,
                                                                                                monitoring,
                                                                                                punishment




Szwarcberg Argentina Political          2005 local          indicator for     semi-closed       unpaid party       municipality         "The negative and significant coefficients
(2010)               returns                 elected        whether           and in-depth      activist (bottom- population, gender,   on unpaid party activist, outsiders, and
                                             representati   candidates took   interviews,       up), outsider(top- Buesnos Aires        minority party show that these
                                             ves (144)      attendance in     participant       down), Union                            independent variables predict
                                                            rallies           observations at   Civica Radical,                         monitoring. Likewise, the positive and
                                                                              political         minority parties,                       significant coefficient on size shows that
                                                                              meetings,         National and                            candidates competing in bigger
                                                                              rallies and on    provincial                              municipalities are more prone to monitor
                                                                              Election Day      incumbent                               voters than candidates competing in
                                                                              by author                                                 smaller districts. " (20)
Tandon    India    Political   1991-2004 regions          vote share of        Election         vote change          employment,            Robust relationship between the regional
(2007)             returns               (472)            parliamentary        Commission       arising from         population             exposure to tariff decreases and Left
                                                          members (Left,                        reform, wage                                gains. Results driven by BJP and far left
                                                          Con, Far Left,                        change from                                 rather than Congress and unaffiliated
                                                          BJP, Reg)                             reform, regional                            regional parties. "The results verify that
                                                                                                exposure to tariff,                         voters hurt by reform changed their
                                                                                                industry opened                             preferences based on party lines and
                                                                                                up to FDI without                           increased their support for parties more
                                                                                                a license., industry                        likely to implement complementary
                                                                                                delicensed, share                           policies to alleviate the distress caused
                                                                                                of Assembly                                 by reform." (11)
                                                                                                Constituencies
                                                                                                held by the party
                                                                                                of the state
                                                                                                government


Thachil   India    Political            2008 voters in    vote for Hindu       survey           membership in          religiosity,         participation in service activities
(2011)             returns                   24 villages nationalist party     conducted by     party auxilliary       communalism,         delivered by Hindu nationalist party's
                                             in 4                              author           organization,          income, caste        auxilliary organization increases
                                             districts in                                       participant in                              likelihood of support for that party even
                                             Chhattisgar                                        party auxilliary                            on the party of low-caste voters
                                             h                                                  organization
                                                                                                activities,
                                                                                                nonparticipant
Thames    Russia   Political   1994-6        individual   individual           Center of        deputy’s faction                            Yeltsin has used the budget process to
(2000)             returns                   deputies     deputy’s vote on     Applied          type, cadre,                                build a coalition of supporters that
                                             (less than   the final budget     Political        corporatist or mass-                        reduced the need for legislation by
                                             450)         vote for each year   Research,        based, ideology                             presidential decree. Deputies from
                                                                               INDEM                                                        corporatist factions has largest impact on
                                                                               Statistics 2.0                                               Yeltsin's budgets.
Zucco     Brazil   Political   2006          municipalit Lula's vote share                      Bolsa Familia          level of             "Bolsa Familia had a considerable impact
(2010)             returns                   y                                                  cash transfer          socioeconomic        on the voting decisions of individuals…
                                                                                                program                development, level   government's support among non-
                                                                                                                       of inequality,       beneficiaries of the program is also very
                                                                                                                       growth, family       high in the less developed regions of the
                                                                                                                       income               country." (25)
Alesina,   Italy   Redistributive 1995      individuals log of hourly        Bank of Italy                        high school degree,   "In summary, the North redistributes to
Danninger,         consequences             (23924)     earnings of full     survey 1995                          college, year work    the South in a way that creates negative
and                                                     employed                                                  experience, female,   cultural and social side effects. But, if
Rostagno                                                workers                                                   married mid           this is the case, why is this redistributive
(1999)                                                                                                            management            system chosen? The answer may be that
                                                                                                                                        redistribution through public
                                                                                                                                        employment is less visible than direct
                                                                                                                                        transfers, therefore it is politically less
                                                                                                                                        costly and it is more effective at creating
                                                                                                                                        patronage and political benefits for local
                                                                                                                                        politicians." (4) "we see that public
                                                                                                                                        sector wages are not statistically different
                                                                                                                                        between the South and the North. On the
                                                                                                                                        contrary the results for the private sector
                                                                                                                                        are quite different….Southern residents
                                                                                                                                        earn on average about 18.9 less than their
                                                                                                                                        northern counterparts." (14)



