Economic Growth in the 1990s Learning from a Decade

Document Sample
Economic Growth in the 1990s Learning from a Decade Powered By Docstoc
					Economic Growth in the 1990s                                                                                                   Page 1 of 2
                                                                                             Contact Us • Help/FAQ • Index • Search




    Home > Economic Growth in the 1990s


                              Economic Growth in the 1990s: Learning from a Decade of
                              Reform
                                                           The World Bank's Poverty Reduction and Economic Management (PREM)
                                                           Network has prepared a study on development lessons of the 1990s. The
                                                           report reviews the growth impact of the main policy and institutional reforms
                                                           introduced in the 1990s, presents a broad perspective on the events, country
                                                           experiences, academic research and controversies of the decade, and
                                                           reflects on how they alter our thinking about economic growth. Economic
                                                           Growth in the 1990s: Learning from a Decade of Reform can be purchased
                                                           from our online bookstore.

                                                          The report complements a series of lectures by leading development
                                                          practitioners such as Larry Summers, President of Harvard University and
                                                          Former U.S. Secretary of the Treasury, and Fernando Henrique Cardoso,
                                                          Former President and Former Minister of Finance of Brazil, discussed their
                              experience as policy makers at the forefront of policy implementation in the 1990s.
                                      Table of Contents
                                      Foreword
                                      Gobind Nankani
                                      1.   Overview
                                      PART 1: FACTS OF THE 1990s
                                      2.   Grist and Mill for the Lessons of the 1990s

                                           Country Note: Economic Growth from the Very Long-Term Perspective of History
                                      3.   Something Special about the 1990s?

                                           Country Note: Lessons from Countries That Have Sustained Their Growth
                                      PART 2: DEVELOPMENT CONTROVERSIES OF THE 1990s
                                      4.   Macroeconomic stability: the More the Better?

                                           Country Note: Poverty and Inequality: What Have We Learned in the 1990s
                                      5.   Trade Liberalization: Why So Much Controversy?

                                           Country Note: The Middle East and North Africa: Performing below Potential
                                      6.   Privatization and deregulation: a push too far?

                                           Country Note: Eastern Europe's Transition: Building Institutions
                                      7.   Financial Liberalization: What Went Right, What Went Wrong?

                                           Country Note: Lessons and Controversies from Financial Crises in the 1990s
                                      8.   Policy Reforms and Growth Performance and Growth: What Have We Learned?

                                           Country Note: Why Blessings become Curses?
                                      9.   Improving Public Sector Governance: The Grand Challenge?

                                           Country Note: Africa's Growth Tragedy: an Institutional Perspective


                                      10. Politics: Does Democracy Help?

                                           Bibliography



http://www1.worldbank.org/prem/lessons1990s/                                                                                   11/27/2006
Overview from Economic Growth in the 1900s: Learning from a Decade of Reform, World Bank, April 2005




           Chapter 1



Overview



          E
                 CONOMIC GROWTH IS A RECENT              policy and institutional reforms. Section 5 sketches
                  event in the history of humanity.      operational implications. Subsequent chapters set
                  During most of 4 million years of      out the facts about growth in more detail, and then
evolution, people made limited economic progress         examine the main areas in which economic and
and their material well-being changed very little. In    institutional reforms concentrated during the
the last few centuries, however, goods and services      1990s—macroeconomic stabilization, trade, finan-
started to be produced at increasingly lower cost in     cial sector, privatization and deregulation, modern-
hours of effort.The hours of work needed to pro-         ization of the public sector, and political reforms.
duce basic goods such as water or heat at the dawn       The chapters aim to draw lessons from gaps
of civilization were several hundred times those         between expectations and outcomes. Most chapters
needed today (DeLong 2000). Similar increases in         are also followed by a Country Note that expands
productivity have been achieved for an expanding         on issues insufficiently dealt with in that chapter, or
range of goods and services. Most of this progress       that considers country-level perspectives.
has taken place in the last two centuries, during
which technological progress has been exception-         FIGURE 11
ally rapid, and economic growth unprecedented
                                                         Worldwide Growth in Real GDP per Capita, 1000–Present
(figure 1.1).
    It is only in the last 50 years that mainstream                                                100%
                                                             Century's growth in real world GDP




