Docstoc

Do regulatory reforms in product and labour markets promote

Document Sample
Do regulatory reforms in product and labour markets promote Powered By Docstoc
					 MAXIMISING POTENTIAL
  IN THE WORKPLACE

A lunchtime seminar series about
    employment relations and
       the world of work


 http://www.dti.gov.uk/er/emar/events.htm
       Purpose of analysis

• Investigate effects of product market
  reforms on employment

• Explore the possible complementarity of
  policies in labour and product markets
    Department of Trade and Industry
       Lunchtime Seminar Series
                 London
              9 March 2005




 Product Market Reforms and
        Employment:
Evidence From OECD countries

         Giuseppe Nicoletti, OECD
       Stefano Scarpetta, World Bank
                Outline
• Competition and employment: theory and
  evidence
• Policy interactions: theory and evidence
• Labour market policies/institutions and
  employment: a reminder
• What’s new with our approach?
• Policies and employment: panel estimates
• Conclusions and future research
Competition and employment: theory
•   Macroeconomic link made surprisingly recently

•   But growing body of research
     –   Krueger and Pischke, 1997; Nickell, 2000; Amable and Gatti, 2001; Pissarides, 2001; Spector, 2002;
         Blanchard and Giavazzi, 2003; Messina, 2003; Ebell and Haefke, 2003


•   Two main channels:
     – Free entry increases equilibrium number of firms (or entrepreneurs)
       eliminates rents, expands long-run output and employment
           •   in both simple imperfect competition/bargaining and search equilibrium models
     – More intensive competition increases labour market tightness and lowers
       unemployment
           •   in search equilibrium models

•   But effect may also depend on other factors:
     – With efficiency wages, reservation wage may rise due to increased risk of lay-
       offs in bad states (Amable and Gatti, 2001)
     – In search equilibrium, precautionary overemployment tends to be reduced
       (Ebell and Haefke, 2004)
     – With bargaining, interaction with labour market policies and institutions (see
       below)
Competition and employment: evidence
•   Relatively little empirical analysis, even less cross-country
•   Proxies for competition (markups, HHI) are few, non-monotone,
    endogenous to market outcomes (and policy irrelevant)
•   If well designed, measures of anticompetitive product market policies
    are monotone, less endogenous (and policy relevant)
     –   Peoples (1998), Bertrand and Kramarz (2002), Kugler and Pica (2004) used single-country sector-specific
         proxies for the stringency of regulation
     –   Boeri et al. (2000), Nicoletti et al. (2001) used cross-country OECD indicators of economy-wide regulation
         for 1998
     –   Messina (2003) uses cross-country proxies of entry costs for 2000 by Djankov et al (2003)
     –   Griffith et al. (2004) uses cross-country indicators by the Fraser Institute available for 1995 and 2000

•   Results generally show a positive effect of competition-friendly
    policies on employment at both sector and macro levels
•   However, existing empirical studies have three main limitations:
     – No studies have looked at the macro effects of a broad range of reforms
       implemented over a long period of time
     – Few studies control for other factors influencing employment and check for
       policy interactions
     – Few robustness and sensitivity tests are usually performed
          Policy interactions: theory (1)
•   Two kinds : “political economy interactions” and “employment outcomes
    interactions”
•   Political complements?
     – By lowering rents, deregulation reduces incentives to fight for capturing them
       (Blanchard-Giavazzi, 2003) or to secure them by increasing insider power (Saint
       Paul, 2002)
     – Free entry makes it harder for firms to bear costs of EPL and reduces incentives to
       protect jobs due to the positive effect of competition on employment (Koeniger and
       Vindigni, 2004)
           Product market reform facilitates labour market reform
     – Positive real wage effects of deregulation are stronger when bargaining power and
       unemployment benefits are modest (Spector, 2003)
           Labour market reform facilitates product market reform