Barkan and Kenya   Redistributive 1983-88   province    District Focus for   private         Distribution                               "the significance of District Focus as a
Chege              consequences             (7)         Rural                correspondenc   patterns of roads                          mechanism to achieve a redistribution of
(1989)                                                  Development          e and           and health                                 resources and power in Kenya is
                                                        Fund (non-           Development     (compared these                            probably greatest in political rather than
                                                        regression           Estimates for   figures to Rural                           economic terms." (450) "Allocations for
                                                        analysis)            the Year        Development                                roads and health, however, indicate that
                                                                             1987/88         Fund figures, non-                         significant redistribution of development
                                                                                             regression                                 expenditure requires the intervention of
                                                                                             analysis)                                  central ministries, because it is they
                                                                                                                                        which unilaterally allocate the most
                                                                                                                                        funds to the rural areas." (450)
Besley and India      Redistributive 1958-1992 states (16) public            Bulletin on        food grain            ratio of urban to     Turnout in previous election does not
Burgess               consequences                         distribution of   Food Statistics,   production per        total population,     affect responsiveness of state
(2002)                                                     food and state    Ministry of        capita, real per      population density,   governments. Greater political
                                                           government        Food and           capita flood          log population,       competition associated with higher levels
                                                           expenditures on   Agriculture,       damage to crops,      Consumer Price        of public food distribution, but not
                                                           calamity relief   Directorate of     newspaper             Index for             calamity relief. Levels of food
                                                                             Economics and      circulation (media    Agricultural          distribution respond to political
                                                                             Statistics,        development),         Laborers and          competition and timing of elections, but
                                                                             Government of      ownership of          Industrial Workers    calamity relief does not. (1435) 10%
                                                                             India, Reserve     newspapers,                                 drop in food production will bring 1%
                                                                             Bank of India      turnout, political                          increase in public food distribution, but
                                                                             Bulletin           competition,                                in states with high newspaper circulation
                                                                                                revenue from                                a 10% drop will give a 2.28% increase in
                                                                                                center                                      public food distribution. This means
                                                                                                                                            incumbents will respond more to a shock
                                                                                                                                            when media is more developed. (1436)


Bodenstein European   Redistributive 2000-2006 region      allocation of    European            Federalism,           GDP per capita,       Objective-1 funding (selection of eligible
and        Union      consequences                         Structural Funds Commission          electoral             unemployment,         regions) is basically driven by regional
Kemmerlin                                                                                       competition           agricultural funds    per capita income, whereas the selection
g                                                                                               (divergence           per capita            of Objective-2 (structural funds
(2008)                                                                                          between two                                 allocation to eligible regions) regions is
                                                                                                largest parties),                           determined by the regional
                                                                                                effective number                            unemployment rate. (15)
                                                                                                of parties
Bouvet and European   Redistributive 1994-99    region     allocation of    European            left-wing national    GDP per capita,       "Left-wing governments and more EU-
Dall'erba  Union      consequences              (120)      Structural Funds Commission          and regional          unemployment,         sceptical countries get more objective 3
(2010)                                                                      (1989–93 are        governments,          agricultural funds    and 4 and 5 funds, but they get less aid
                                                                            from                percentage of EU-     per capita            through objective 2." (523) "Political
                                                                            ‘Community          sceptics in                                 bargaining will always be part of the
                                                                            Structural          national                                    allocation process because
                                                                            Interventions’,     populations,                                there are too many potential recipient
                                                                            Statistical         National and                                regions, and the decision process
                                                                            Report No. 3        regional electoral                          involves interaction between several
                                                                            and No. 4           margins, Dummy                              levels of the political arena." (524)
                                                                            (European           variable for a safe
                                                                            Commission,         electoral margin,
                                                                            1992a, b) and,
                                                                            for 1994–99
Brown and multi-       Redistributive 1973-1997 countries       agricultural/resid   World Bank,     democracy             population, urban "democratization increases the residential
Mobarak   country      consequences             (57)            ential/"other"       World           indicator,            population,          sector's share of consumption relative to
(2009)                                                          sector's share of    Development     democracy*GDP         electricity          industry" (200)
                                                                electricity          Indicators      interactions          production,
                                                                consumed relative                                          electricity
                                                                to industry                                                privatization, total
                                                                                                                           electricity
                                                                                                                           consumption, GDP
                                                                                                                           per capita, GDP
                                                                                                                           per capita squared