economics has focused on the determinants of                                                        90%
Adam Smith’s “natural progress of opulence” and                                                     80%
on how growth could be accelerated. Many ques-                                                      70%
tions about growth still lack satisfactory answers.Yet                                              60%
                                                                         per capita




few issues are more important for the world’s future                                                50%
than the ability of developing countries to raise                                                   40%
both productivity and the rate at which they accu-                                                  30%
mulate capital.                                                                                     20%
    This overview chapter first briefly reviews our                                                 10%
understanding of growth before turning, in section                                                   0%
2, to the facts and controversies of growth and pol-                                              –100% 11th 12th 13th 14th 15th 16th 17th 18th 19th 20th
icy reforms in the 1990s. Section 3 draws the broad                                                                       Century
lessons coming out of the growth experience of the       Source: DeLong 2000.
1990s, and section 4 offers lessons specific to key

                                                                                                                                                   1
2                                                                    E C O N O M I C G ROW T H I N T H E 1 9 9 0 s



1. Understanding Economic                                      Up to then, thinking about growth had been
                                                           dominated by the Solow model, the basic model
   Growth
                                                           with which we still think about economic growth,
Absent definitive theories, views on growth have           in which growth is a function of the accumulation
been shaped by facts and changed by experience.            of capital, accumulation of labor, and productivity
Until the 1970s, the growth strategies of developing       growth.This model leaves out much of what needs
countries focused on accelerating the rate of capital      explaining. In particular, it views long-run growth
accumulation and technological adoption. Import            as entirely determined by exogenous factors, inde-
substitution, state-owned enterprises, controls over       pendent from structural characteristics of the econ-
the financial sector, central planning, and a variety of   omy such as openness, scale, and saving rate, and,
price controls and state interventions in the econ-        most important, from the policies influencing such
omy were some of the policies that governments             variables.Also, while left unexplained in the model,
used to take the “commanding heights” of the econ-         productivity growth drives the empirical story.
omy and guide resource allocation to areas thought         Solow himself estimated that technological change
to be most conducive to long-term growth. Confi-           explained more than half of per capita output
dence in governments was born from their (partial)         growth in the first half of the 1900s in the United
success in addressing the Great Depression, in             States. Calculations by the World Bank indicate that
expanding production during World War II, and              it explained one-third of the increase in per capita
reconstructing Europe and Japan. Economists and            income in East Asia up to the early 1990s (World
policy makers saw that market forces disrupted             Bank 1993). Other exercises reach similar conclu-
growth and that governments were able to restore it,       sions on the large role of productivity gains in
and to expand capacity efficiently.The generation of       growth experiences.
economists that followed, however, familiar with               At first the New Growth Theory seemed to hold
experiences of developing countries in the 1970s and       the promise of linking policies to growth perform-
1980s, saw the waste of enormous resources in ill-         ance. It appeared at a time when evidence was accu-
conceived government initiatives, the costs of poor        mulating—from the growth experience of the 1970s
macroeconomic management, and the ease with                and 1980s—suggesting that the accumulation of cap-
which well-intentioned public policies could be            ital was not a panacea, and that misguided policies
diverted to serve narrow political or economic inter-      were costly for growth.The new evidence provided
ests. Understandably, this later generation of econo-      the conceptual foundation for aggregate cross-coun-
mists and policy makers came to believe that the cost      try regressions, which throughout the 1990s sought
of government failures was considerably larger than        to capture the effect of policies on long-term growth
the cost of market failures, that government inter-        (Barro 1991; Temple 1999) and provided the
ventions interfered with development, and that con-        strongest intellectual foundation for the view that
taining the role of the public sector in the economy,      better policies would deliver faster growth.
reducing its use of resources, and limiting its discre-        A number of empirical problems became evi-
tion were essential for economic growth.                   dent, however, related to the crude manner in
                                                           which policy variables enter the cross-country
                                                           regressions; the fact that differences in the institu-
New Growth Theory                                          tions underlying policy design and policy imple-
This shift in views was supported by a new strand of       mentation are not captured; the lack of robustness
academic research that started in the second half of       to changes in time periods and specifications; the
the 1980s and gathered impetus during the 1990s,           crudeness of the assumption that the same model
when there was a resurgence of academic and                explaining growth in the Republic of Korea or
empirical work on growth.                                  Brazil could be used for Bolivia or Rwanda; and the
I N T RO D U C T I O N                                                                                                                  3