•   Or political substitutes?
     – Employment effects of deregulation may be smaller if rigidities related to efficiency-
       wage mechanisms remain: for given rigidities, firing costs should increase with
       deregulation (Amable and Gatti, 2001)
    Policy interactions: theory (2)
• Sign of interaction in outcomes unsure (little theoretical research)

• Complementary?
    – High rents and high insider compound their negative effects on
      employment  % gains from PM reform higher than with lower insider
      power (also, less “Amable and Gatti” effect)
    – PM reform yields double dividend in the presence of strict LM policies
      (due to political complementarity)  % gains from PM reform higher
      than with easier LM policies (Ebell and Haefke, 2004)
    – Gains from LM reform are lower if startup costs are high (Kugler and
      Piga, 2003)  % gains from LM reform higher with easier PM policies

• Substitutes?
    – Adjustment costs higher with rigid LM  % gains from PM reform lower
      than with easier LM policies
       Policy interactions: evidence

• Nicoletti et al. (1999) highlighted a close correlation between a
  measure of economy-wide regulation and EPL in 1998
• The correlation is robust to changes in the time period and the
  definition of regulation
• It is also somewhat robust to accounting for a wider set of labour
  market policies
• Moreover, there is evidence that lagged PM reforms are positively
  correlated with LM reforms (Burniaux et al. 2005)
• Kugler and Pica (2003) show (a contrario) on Italian data that tight
  entry regulation curbs gains from easing of EPL
• Not aware of other empirical work so far exploring complementarity
  vs. substitutability issue
   Labour market policies/institutions and
       employment rate: a reminder
• High tax wedges
   – Theory and evidence point to negative effect on employment rates
• Generous unemployment insurance
   – In theory negative on unemployment, but ambiguous on participation
   – Evidence mixed, need to define UI in comprehensive way and control for
     likely trade off with EPL
• Strict EPL
   – Conflicting effects at work, but negative if costs are unpredictable and/or
     cost shifting difficult
   – Mixed evidence, need to account for likely interaction with bargaining
• Bargaining systems
   – Evidence tends to support Calmfors-Driffil effect
   – Need to consider jointly with unionisation (though coverage would be better)
• Public employment
   – Crowding out of private employment likely under certain conditions
   – Evidence tends to support this effect, but endogeneity and aggregation are
     crucial issues
     What’s new with our approach?

• Look at the employment effects of regulation over 1980-2002 using
  new time-series proxy for pro-competitive reforms in a large number
  of industries
     Used by Nicoletti-Scarpetta (2003) to study productivity and Alesina et al. (2003)
      to study investment
     Updated to 2003 by OECD

• Control for the main known influences of labour market
  policies/institutions on employment (notably using new time series
  proxy for EPL and bargaining)
     Revised and updated to 2003 by OECD

• Account and test for interactions/complementarities between labour
  and product market policies/institutions
• Look at effects of regulation on shifts in trend of employment rate
• Look at robustness of results to outliers, changes in specification,
  omitted variables, multicollinearity
             The empirical model
• Standard Layard-Nickell-Jackman framework
 ebr      t   egr    lmr   inst    (inst lmr)
    it    0 i i          it
                            k k     kit k k    kit k k            kit
        pmr   ( pmrlmr)    gap 
             it                         it it
                 k k           kit

• accounting for
   –   bargaining and market power (generated by anticompetitive regulations),
   –   Calmfors-Driffill interactions between LM policies and institutions,
   –   Interactions between product and labour market policies, and
   –   country-specific time trends reflecting other factors affecting employment rates
       (preferences, women participation, education, asymmetric effects of productivity
       growth)

• Fixed effects estimation with outlier control
Panel estimates: baseline equation
                                             Baseline equation
                                     Fixed effects estimation, 1980-2002
 Dependent variable                                 Non-agricultural business employment rate
                                                                     with control for            with country
 Independent variables                                                   outliers                time trends

 Lagged public employment rate                       0.047                  -0.031                     0.289
                                                    (0.108)                  (0.1)                    (0.165)*

 EPL                                                -2.068                 -1.822                      -0.753
                                                  (0.398)***             (0.363)***                   (0.419)*