Bungey,    Australia   Redistributive 1956-57 --   states (6)   special purpose      Commonwealt     proportion of         proportion of urban   "Grants may well go to states where the
Grossman,              consequences 1985-86                     recurrent grants,    h               state's seats held    population,           federal government's need to increase
and Kenyon                                                      special purpose      Governments'    by the federal        unemployment          voter support is greatest. However, as the
(1991)                                                          capital grants,      Budget Paper    government's          rate, taxable         position of the government in the state
                                                                special purpose      Number 7,       party, proportion     income per capita,    improves, pork-barelling behavior is
                                                                total (sum),         Payments to     squared, dummy if     population density    likely to take over" (negative slop on
                                                                general purpose      or For the      federal and state     in rural areas,       proportion of state seats held by federal
                                                                grants               States, the     governments are       population, school    government party but positive slope on
                                                                                     Northern        of the same           pupils as             term squared) (665). Marginal seats not
                                                                                     Territory and   political party,      proportion of         significant. Dummy on state and central
                                                                                     Local           proportion of seats   population,           government sharing a party is significant
                                                                                     Government      held with less than                         "but has the wrong sign" (666).
                                                                                     Authorities     a 5% two party
                                                                                                     preferred vote
                                                                                                     ("marginal seats"
                                                                                                     or unsafe seats),
                                                                                                     proportion private
                                                                                                     school pupils
                                                                                                     (capturing
                                                                                                     preferences in
                                                                                                     DLP), proportion
                                                                                                     of federal seats
                                                                                                     decided on
                                                                                                     preferences
                                                                                                     ("marginal seats"
                                                                                                     or unsafe seats)
Castells and Spain   Redistributive 1987-1996 electoral   infrastructure   Fundacion   Position in the       output growth,        (i) "regional governments need more time
Solé-Ollé            consequences             regions     investments      BBVA        electoral cycle,      population growth,    to adjust to the long-run equilibrium, (ii)
(2005)                                        (NUTS3)                                  plus Electoral        growth in ratio of    are inclined towards efficiency, and (iii)
                                              (50), which                              Productivity:         trucks to output,     are more responsive to infrastructure
                                              receive                                  pivotal               growth in vehicle-    needs" (1200) Vote margin (inverse of
                                              allocations                              representative        km per year,          number of votes needed to gain/lost a
                                              from both                                dummy, number         growth in rail        seat) and turnout are mostly positive and
                                              the central                              of votes needed for   passengers per        statistically significant. The measure of
                                              and                                      incumbent to          year, growth in       swing voters (cut point density) is not
                                              regional                                 gain/lose a seat,     tons transported      stat sig. When combined into an
                                              governmen                                turnout in            per year, growth in   electoral productivity composite, they
                                              ts. Central                              legislative           air passengers per    become positive and stat sig. Thus
                                              and                                      elections, cut-       year, growth in       evidence "in favour" of the Dixit-
                                              regional                                 point density of      regional general      Londregan model (1199). Leftist
                                              allocations                              legislative           revenues, growth of   regional governments (sharing party with
                                              investigate                              elections,            regional capital      center) and a higher share of socialist
                                              d                                        Partisanship:         revenues, growth in   regional representatives attract more
                                              separately                               Incumbent's share     net debt of region    investment. The effect of vote share for
                                                                                       of votes in last                            incumbents is positive but not stat sig..
                                                                                       election, dummy                             "Constituency support (i.e., vote share)
                                                                                       for regional                                does not influence investment (Cox and
                                                                                       socialist                                   McCubbins 1986)..." (1199).
                                                                                       government
                                                                                       (socialists always
                                                                                       the central
                                                                                       government
                                                                                       incumbent),
                                                                                       socialist share in
                                                                                       regional
Cingano     Italy   Redistributive 1985-97   manufactur firm revenues    Italian    fraction of sample   According to our results, access to
and Pinotti         consequences             ing firms deflated using 2- National   period in which      political connections increases firm
(2009)                                       (1227)     digit industry   Accounts   firm was             revenues by approximately 5%, yielding
                                                        indexes                     connected to a       to an almost equivalent change in current
                                                                                    politician           profits. These gains only accrue to firms
                                                                                    (measured by         establishing a connection through
                                                                                    indicator variable   politicians appointed with the party that
                                                                                    of at least 1        won the elections: firms connected
                                                                                    employee of firm     through other politicians see no increase
                                                                                    being appointed in   in market shares, just as non-connected
                                                                                    a local              firms." (5-6) "estimates from alternative
                                                                                    government),         production function specifications
                                                                                    politician elected   indicate that political connections do not
                                                                                    with a party that    have any impact on productivity. Rather,
                                                                                    won elections        the average effect on revenues turns out
                                                                                    (measured by an      to be driven by firms operating in
                                                                                    indicator variable   markets in which public demand plays a
                                                                                    for at least 1       major role." (7)
                                                                                    employee of firm
                                                                                    being appointed in
                                                                                    a local government
                                                                                    with the winning
                                                                                    coalition)
Cole,      India       Redistributive 1977-99   administrat    crop yield, state   Sanghi,       weather measured   "Extreme amounts of rain lead to higher
Healy, and             consequences             ive            expenditure on      Kumar,        by rainfall        amounts of spending -- particularly
Werker                                          district*ele   disaster relief,    McKinsey                         droughts. Budget transfers from central
(2008)                                          ction in 28    vote share for      (1998, crop),                    government for the purpose of calamity
                                                states         incumbent ruling    Reserve Bank                     relief were often provided as matching
                                                (21,532        party candidate     of India,                        funds, directly proportional to the
                                                elections,                         Election                         amount spent by the state." (14) Ruling
                                                594                                Commission of                    party does very poorly during extreme
                                                districts)                         India                            droughts but its performance increases
                                                                                                                    steadily with rainfall, reaching an
                                                                                                                    optimum at a point between 0 and 1
                                                                                                                    standard deviations above the mean. As
                                                                                                                    rainfall exceeds this optimum, support
                                                                                                                    for the ruling party declines." (14-15)
                                                                                                                    Voters punish ruling coalition even for
                                                                                                                    circumstances beyond its control.
                                                                                                                    "Voters are twice as sensitive to relief
                                                                                                                    spending during extreme weather as they
                                                                                                                    are to relief spending during bad
                                                                                                                    weather" (19)




Dall'erba   European   Redistributive 1989-1999 145 regions    regional per      New Cronos                         "We first used Moran’s I to detect the
(2005)      Union      consequences             at NUTS II     capita GDP series Regio by                           presence of positive global spatial
                                                (Nomencla      in Ecu current    Eurostat                           autocorrelation in the distribution of per
                                                ture of        prices (spatial                                      capita GDP to explore whether the rich
                                                Territorial    distribution)                                        (poor) regions have a propensity to be
                                                Units for      (exploratory)                                        clustered close to other rich (poor)
                                                Statistics)                                                         regions. Global spatial autocorrelation
                                                over 12 EU                                                          also characterizes the regional growth
                                                countries                                                           rate, structural funds and Community
                                                                                                                    projects total cost." (146) Evidence
                                                                                                                    shows this persists over time.
de la      Spain    Redistributive 1981--1990 regions    infrastructure      National        number of years       expected return to infrastructure investments are not higher
Fuente,             consequences              (17)       investments         Institute of    regional              investment, income in regions that have the same party as the
Vives,                                                                       Statistics,     government            per capita, effective central government (Socialist) and are
Dolado and                                                                   national        controlled by         public capital        not higher in regions with
Faini                                                                        Census,         Socialists (the       stock, dummy 1986- nationalist/regionalist parties (taken to be
(1995)                                                                       Municipal       national              -1990                 more active in lobbying for regional
                                                                             Census, Labor   incumbent),                                 funding). "The party in power remains
                                                                             Force Survey,   number of years                             insignificant in all specifications" (34)
                                                                             Mas et al       regional                                    Negative slope for income -- "on average
                                                                             (1993)          government                                  redistributive concerns outweighed the
                                                                                             controlled by                               political influence of the richer regions"
                                                                                             nationalist/regiona
                                                                                             list parties (more
                                                                                             active in
                                                                                             lobbying), dummy
                                                                                             for Basque-
                                                                                             Navarre