poor predictive power of policies as indicators of                       Nonetheless, in the process, greater clarity was
performance.                                                             reached on the facts about growth, analysts paid
    If, as suggested by the growth regressions, poli-                    greater attention to the role of institutions,and stud-
cies matter for growth, policy improvements should                       ies brought the issue of inequality—both within
lead to higher growth. Both in the 1980s and 1990s,                      and between countries—increasingly to the fore.
policies improved relative to other decades, but
growth performance remained well below that of
the 1960s and 1970s (Easterly 2001). More recently,                      Growth in Developing Countries:
empirical research has argued that when a measure                        Divergence,Variability, and Unpredictability
of “institutional quality” is included in cross-country                  Research during the 1990s was able to extend the
regressions, the explanatory power of other vari-                        availability of data over long periods.This made it
ables, including all measures of “policies,” becomes                     clear that growth was not a linear process, and that
negligible (Acemoglu, Johnson, and Robinson                              it did not conform to the theoretical prediction
2001; Rodrik, Subramanian, and Trebbi 2002; East-                        that per capita income in developing countries
erly and Levine 2003; and IMF 2003e).This suggests                       would eventually converge with that of industrial-
that “good” institutions matter more for growth                          ized countries. In fact, there has been “divergence
than “good” policies—that “institutions rule.”                           big time” in the evolution of per capita incomes
    In hindsight, the breakthroughs expected from                        (Pritchett 1997), both between industrialized and
the New Growth Theory have not materialized.                             developing countries and among developing coun-

FIGURE 12
Economic Growth in Perspective, 1960–2002
20%




15%




10%




 5%




 0%
      19         19         19        19         19        19           19      19         19         19        19          19     19       19   20
      60         63         66        69         72        75           78      81         84         87        90          93     96       99   02

                                              Latin America and the Caribbean                       Sub-Saharan Africa
                                              South Asia                                            Middle East and North Africa
                                              Europe and Central Asia                               East Asia and Pacific

Source: WDI 2003.

Note: Regions’ GDP per capita is shown as a percentage of the OECD GDP per capita (total regional GDP over total population).
                                            4                                                                                                  E C O N O M I C G ROW T H I N T H E 1 9 9 0 s



                                            tries themselves.This is the case whether the period
                                            being considered is the last 40 years (figure 1.2) or     FIGURE 1.4

                                            the last 10 (figure 1.3).                                 Fraction of World Inequality Accounted for
                                                As a result, worldwide inequality has changed         by Differences across Countries
                                            from being the result almost exclusively of differences




                                                                                                      Fraction of total inequality due to
                                                                                                                                            70.0%
                                            among people within countries to being the result pri-




                                                                                                         differences across countries
                                                                                                                                            60.0%
                                            marily of differences across countries (figure 1.4).
                                                The consideration of growth over longer periods                                             50.0%
                                            also highlights the variability of growth in developing                                         40.0%
                                            countries. The experience of Latin America since
                                                                                                                                            30.0%
                                            the 1980s, the collapse of growth in Africa in the last
                                            two decades, and the economic collapse of Eastern                                               20.0%

                                            Europe after several decades of sustained growth                                                10.0%
                                            stand in sharp contrast to the stability of growth                                              0.0%
                                            among industrialized countries, which have grown