 Corporatism - low                                   1.853                  1.518                      0.435
                                                  (0.545)***             (0.493)***                   (0.599)

 Corporatism - high                                  0.686                   0.703                      1.252
                                                   (0.407)*                (0.373)*                  (0.446)***

 Union density                                      -0.274                 -0.268                      -0.067
                                                  (0.019)***             (0.018)***                  (0.029)**

 Tax wedge                                          -0.147                 -0.104                      -0.255
                                                  (0.034)***             (0.032)***                  (0.033)***

 Unemployment benefits                              -0.146                  -0.12                      -0.123
                                                  (0.032)***             (0.029)***                  (0.027)***

 Output gap                                          0.451                  0.432                       0.419
                                                  (0.039)***             (0.036)***                  (0.026)***

 Constant                                           71.141                 69.533                      57.069
                                                  (1.823)***             (1.684)***                  (2.551)***

 Observations                                        445                     433                        433
 Number of countries                                  20                      20                         20
 R-squared                                           0.64                    0.66                       0.85
 Country effects                                      Yes                     Yes                       Yes
 Country time trends                                  No                      No                        Yes
 Control for outliers                                 No                      Yes                       Yes

 Standard errors in parentheses. * significant at 10%; ** significant at 5%; *** significant at 1%
Panel estimates: with bargaining interactions and regulation
                       Equation with bargaining interactions and regulation
                                   Fixed effects estimation, 1980-2002
          Dependent variable                   Non-agricultural business employment rate
                                             With interactions EPL-
          Independent variables                   bargaining.            With PM regulations


          Lagged public employment rate               -0.145                    -0.074
                                                     (0.103)                   (0.091)

          Corporatism – low                            2.049                     1.682
                                                    (0.503)***                (0.438)***

          Corporatism – high                          0.423                      1.609
                                                     (0.403)                  (0.369)***

          Union density                                -0.26                    -0.176
                                                    (0.018)***                (0.018)***

          Tax wedge                                    -0.11                    -0.151
                                                    (0.031)***                (0.027)***

          Unemployment benefits                       -0.107                    -0.132
                                                    (0.031)***                (0.027)***

          Output gap                                   0.428                     0.403
                                                    (0.035)***                (0.031)***

          EPL in low corporatist                      2.079                      1.636
                                                     (1.058)*                  (0.919)*

          EPL med corporatist                         -4.265                    -2.612
                                                    (0.721)***                (0.641)***

          EPL high corporatist                         3.846                     4.303
                                                    (0.874)***                (0.768)***

          PM regulations                                                          -1.2
                                                                              (0.107)***

          Constant                                    69.726                    67.929
                                                    (1.714)***                (1.487)***

          R-squared                                    0.67                      0.78
          Country time trends                          No                        No
      Summarising LM features and
         testing interactions
• Relationships among variables make identification and
  interpretation of effects difficult (collinearity):
   – LM features (e.g. bargaining and unionisation, EPL and UI)
   – complementarity between LM and PM features (e.g. EPL and PMR)
 To address this we summarise LM policies and institutions
  by means of principal components

• Complementarity between PM and LM policies may make
  effects of regulation depend on LM regime
 To address this we divide countries in three groups
  depending on LM policy regimes (highly restrictive, medium
  restrictive, liberal) and interact these regimes with our
  measure of product market regulation
Panel estimates: synthetic LM features and PM/LM interactions
              Equation with synthetic labour market policies/institutions and regulation
                                            Fixed effects estimation, 1980-2002
Dependent variable                                         Non-agricultural business employment rate
                                         Without country trends                          With country trends
Independent variables               With synthetic        With LM institutions/      With synthetic               With LM             With PM reg.
                                   indicators of LM        policies & PM reg.       indicators of LM       institutions/policies & interacted with LM
                                  institutions/policies                           institutions/ policies           PM reg.               policies