De Rynck European   Redistributive 1999 to    EU regions allocation of                       cohesion policy                            "While the evidence points to a stronger
and       Union     consequences present                 Structural Funds                                                               emphasis for social and local problems
McAleavey                                                (non-regressional                                                              and more intense interaction between the
(2001)                                                   analysis,                                                                      Commission and local actors, the
                                                         exploratory)                                                                   dominant focus on economic problems
                                                                                                                                        and regional disparities remained intact
                                                                                                                                        in the 1999 reform." (553) GDP per
                                                                                                                                        capita remains central to the Commission
                                                                                                                                        in deciding redistribution.
Dellmuth   European   Redistributive 2000-2006 EU regions amount of            European          ratio of paid SF in GDP, employment,   GDP per capita and unemployment rate,
(2009)     Union      consequences             (148)      structural funding   Commission        relation to         share of public    are relatively identical and statistically
                                                          (SF) granted to      decisions         potentially         sector employees   significant. "The measure introduced to
                                                          regional             (various years)   available funds                        capture the capability of regions to cater
                                                          development                            (information), left                    to the Commission’s demand for
                                                          plans (log)                            parties vote                           information about the implementation of
                                                                                                 shares, political                      SF on the ground supports the argument
                                                                                                 centre of gravity                      that the Commission distributes higher
                                                                                                 proposed by Gross                      funding to regions which can help
                                                                                                 and Siegelman                          overcoming informational asymmetries
                                                                                                 (1984), effective                      and increase the prestige and reputation
                                                                                                 number of parties,                     of the Commission. The variable is
                                                                                                 support for the                        statistically significant across all models
                                                                                                 EU, Federalism                         and has the expected sign." (17) "The
                                                                                                 index, EU support,                     empirical results show that the nature of
                                                                                                 election year                          the allocation process appears to ensure
                                                                                                 dummy                                  that national politicians cannot use the
                                                                                                                                        Community resources to build electoral
                                                                                                                                        support." (19)



Doménech, European    Redistributive 1986-199              transfers per       Commission                          per capita income    Regional funds are negatively related to
Maudes,    Union      consequences                         capita              accounting                                               per capita income and redistributive
and Varela                                                                     system                                                   effect has increased over time. Social
(2000)                                                                         (SINCOM)                                                 funds also negatively related but
                                                                                                                                        redistributive effect is constant over time.
                                                                                                                                        Total expenditures show a negative
                                                                                                                                        relationship with per capita income. "we
                                                                                                                                        have found that per capita contributions
                                                                                                                                        increase
                                                                                                                                        proportionally to per capita income as a
                                                                                                                                        country becomes richer, but the
                                                                                                                                        Community funds received remain
                                                                                                                                        constant or even decrease. This result
                                                                                                                                        implies that the EU budget has a clear
                                                                                                                                        redistributive effect on
                                                                                                                                        the income of its members in per capita
                                                                                                                                        terms." (648)
Fattore    European   Redistributive 1975-1995 state       percentage of the European        % of regions        GDP, New State       "The coefficient for the lagging region
(2004)     Union      consequences                         European           Commission     within a state that Year, year           variable is positive as expected, as well
                                                           Regional                          the European        dummies              as statistically significant. Substantively,
                                                           Development                       Commission has                           for every one unit increase in the
                                                           Fund (the largest                 identified as                            percentage of lagging regions within a
                                                           portion of the                    eligible for ERDF                        specific state, it can be expected that that
                                                           current Structural                grants via                               state will also receive a .08% increase in
                                                           Funds) yearly                     Objective 1 or 2 of                      their share of the ERDF disbursements
                                                           budget allotted to                SF.                                      for that year." (15) "richer states
                                                           a specific state                                                           contribute more to the EU budget, and,
                                                                                                                                      therefore, exert more influence over
                                                                                                                                      decision-making regarding monetary
                                                                                                                                      issues such as the Structural Funds.
                                                                                                                                      However, this hypothesis was not
                                                                                                                                      supported by the model’s results." (15-
                                                                                                                                      16)