                                                                                                                                                1800
                                                                                                                                                       1820
                                                                                                                                                                1840
                                                                                                                                                                       1860
                                                                                                                                                                              1880
                                                                                                                                                                                      1900
                                                                                                                                                                                             1920
                                                                                                                                                                                                    1940
                                                                                                                                                                                                           1960
                                                                                                                                                                                                                  1980
                                                                                                                                                                                                                         2000
                                            at roughly a constant rate (except for the interrup-
                                            tion of World War II and recovery years) for more                                                                                        Years
                                            than 100 years. It also contrasts with the experience                                                             Theil coefficient                 Mean ln deviation

                                            of East Asian countries. What is remarkable about         Source: Source: Bourguignon and Morrison 2002.
                                            East Asia is not that it experienced a crisis in 1997,
                                            but that it experienced so few crises over the pre-
                                            ceding decades. By and large, developing countries        half that rate. Korea has only had only three years of
                                            have one year of negative per capita growth roughly       negative per capita growth since 1961.1
                                            once every three years. In East Asia, the average is          The variability of growth helps to explain why
                                                                                                      growth in the developing world is so difficult to
                                                                                                      predict. Instances of economists (including, for
                                                                                                      example, 1977 Nobel Laureate James Meade on
FIGURE 1.3                                                                                            Mauritius) making highly inaccurate predictions
Regional Perspectives on Growth in the 1990s                                                          have become part of the economic folklore. Many
                                                                                                      of the economic successes of today—Bangladesh,
                              200
                                                                                                      Indonesia, Korea, or Mauritius—were considered
 Per capital GDP (1990=100)




                              180                                                                     “basket cases” in the 1960s, when Africa’s growth
                              160                                                                     prospects were seen as superior to those of over-
                              140                                                                     populated Asia—a view captured in Asian Drama
                                                                                                      (Myrdal 1972). In the later 1990s, just before the
                              120
                                                                                                      second most dramatic economic crisis in its history,
                              100                                                                     Argentina was seen as a model for developing coun-
                               80                                                                     tries and believed to have found the path to sus-
                               60                                                                     tained growth. At a more technical level, World
                                    1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000            Bank growth projections, as well as growth projec-
                                    East Asia and Pacific                   South Asia                tions by other forecasters, tend to be systematically
                                    Eastern Europe and Central Asia         Sub-Saharan Africa
                                    Latin America and the Caribbean                                   overoptimistic (a point that was highlighted in the
                                                                            Organisation for
                                    Middle East and North Africa            Economic Co-operation     World Bank’s World Development Report 1991).
Source: WDI 2003.                                                                                         While a rare occurrence thus far, sustained
                                                                                                      growth has improved the lives of millions. Coun-
I N T RO D U C T I O N                                                                                            5