Lagged public employment rate           -0.159                  -0.315                  0.276                     0.48                  0.505
                                       (0.102)                (0.087)***               (0.157)*                (0.168)***            (0.164)***

Output gap                              0.581                    0.54                   0.469                     0.469                 0.454
                                     (0.045)***               (0.038)***             (0.026)***                (0.026)***            (0.025)***

LM institutions                         -2.01                   -0.824                   -0.66                   -0.808                -0.788
                                     (0.356)***               (0.307)***               (0.390)*                (0.388)**             (0.377)**

LM policies                            -4.226                   -2.047                 -3.812                    -3.118                -1.921
                                     (0.614)***               (0.540)***             (0.545)***                (0.581)***            (0.616)***

PM regulations                                                  -1.341                                           -0.769
                                                              (0.101)***                                       (0.241)***

PM reg. with strict LM policies                                                                                                        -1.009
                                                                                                                                     (0.241)***

PM reg. with med. LM policies                                                                                                          -0.747
                                                                                                                                     (0.246)***

PM reg. with lax LM policies                                                                                                            -0.346
                                                                                                                                       (0.249)

Constant                               50.668                   57.711                 42.683                     44.56                44.065
                                     (1.244)***               (1.167)***             (1.874)***                (1.944)***            (1.891)***

R-squared                                0.43                     0.6                     0.84                     0.84                  0.85
Country time trends                       No                      No                      Yes                      Yes                   Yes
Panel estimates: sensitivity to public employment
                             Sensitivity to public employment
                              Fixed effects estimation, 1980-2002
   Dependent variable                        Non-agricultural business employment rate
   Independent variables                   Without pub.       Without pub. employment &
                                           employment                interactions

   Output gap                                  0.466                      0.451
                                             (0.026)***                 (0.026)***

   LM institutions                             -0.903                     -0.885
                                             (0.390)**                  (0.380)**

   LM policy                                   -3.249                     -2.096
                                             (0.584)***                 (0.620)***

   PM regulations                              -0.509
                                             (0.225)**
   PM reg. with strict LM policies                                        -0.724
                                                                        (0.225)***


   PM reg. with med. LM policies                                          -0.456
                                                                        (0.229)**

   PM reg. with lax LM policies                                           -0.086
                                                                         (0.237)

   Constant                                    48.803                     48.497
                                             (1.264)***                 (1.240)***
   R-squared                                    0.84                       0.85
   Country time trends                          Yes                        Yes
  Are the estimates reasonable?
• Given the sample variation of the summary policy
  indicators, the estimates suggest that product market
  reforms:
   – have around half the impact on employment rates of labour market
     reforms
   – have a larger impact than changes in LM institutions (perhaps
     because changes in institutions are indirectly caused by PM
     reforms)
   – over the past two decades have contributed to increase
     employment rates by (at most) around 3 per cent in certain
     countries (e.g. the UK)
   – given policy interactions, employment rates could increase by 2
     per cent in Italy if further PM reforms were to align with current
     best practice
• These are lower bound estimates because they only take
  into account the effects of reforms in non-manufacturing
                     Conclusions
• Evidence broadly confirms previous results on effects of
  LM policies and institutions
• Evidence supports recent theoretical insights on
  regulation/employment link
• There is also evidence of political complementarity
  between LM and PM policies
• Overall, estimates seem reasonable and suggest that
  policy interactions are worth of consideration
• Three main policy implications:
   – PM reforms can be useful complement to, although not substitute
     for, LM reforms
   – In restrictive countries, reforms in both areas are needed to
     significantly raise labour utilisation
   – Bang-for-the-buck of PM reforms higher in countries that have rigid
     labour markets
     Employment rate and economy-wide
             regulation, 1998
.8




                                              DNK                             NOR
                                                            JPN
.7




                                                      SWE
                     GBR                      AUT
                                                    CAN
                                       USA
                                       AUS                    DEU


                                                            PRT
.6




                                                                        FIN
                                                            BEL         FRA
                                               IRL    NLD
                                                NZL                             ITA
.5