Gauthier   Chad       Redistributive      2004 health      whether health       Healthy       frequency of         location, no. of   "results show discretion of district and
and Wane              consequences             centers     center received      Facilities   supervision visits    employees          regional administrators has a significant
(2008)                                         (281) and   positive amount      Survey by    by regional or                           impact on probability of receiving public
                                               hospitals   of public            World Bank   district officials,                      resources, in that the probability of
                                               (30)        resources            2004         foreign donor                            receiving material resources increases
                                                           (financing, drugs,                support                                  significantly for health centers that have
                                                           material)                                                                  been visited by the regional delegate."
                                                                                                                                      (70)
Kauppi,      European   Redistributive 1976-2001 region      country i's EU   European     political power      agricultural        "Our results indicate that at least 60% of
Widgén,      Union      consequences                         budget share     Commission                        production, GDP     the budget expenditures can be attributed
and Carrillo                                                                                                    per capita versus   to selfish power politics and the
(2004)                                                                                                          EU wide GDP per     remaining 40% to the declared
                                                                                                                capita              benevolent EU budget policies. However,
                                                                                                                                    when we apply specific voting power
                                                                                                                                    measures that allow correlated
                                                                                                                                    preferences and cooperative voting
                                                                                                                                    patterns between the member states, our
                                                                                                                                    estimates indicate that the power politics
                                                                                                                                    view can explain as much as 90% of the
                                                                                                                                    budget shares. We conclude that power
                                                                                                                                    politics can explain a major part of the
                                                                                                                                    Council decisions and that correlated
                                                                                                                                    preferences and voting cooperation
                                                                                                                                    between EU countries potentially play
                                                                                                                                    significant roles in EU decision making."
                                                                                                                                    (223)


Kemmerlin European      Redistributive      1999 countries   allocation of    EU           Effective Number     GDP,                "As expected, GDP per capita has a
g and      Union        consequences             (71)        Structural Funds Commission   of Parties (ENP),    unemployment        negative and strongly significant impact
Bodenstein                                                   (SF) per capita               presence of          rate, agriculture   on the regional allocation of SF. The sign
(2006)                                                                                     federalist systems   sector size         of the coefficient of the unemployment
                                                                                           (FED).                                   rate (UNEMP) is positive but
                                                                                                                                    insignificant. Allocation of Agricultural
                                                                                                                                    Funds per capita (AGRAR/CAP) has a
                                                                                                                                    positive coefficient, suggesting that
                                                                                                                                    beneficiaries of Agricultural Funds also
                                                                                                                                    receive more SF....The share of left
                                                                                                                                    parties in the European elections at the
                                                                                                                                    regional level (LEFTPAR) and the share
                                                                                                                                    of Eurosceptic parties (ESCEP) have the
                                                                                                                                    expected positive signs and are also
                                                                                                                                    significant at the 10% level... Neither the
                                                                                                                                    federalism (FED) nor the multiparty
                                                                                                                                    system indicator (ENP) performs
                                                                                                                                    particularly well" (13)
Kemmerlin Germany   Redistributive 1980, 1986, cities (83) grant allocation   Statistical      congruence           private capital        "The main findings of our analysis are (i)
g and               consequences 1988                                         Yearbook of      between the local    stock, gross value     the contribution of local public capital to
Stephan                                                                       German Cities    city government      added of a city's      private production in cities is posit- ive
(2002)                                                                        and              and the state        manufacturing          and significant (ii) political affiliation,
                                                                              Municipalities   ('Bundesland')       sector,                measured by the coincidence of party
                                                                                               government,          infrastructure         colour between state and local
                                                                                               percentage           capital stock,         government, is decisive in explaining the
                                                                                               difference of the    Annual investment      distribution of investment grants across
                                                                                               two major parties,   in infrastructure,     cities (iii) cities with a prevalence of
                                                                                               which are the        labour input,          'marginal voters' neither spend more on
                                                                                               parties SPD and      number of working      public infrastructure nor receive more
                                                                                               C여                   hours in the           investment grants from higher-tier
                                                                                                                    manufacturing          governments (iv) redistributive concerns
                                                                                                                    sector, special        of higher-tier governments matter for the
                                                                                                                    grant-in-aid for       allocation of grants, whereas efficiency
                                                                                                                    investments,           considerations appear to be less
                                                                                                                    financial situation    important." (421)
                                                                                                                    of a city, trade tax
                                                                                                                    revenues, number
                                                                                                                    of (four-wheel)
                                                                                                                    motor vehicles,
                                                                                                                    number of
                                                                                                                    manufacturing
                                                                                                                    firms,
Khemani    India      Redistributive 1972-1995 major       disaggregated       Reserve Bank      Political affiliation   real state income       "Plan grants are directed toward
(2003)                consequences             states (15) intergovernmental of India            dummy (=1 if            per capita, total       affiliated states" especially where the
                                                           transfers: (1)      Bulletin          center and state        population, state       proportion of seats is small -- "greater
                                                           statutory transfers                   governments             election cycle, state   transfers are directed towards those states
                                                           (Finance                              belong to the same      government              where the national ruling party has more
                                                           Commission), (2)                      party),                 coalition dummy,        seats to gain, rather than to the core
                                                           plan grants                           affiliation*seats       interaction of          support states where it already controls a
                                                           (Planning                             (proportion of          affiliation after       high proportion of seats to the national
                                                           Commission), (3)                      seats controlled by     1990 (national          legislature" (21, 22). "On the other hand,
                                                           central schemes                       national ruling         coalition)              statutory transfers...are [lower]" for
                                                           (discretionary)                       party),                                         affiliated states, perhaps to correct for
                                                                                                 affiliation*(1-                                 politically-motivated plan transfers (22).
                                                                                                 seats), Congress                                Central scheme transfer results are
                                                                                                 Party dummy                                     insignificant except for a "very restricted
                                                                                                                                                 political variable" (25). Suggests district-
                                                                                                                                                 level is more appropriate for central
                                                                                                                                                 scheme transfers