tries where sustained growth has taken place                 larly in capital-intensive industries, because old eco-
(mostly in South and East Asia, including                    nomic interests resisted change and were unwilling
Bangladesh, China, Indonesia, India, and Vietnam)            to take on risks inherent in new industrial activities;
account for a large proportion of world population.          price controls did not have serious economic con-
Out of 117 countries with populations of more                sequences because the concentration of wealth pre-
than half a million people, only 18 have been able           cluded the redeployment of resources in response
to sustain growth rates exceeding industrialized             to changes in demand (Seers 1962).
countries’ growth and hence narrow their per capita              While there are some functions that institutions
income gap with those countries.2                            need to perform in any society, the form through
                                                             which institutions can perform these functions can
                                                             vary considerably (Virmani 2004). Most of the
Institutions                                                 empirical work on the importance of institutions
Defined as the rules and norms constraining human            leaves open the question of how to improve institu-
behavior (North 1990), institutions include the              tional performance. Merely adopting some other
informal rules and norms that govern personal and            country’s laws and formal regulations is no guaran-
social behavior and the formal rules and norms               tee of achieving the same institutional performance.
governing economic, social, and political life. Insti-       Recently, accession to the World Trade Organiza-
tutions enable societies to organize themselves and          tion and integration into regional supra-national
function in an orderly manner by solving problems            entities such as the European Union and the New
central to life in society, particularly agency prob-        Economic Partnership for Africa have strengthened
lems, containment of predation by individuals or             incentives for institutional improvements. East Asian
the state, and collective decision making. Societies’        countries have long realized the importance of
performance depends on how effectively their insti-          institutional change and innovation, and the 1997
tutions resolve these problems.                              crisis made this realization all the more acute, creat-
    The importance of institutions for economic              ing renewed impetus to modernize institutions,
prosperity is not a novelty learned from the 1990s.          including political institutions. But for most devel-
From different perspectives, Adam Smith, Karl                oping countries, improving the quality of their
Marx, and Max Weber highlighted the role of insti-           institutions remains a challenge.
tutions in the development of a market economy
and formation of a capitalist society. Economists
dealing with development in the 1950s and 1960s              Fairness, Growth, and Institutions
were aware that the development challenges faced             Another important strain of ideas in the 1990s came
by a plantation economy differed from those faced            from the resurgence of interest in inequality as an
by a society where economic and political power              apparent influence on growth and institutional per-
were not concentrated (Rostow 1952, 1960; Adel-              formance. A recent body of literature suggests sev-
man and Morris 1965). Latin American economists              eral channels through which inequality affects
of the Structuralist school saw in the legacy of colo-       economic growth. Fairer societies offer their citi-
nialism, embedded in institutions serving the inter-         zens more public goods, more social support, and
ests of a small, landed elite, the source for economic       more social capital. Hence they are more capable of
performance inferior to that of the United States or         sharing the costs and benefits of improving eco-
Canada (Furtado 1963). This imbalance formed a               nomic policies, and in turn facilitating consensus
part of the justification for an activist state: inflation   building and decision making (Deaton 2003a). Fair-
helped to mobilize resources from the wealthy elite          ness also facilitates agreement on the provision of
who resisted more efficient forms of taxation; states        public goods that have strong beneficial side effects
sponsored investments in manufacturing, particu-             on society, such as health services, water supply, or
6                                                                    E C O N O M I C G ROW T H I N T H E 1 9 9 0 s



waste disposal. Other channels through which              States in the early 1900s, when the government
inequality affects growth are market structures and       decided to regulate matters hitherto left to private
microeconomic incentives. A better distribution of        parties and the courts; the reason for the shift was a
wealth reduces credit constraints, and broader avail-     perception that judges and the courts, having been
ability of credit is found to have a significant and      corrupted by powerful economic interests, were
positive effect on growth rates. If individuals are       unable to render fair and equitable judgments. World
limited in their borrowing capacity, reallocating         Development Report 2001 provides other examples of
capital toward the poorest will increase aggregate        how economic incentives affect the emergence of
productivity.                                             institutions that sustain the functioning of markets,
    Even if one concludes that greater equality           and the different coordination or risk-reducing
influences growth positively, there is still consider-    problems that they are meant to resolve.
able ignorance about the means through which
greater equality can be achieved. Governments have
long sought, with varied degrees of success, to redis-    2. Facts and Controversies of the
tribute income through land redistribution,                  1990s
employment programs, subsidies, and promotion of
broad access to credit, infrastructure, health, and       At the beginning of the 1990s, most economists
education.The large, underresearched area for fur-        working on development and many policy makers
ther study includes questions related to the impact       shared the conviction that more efficient use of
of public spending on equity, both in a static sense      resources would lead to growth.This was believed
(incidence of public spending) and a dynamic sense        to require, first, macroeconomic prudence, domes-
(changes in individuals’ earnings potential).             tic liberalization, and outward orientation, which in
    Recent literature has emphasized the important        turn required freeing market incentives and open-
links between the distribution of assets in a society     ing the economy. Hence fiscal deficit reduction,
and the institutions that emerge. Knowledge is still      realignment of exchange rates to eliminate black
rudimentary about how institutions emerge and are         market premia, lifting controls on prices, deregula-
established in a society, but economic research in the    tion of interest rates and liberalization of the finan-
1990s has provided some insights. First, economic         cial sector, and reduction of tariffs and other
incentives influence what type of institutions            restrictions on imports all became central to the
emerge and when. The enforcement of property              policy reform programs implemented in the 1990s.
rights to land, for example, will depend on the ben-          Second, conventional wisdom held that to
efits of enforcement relative to its costs, which for     achieve greater efficiency required a reduction in
each owner depends on the extent to which other           the role of the state. There was evidence that the
owners enforce their property rights. In an extrac-       state discretion that was inherent in growth strate-
tive economy, for example if landowners in general        gies based on infant industry, import substitution
do not enforce their property rights, it is uneco-        policies, and the growth of public enterprises had
nomical for one landowner to enforce his: workers         been misused more often than anticipated, had
will find it attractive to exploit land and appropriate   often been captured by narrow interest groups, and
the rents for themselves. Only when this coordina-        served as the source of endemic corruption.
tion problem is resolved will economic incentives         Addressing this problem required reducing state
be sufficient for enforcement of property rights          discretion, downsizing governments, and encourag-
(Hoff and Stiglitz 2001). Second, concentrated eco-       ing a much greater role for the private sector. Hence
nomic and political interests influence institutions.     privatization, deregulation, elimination of quantita-
This can be seen from experiences with land distri-       tive restrictions and of licensing requirements, and
bution in Latin America, and also from the United         dismantling agricultural marketing boards and
I N T RO D U C T I O N                                                                                       7