                                                                  ESP
           Correlation coefficient = -0.4
                                                                              GRC
           P-value = 0.08
.4




      .5                    1                1.5               2              2.5
                                Product market regulation, summary index
Employment rate and non-manufacturing
 .8
          regulation, 1998


                                        NOR     DNK
                                               JPN
 .7




                                 SWE
               GBR                                AUT
                                   CAN
                       USA
                         AUS             DEU


                                                             PRT
 .6




                                         FIN
                                                     BEL   FRA
                                              NLD            IRL
                       NZL                                         ITA
 .5




                                                    ESP
      Correlation coefficient = -0.55
                                                                             GRC
      P-value = 0.01
 .4




           1               2             3              4                5
                        Product market regulation, time-series index
Employment rate and non-manufacturing
 .8
       regulation, 1980-2002


                                            SWE

                                                             DNK
 .7




                                                NOR     AUT
                                                 JPN
                           GBR
                                    CAN
                                                   DEU
                                          AUS
                   USA                                 FIN
 .6




                                                                     FRA
                                   NZL                         BEL      PRT
                                                                          ITA

                                                             NLD
 .5




      Correlation coefficient = -0.52                                IRL

      P-value = 0.02                                    ESP
                                                                           GRC
 .4




            2                   3                4                 5             6
                        Product market regulation, time-series index
     and economy-wide regulation in
EPL Figure 4. Product market regulation and employment protection legislation 1998                                1


 Employment protection legislation
  4.0

              Correlation 0.73                                                             Portugal
               t-statistic 4.72
  3.5                                                                                                            Greece
                                                          rd
                                                      3 cluster                                                   Italy
                                                                                       Spain
                                                                                                        France
  3.0
                                                                                                  Norway
                                                                            Germany
                                                                               Japan
  2.5                                                          Sweden
                                                                          Austria
                                                                         Netherlands
                                nd
                            2        cluster
                                                                            Finland              Belgium
  2.0



  1.5                                                                      Denmark
                                                                              Switzerland
                                                   Australia
  1.0                                  Ireland
                                                           New Zealand

                                                                  Canada
  0.5                             United Kingdom                                          st
                                                                                         1 cluster
                                                      United States

  0.0
        0.0                 0.5                     1.0                     1.5                   2.0                     2.5
                                                                                                 Product market regulation
EPL and non-manufacturing regulation
4           1980-2002
                                                                     PR T
        correlation = 0.72
        p-value = 0.0003                                                    GRC
                                              ESP
                                                                        ITA
3




                                   SWE    D EU          BEL   FRA
                                     N OR
                                                     N LD

                                              AU T
                                             FIN
                                       JPN
2




                                                     DNK




                        N ZL                     CHE
1




                                AU S                          IR L

                GBR      C AN

        U SA
0




    2                3                  4                 5                       6
                 Product m arket regulation (non-m anufacturing)
                  U SA
       Labour market policies and non-
      0
             Produc t m arket regulation (non-m 1980-2002
      manufacturing regulation anufac turing)
      1
       2         3                   4             5                                            6



                                              SWE
                                                                     BEL             ITA

                                                               N LD        FR A
                                                            ESP
                                                         D EU                     PR T
.5




                                                           FIN
                                                                                         GR C
                                                                   DNK

                                                  N OR
      0




                                                            AU T
              correlation = 0.65
              p-value = 0.002
-.5




                                                                           IR L


                           GBR                      JPN
                                   N ZL                      CHE
-1




                  U SA              C AN
                                           AU S

          2                     3                   4                  5                        6
                            Produc t m arket regulation (non-m anufac turing)
  Principal components
           Principal components: Scoring Coefficients
                       (based on rotated factors)


                 Labour                                     Labour
                 market                                     market
Variable         institutions             Variable          policies

                 weights                                     weights
coordination         0.26                 tax wedge             0.47
centralisation       0.54                 benefit generosity    0.10
union density        0.18                 epl                   0.38

				
DOCUMENT INFO