Kudamatsu 28 African Redistributive latest        individual   individual-level    Demographic   dummy for               mother fixed            "Democratization since 1990 is followed
(2012)    countries consequences survey per       baby by      infant mortatiliy   and Health    democratic year         effects, mother's       by a reduction in the infant mortality rate
                                    country       mother       dummy for           Surveys                               birth cohort by         by 1.2 percentage points, which is as
                                    since 1996-   (643846)     whether a child                                           child's birth year      much as 12% of the sample mean" (16).
                                    2006                       dies before age                                           fixed effects,          "The probability that a mother sees her
                                                               one.                                                      dummy for girls,        baby die within the first month of life
                                                                                                                         multiple births, and    falls by 0.6 percentage points after
                                                                                                                         birth order             democratization" (16).
Lake and   multi-    Redistributive various     countries   education           World Bank   level of democracy     per capita GNP,       Level of democracy is positively
Baum       country   consequences years 1975, (approx. 55   (illiteracy, pupil-              as measured by         land area,            significant with the level of public goods
(2001)                              1985,       to 85)      teacher ratio,                   contestability of      population, percent   provision at the 0.05 level or better in
                                    1987, 1990,             school                           the political          of urban              84% of the 38 OLS regressions reported
                                    1992                    enrollment, etc.)                market and lags        population, OECD      (609). In general, impact of regime type
                                                            and public health                                       dummy                 on public services is strongest within 2
                                                            (health care                                                                  years or less.
                                                            access, clean
                                                            water access, no.
                                                            of physicians,
                                                            unintended births,
                                                            infant mortality
                                                            etc.) indicators



Luna and   Chile     Redistributive 2000-2008 municipalit 8 types of social    National      electoral results      municipality's        Overall programs effectively target poor
Mardones             consequences             y (1911)    transfers and        System of     for 5 bigger parties   dependence on         communities with higher social
(2010)                                                    targeted subsidies   Municipal     for 3 elections at     Municipal             (education and health) needs. (11) "even
                                                          (i.e. family,        Indicators    municipal level        Common Fund,          in context of efficient socioeconomic
                                                          drinkable water                    (1996, 2001,           municipality's        targeting and fiscal oversight, political
                                                          and sewerage,                      2004), dummy for       health and            targeting is also present; and, political
                                                          pension,                           reelected mayors,      education needs,      investment rationales in Chile diverge
                                                          education, health)                 participation rate,    bureaucratic          from the standard predictions that have
                                                                                             competitiveness of     capacity, year        been derived from machine party ridden
                                                                                             the race                                     and/or weak states cases. (2)
Miguel   Kenya and Redistributive Kenya:       villages (66   various local         Busia, Kenya:     local ethnic          average years of     "The school funds result implies that the
(2004)   Tanzania consequences 1995-1996       in Meatu,      school funding:       1996 ICS          diversity             education,           change from ethnic homogeneity to
                                  (well data   Tanzania)      annual school         School and        (ethnolinguistic      proportion formal    average levels of diversity in Busia,
                                  from 2000-   and            spending per          Pupil             fractionalization,    sector employment,   Kenya is associated with a drop of
                                  1)           primary        pupil, desks per      Questionnaires    ELF), interaction     proportion homes     approximately 25 percent in average
                                  Tanzania:    school         pupil, latrines per   , 1996            term for difference   with iron roofs,     funding - a large effect - while the
                                  1997-2002    community      pupil, classrooms     Government        between the           proportion           estimated effect for Meatu, Tanzania, is
                                               which          per pupil,            Examinations      impact of ethnic      households grow      positive and statistically significant."
                                               includes       proportion wells      Namelists,        diversity on the      cash crops,          (353) "local ethnic diversity has a
                                               area within    with normal           Global            public goods          proportion           significantly more negative effect on
                                               5 km of a      flows                 Positioning       outcome in Kenya      households own       local public goods provision in Busia,
                                               primary                              Systems (GPS)     versus Tanzania       cattle, proportion   Kenya, than in Meatu, Tanzania." (357)
                                               school (84                           readings taken                          Catholic
                                               in Busia,                            by NGO field-
                                               Kenya)                               workers,
                                                                                    survey on well
                                                                                    conditions by
                                                                                    NGO field-
                                                                                    workers for
                                                                                    this study.
                                                                                    Meatu,
                                                                                    Tanzania:
                                                                                    2001-2
                                                                                    household
                                                                                    village council
                                                                                    surveys
Min        India   Redistributive 1992-2003 villages in electrification of     satellite images   indicator for         village population,    "When it comes to village electrification,
(2009)             consequences             Uttar       village                from Defense       village represented   presence of school     the differences across parties are
                                            Pradesh                            Meteorological     by a BSP              or medical facility,   substantial, and the most positive effects
                                            (approx.                           Satellite          (Bharatiya Samaj      literacy rate,         are in villages located in BSP
                                            1.2 million                        Program's          Party) legislator     distance to nearest    constituencies." (24) "Using matching
                                            = approx.                          Operational        or BJP (Bharatiya     town, paved            technique to evaluate similar villages that
                                            98,000                             Linescan           Janata Party) in      approach road,         differ only on whether they switched to
                                            villages                           System             the State             income index,          BSP representation in the critical 2002
                                            *14 years)                                            Assembly,             nighttime light        elections, I also show a positive BSP
                                                                                                  population            output in              treatment effect." (34)
                                                                                                  classified as         constituency
                                                                                                  Scheduled Caste       (1992)
                                                                                                  according to 2001
                                                                                                  Indian Census,
                                                                                                  interaction term of
                                                                                                  the two, other
                                                                                                  party indicators
                                                                                                  Samajwadi Party,
                                                                                                  Congress