other forms of state monopoly all became central          reducing hyperinflation and reversing several
to reform programs. Seeing the need to strengthen         decades of state-led import-substituting industrial-
the organizational effectiveness of the state, and the    ization. In Argentina, President Menem set the
efficiency with which the state used public               country on a course of eliminating hyperinflation
resources, reformers rationalized government func-        through a currency board as well as ambitious mar-
tions and undertook civil service, legal, and budget      ket reforms, which saw the privatization of state-
reforms. Democratic processes were expected to            owned businesses and liberalization and opening of
provide checks and balances and further incentives        the economy. In Bolivia, reforms by Paz Estensoro
to this process.                                          that had brought hyperinflation to a halt in the
    Third, it was believed, reforms had to be rapid.      mid-1980s were continued in the 1990s, regardless
Earlier, some of the first authors to argue in favor of   of the parties in government. In Africa, the devalu-
abandoning the dirigiste framework of early devel-        ation of the African Financial Community (CFA)
opment economics (notably Little, Scitovsky, and          franc increased competitiveness and many other
Scott 1970; McKinnon 1973) had argued explicitly          reforms were implemented throughout the region.
in favor of a gradualist reform strategy (in respect to   In Tanzania, President Mkapa started an ambitious
trade and the financial sector, respectively).3 But in    program of reforms. In South Africa, the transition
the course of the 1980s the economics profession          to a multiracial democracy was followed by steps
began to be influenced by the enthusiasm of lead-         toward liberalizing the economy.
ing politicians for “the magic of the market.”Argu-           Leaders such as Rawlings of Ghana and Musev-
ments in favor of “big bang” and “shock treatment”        eni of Uganda strengthened fiscal fundamentals,
became prominent. By the time that the transition         achieved macroeconomic stability, liberalized the
to a market economy got under way in the former           economy, and reduced the role of the state. Privati-
socialist economies, “a belief in gradualism had          zation, retrenchment of the public sector, and liber-
almost become tantamount to a confession of a lack        alization of trade were the focus of economic policy
of reforming virility” (Williamson and Zagha              changes in countries as diverse as the Central
2002).                                                    African Republic, Ghana, and Tanzania. Reforms in
                                                          the Middle East and North Africa were less ambi-
                                                          tious but were nonetheless significant in the Arab
A Decade of Significant Change                            Republic of Egypt, Jordan, Morocco, and Tunisia.
The 1990s provided ample opportunity for these            On the political front, democracy spread in former
views to be implemented.The Russian Federation,           communist countries and Africa, and was consoli-
Eastern Europe, and Central Asia embraced capital-        dated in Latin America. These and other changes
ism and a new generation of leaders made it a pri-        gave rise to expectations that the 1990s would
ority to rebuild their economies on the basis of          accelerate growth and social progress in the devel-
capitalist principles, markets, and privatized firms.     oping world.
Regarding the speed of reform, while there were
divergences, the balance of opinions supported            Rapid Growth, Take-offs, and Social Progress
rapid rather than gradual reform. China, the largest      India and China, together accounting for 40 per-
developing economy, continued the reforms it had          cent of the developing world’s population, grew fast
begun in 1978 with further liberalization of the          in the 1990s for a second decade in a row, as did
domestic economy, and increased openness. After           many other countries in South and East Asia,
its crisis in 1991, India, the second-largest develop-    including Bangladesh, Sri Lanka, and Vietnam.
ing economy, speeded up liberalization started in         Chile continued to grow in Latin America,Tunisia
the 1980s. President Collor of Brazil announced a         in North Africa, and Botswana and Mauritius in
radical program of economic reform aimed at               Africa. New high performers appeared and annual
8                                                                   E C O N O M I C G ROW T H I N T H E 1 9 9 0 s



gross domestic product (GDP) growth for the              positive developments occurred in the later 1990s,
decade was rapid in an array of countries: Mozam-        such as in Mozambique,Tanzania, and Uganda, and
bique (7.8); Uganda (6.8), Dominican Republic            still persist at the time of writing, it is too early to
(6.0), Tunisia (5.0), and Poland (4.5). Countries        conclude that Africa has turned the corner.
affected by the crisis in East Asia made an unex-
pectedly rapid recovery.                                 Financial Crises
    Notwithstanding unevenness across regions, the       Financial crises in the 1990s were less predictable
incidence of poverty continued to decline through-       than in the 1970s and 1980s. Macroeconomists,
out the 1990s—more rapidly in East Asia than in          bank restructuring experts, and emerging-market
South Asia and more rapidly in South Asia than in        private traders rolled from crisis to crisis: from Mex-
Latin America. In Africa, however, the incidence of      ico during 1994–95, to Korea, Malaysia, Thailand,
poverty increased slightly. Growth was the main          Indonesia during 1997–98, Russia in 1998, Brazil
force behind virtually all cases of significant reduc-   in 1998, and Turkey in 2001 to the latest and per-
tions of poverty, including in China and India. But      haps most worrisome of all, in Argentina during
particularly in Latin America, there were instances      2001–02. The evolution of spreads in the months
such as Brazil and Bolivia where social indicators       preceding the financial crises suggests that few were
improved without significant growth.                     anticipated.
    Alongside these positive developments, how-
ever, there were several negative surprises.             Delay in Recovering Growth, Particularly in
                                                         Latin America
“Transition Recession” in the Former Soviet              It was hoped that the “lost decade” of the 1980s
Union and Eastern Europe                                 would be reclaimed in the 1990s. Macroeconomic
The transition from a communist, centrally planned       stabilization, fiscal austerity, trade liberalization, and
economy to a capitalist one was expected to be dif-      privatization were expected to lead to rapid growth.
ficult. But the depth of the output collapse was not     Although growth was the fastest in two decades
widely predicted. The length of the transition—in        until 1998, its collapse thereafter following the
which many countries in 2003, more than a decade         reversal in capital flows created the general percep-
later, remain far below their previous levels of out-    tion that the growth payoffs have been smaller than
put—was not widely forecast. Nor was the variabil-       expected.
ity among countries in the depth and duration of
the output collapse.Though recoveries have started       Argentina: The Collapse of the Hard Peg
to emerge—in the Czech Republic, Hungary, and            Argentina was the most successful example of a
Poland, for example—it will take years, and in some      trend in the 1990s to create macroeconomic stabil-
cases, decades, for most former Soviet countries to      ity by legal and institutional changes intended to
regain their per capita income levels prevailing at      reduce the scope and latitude of government’s dis-
the beginning of the transition.                         cretion. Exchange-rate arrangements that set a fixed
                                                         rate for peso convertibility were not only incorpo-
Continued Stagnation in Sub-Saharan Africa               rated into law but also made especially difficult to
The failure of growth in Africa—either of powerful       alter, and changes were made in the operation of
and rapid growth in a single large country or in a       the central bank to make these limitations a reality.
substantial number of smaller ones—was a surprise.       As part of a package of reforms, this convertibility
Despite good policy reforms, debt relief, continued      plan eliminated Argentina’s hyperinflation and, for a
high levels of official assistance, promising develop-   period, it restored economic growth.
ments in governance, and a relatively supportive             Once Argentina achieved stability with growth,
external climate, no take-off has ensued.While some      there was considerable discussion—particularly
I N T RO D U C T I O N                                                                                              9