Palaniswam India   Redistributive 2005-2006 gram          income and           gram               proportion of       district dummies,        SGRY expenditure at GP level is not
y and              consequences             panchayat     expenditure from     panchayat,         reserved seats for population                significantly influenced by composition
Krishnan                                    in            gram panchayat's     survey by IF       various social                               of GP or by elite bargaining power
(2008)                                      Karnataka     Sampoorna            Panchayati Raj     groups, minorities,                          variables, but they are large in
                                            (79)          Grameen Rozgar       Institutions       women, total                                 magnitude. GP population and district
                                                          Yojana (SGRY),       and Institute      number of seats in                           dummies, which are used in the
                                                          a program for        for Social and     GP, elite capture                            formulas, are most significant in
                                                          wage employment      Economic           measured by land                             determining SGRY income (27-28).
                                                          and infrastructure   Change             ownership and
                                                                                                  political
                                                                                                  competition of
                                                                                                  councilors
Peisakhin   India      Redistributive    2008 individual   right/register to    field           field experiment                           Recourse to the RTI substantially speeds
(2009)                 consequences           (61 from     vote (placement      experiment      with 3 groups:                             up the voter registration process by
                                              slums, 61    on electoral roll)   conducted by    information                                comparison to standard application
                                              middle-      measured by          authors in      treatment (Right to                        procedure. Urban poor are 24 times more
                                              class)       length of time       2008            Information Act),                          likely to be placed on the roll than the
                                                           before an                            bribe treatment,                           control, and middle class 14 more times.
                                                           applicant is added                   control                                    Threat of information disclosure forces
                                                           to the draft                                                                    public officials to operate more
                                                           district electoral                                                              efficiently, and can be a legal substitute
                                                           roll                                                                            to bribery (23)


Pereira     Portugal   Redistributive    1989 communiti intergovernmental Direccao-Geral        "the proportion of    structure of local   "The 'political' variable P under this
(1996)                 consequences           es (186)  lump-sum grants da                      votes in local        government lobby     alternative political hypothesis is not
                                                        from a fund for   Administracao         elections for the     organization, per    statistically significant even at a 90%
                                                        financial         Autarquica            political party       capita tax base,     degree of confidence, and there are at
                                                        imbalance "which (DGAA)                 which form the        economies of         least two reasons to explain this result."
                                                        consolidates                            government" (196)     sharing the          One is that rewarding core supporters
                                                        almost all                                                    consumption of       and buying off opposing voters might
                                                        transfers from                                                capital intensive    cancel each other out. The other is that
                                                        central to local                                              public goods         parties care only about the total number
                                                        governments"                                                  (economies of        of votes, not about the vote share
                                                        (194)                                                         scale)



Porto and Argentina Redistributive 1960-1999 provinces transfers per            Porto (1990), deputies per            population share,    "The results of the estimation strongly
Sanguinetti         consequences             (88=22      capita                 population      capita, senators      income per capita    support the hypothesis that the political
(2001)                                       provinces *                        share from      per capita            and density          representation at the National Congress
                                             4 decades)                         national census                                            has been a key variable that has
                                                                                (INDEC 1993)                                               determined the distribution of transfers
                                                                                                                                           across provinces in Argentina in the last
                                                                                                                                           decades. The estimated coefficients for
                                                                                                                                           the variables deputies and senators, both
                                                                                                                                           expressed in per capita terms, are
                                                                                                                                           positive and significant in all the
                                                                                                                                           regressions." (252)
Rodden      European   Redistributive 1977-1999 countries   real per capita net                 votes per capita     agricultural share Overrepresented states are rather
(2002)      Union      consequences             (14)        transfers paid or                                        of workforce, GDP dramatically favored in the distribution
                                                            received, real per                                       per capita,        of both agricultural and structural funds.
                                                            capita transfers                                         population,
                                                            received by each
                                                            member state



Sapienza    Italy      Redistributive 1991-1995 firms       interest rate     Centrale dei      local political      size of bank,       "The political strength of the party has a
(2004)                 consequences                         charged to firm I Bilanci           strength --          percentage of non- negative and significant effect on the
                                                            by bank k at time                   measured by the      performing loans, interest rate charged to borrowers"
                                                            t minus the prime                   percentage of        market
                                                            rate at time t                      votes received by    concentration, size
                                                                                                the party to which   of firm
                                                                                                the chairperson of
                                                                                                the state-owned
                                                                                                bank is affiliated
                                                                                                in the area where
                                                                                                the firm is
                                                                                                borrowing

Stasavage   44 African Redistributive 1980-1996 country-   public spending        UNESCO,       multiparty           GDP per capita       Multipary political competition results in
(2005)      countries consequences              year (365) on                     World         competition,         (log), aid (%        1.1% of GDP increase of total education
                                                           eduction/primary       Development   election year        GDP), rural          and 4.4% of total expenditure increase to
                                                           education/univers      Indicators                         population (%),      education. Multiparty competition would
                                                           ity a(% GDP and                                           population under     increase primary education spending by
                                                           % total govt.                                             15 (%)               0.45% of GDP and 1.5% in share of total
                                                           spending)                                                                      govt. spending (351). As hypothesized,
                                                                                                                                          spending on unversities remains
                                                                                                                                          unchanged with electoral competition.
Vicente     Sao Tome Redistributive        2006 household     change in vote       survey         anti vote-buying      demographic          "vote buying is effective in changing
(2007)      and      consequences               (1275 in      buying               conducted by   campaign, dummy       controls, location   voting preferences/behavior. In
            Principe                            preelection   (perceptions on      author         for urban location,   controls, dummy      particular, when compared with the
                                                survey,       parliamentary vs.    (randomized    dummy variable        variable of 1 in     incumbent, the challenger seems to
                                                1034 in       presidential         field          for swing zones       treated areas,       benefit more from vote buying. In
                                                post-         elections); change   experiment)                          information          addition, we show that vote buying
                                                election      in voting                                                 gathered about       seems to energize the electorate." (32)
                                                survey)       behavior for                                              candidates,          "Regarding the prices of votes, ... higher
                                                              presidential                                              electoral            in urban areas and swing zones." (27)
                                                              election (intended                                        observation          Effect of anti-vote buying campaign on
                                                              vs. actual                                                mission of CPLP      change in voter turnout is not significant
                                                              reported)                                                 (minutes spent at    (30).
                                                                                                                        each location,
                                                                                                                        survey-based
                                                                                                                        measures of
                                                                                                                        psychological
                                                                                                                        consistency of
                                                                                                                        respondents

Wilkinson   India     Redistributive 1960-1992 major       capital road            Khemani        political             ethnic               Political competition has significant
(2006)                consequences             states (14) expenditure             (2004),        competition (state    fractionalization,   effect on capital spending on roads. "an
                                                                                   Reserve Bank   electoral             religious            increase of one standard deviation from
                                                                                   of India and   volatility), state    fractionalization,   the mean level of electoral volatility in a
                                                                                   State data     party                 total state          state would increase the level of capital
                                                                                                  fractionalization     population,          road expenditure from $7.349 million to
                                                                                                  adjusting for         literacy, police     $8.542 million." (21) Political
                                                                                                  independent,          density, deaths in   competition also drives actual road
                                                                                                  electoral cycle       communal riots in    construction but road quality is
                                                                                                  (state and national   previous 5 years     questionable. (22)
                                                                                                  election)
Su and    China   Non-        1978-1994 province/ye provincial shares Database           Central Committee   provincial income    Central Committee Index positive and
Yang              democracy             ar          of state sector                      Index (CCI) which   per capita,          significant at 0.01 level, implying a
(2000)                                              capital                              shows provincial    population size.     province that enjoys greater
                                                    construction                         representation      Geographic           representation on Party Central
                                                    investment                                               location indicator   Committee is likely to have more
                                                                                                             (central, coastal)   resources allocated to that province from
                                                                                                                                  the central government. (223) As far as
                                                                                                                                  central cities, Beijing, Shanghai, Tianjin,
                                                                                                                                  greater representation on the Central
                                                                                                                                  Committee is associated with less state
                                                                                                                                  sector capital construction investment,
                                                                                                                                  and estimate is insignificant at 0.01 level.
                                                                                                                                  Thus supports autonomous center model.
                                                                                                                                  (225)



Blaydes   Egypt   Non-        1986-1996 governorat change in the        Central          previous level of                        "Governorates with high levels of
(2011)            democracy             es (26)    percentage           Agency for       infrastructure                           support for the Brotherhood-Wafd
                                                   of residences in a   Public           coverage, urban-                         alliance (i.e.
                                                   governorate that     Mobilization     rural status,                            vote share around 50 percent) saw 30
                                                   have piped public    and Statistics   Brotherhood-Wafd                         percent less water infrastructure
                                                   water, change in                      voteshare in the                         development than governorates that saw
                                                   the percentage of                     1984 election,                           low levels of support for the Brotherhood-
                                                   residential                           percent of non-                          Wafd alliance (i.e. voteshare
                                                   buildings in a                        NDP                                      around 0 percent)" (105). Similar results
                                                   governorate that                      parliamentary                            for sewerage infrastructure.
                                                   are connected to                      representatives
                                                   the public                            between 1984-
                                                   sewerage network                      1995
Berman,    Iraq   Non-        2004-2007 district      intensity of         "significant     spending by         district           "spending on public goods is
Shapiro,          democracy             (832 =        insurgent activity   activity"        Coalition forces    characteristics,   unconditionally correlated with greater
Felter                                  104district   (rate of attacks     (SIGACT)         on small-scale      year indicators,   violence…. once we condition on
(2008)                                  s*8half-      per capita against   reports by       reconstruction      interactions of    community characteristics, we find that
                                        years)        Coalition and        Coalition        projects through    these              greater service provision leads to less
                                                      Iraqi government     forces, MNF-I-   programs, such as                      violence. This violence-reducing effect
                                                      forces)              SIGACTS III      CERP, intended to                      appeared in the second half of 2007,
                                                                           Database         provide local                          when operational changes meant that
                                                                                            public goods,                          Coalition forces nation-wide had a better
                                                                                            Sunni vote share,                      understanding of their communities'
                                                                                            Shia vote share                        needs. In that period, every dollar per
                                                                                                                                   capita of CERP spending predicted 1.9
                                                                                                                                   less violent incidents per 100,000
                                                                                                                                   population." (37)

								
To top