after the devaluation of Brazil’s real in January                It has been suggested that lower OECD growth
1999—as to whether the country should abandon                might have depressed developing countries’ growth
its rigid exchange rate system. Views diverged               in the 1990s (Easterly 2002). In reality, the 1990s
among economists. Looking back, the former Gov-              was favorable for developing countries, even if not
ernor of Argentina’s Central Bank described the              every country found ways to benefit. Exports from
abandonment as a marriage to be broken when it               developing countries as a group grew much faster
was going well (Mario Blejer,World Bank 2005b).              than in previous decades. Real interest rates were
For fear that markets could overreact, Argentina’s           lower. Debt obligations claimed fewer resources,
authorities maintained the system. When the plan             and foreign direct investment and financial flows to
collapsed, the result was politically and economi-           developing countries were much larger. If com-
cally costly by design. Thus the damage was not a            modity prices affected developing countries
surprise. But the demise of the convertibility plan          adversely, the damage was not dramatic and should
itself was a surprise, for two reasons. First, its initial   have been offset by the increasing share of manu-
successes had suggested longevity was possible; it           facturing exports—except in a small group of least
had reduced rapid inflation and initiated a boom in          developed and Sub-Saharan African countries that
the early 1990s, and it had weathered the “Tequila”          remained highly dependent on agricultural exports.
after-shocks of the Mexican crisis reasonably well.              All in all, while external factors played a role,
Second, while the end of convertibility was costly           explanations of performance must be sought pri-
by design, its actual cost exceeded the most pes-            marily in developing countries’ domestic policies.
simistic forecast.                                               Good performance has been associated with
                                                             domestic and external liberalization; Chile, India,
                                                             China, and other countries in East Asia are all more
Interpreting the Results                                     open than in previous decades and have moved
From a growth perspective, the net result of the             toward greater reliance on market forces. But many
contrasting experiences of the 1990s is that devel-          aspects of these countries’ policies are still far from
oping countries as a group grew faster than in the           compliant with conventional wisdom. For example,
1980s. In East and South Asia this reduced the               India has registered fiscal deficits several times higher
income gap with industrialized countries, but in             than Brazil’s or Argentina’s, with lower inflation and
other regions, the gap increased. In Latin America,          lower interest rates.While this fiscal trend is clearly
there were clear gains up to 1998, reversed in the           unsustainable in the long run, and measures have
late 1990s and early 2000s (figures 1.2 and 1.3).            been taken to correct it, it is clear that there is more
    Analyzing policy reforms of the 1990s, several           to macroeconomic stability than a superficial read-
studies (Loayza, Fajnzylber, and Calderon 2002; Lora         ing of the size of the fiscal deficit. China has built
2001a; Easterly 2001) find that countries that               extremely large contingent liabilities related to
improved their policies—strengthening macroeco-              unfunded pensions and nonperforming loans in the
nomic management, opening up their economies,                banking system.While, again, this is not a sustainable
liberalizing their financial sectors—grew faster in the      situation, it suggests that economies do not operate
1990s. However, they also find a large unexplained           in mechanical ways, and that dynamism in one sec-
negative effect associated with both the 1990s and           tor can offset the cost of inefficiency in others. Sim-
the preceding decade.Together with analysis of indi-         ilarly, India’s and China’s industries, though
vidual country experiences and overoptimistic fore-          increasingly competitive in export markets, remain
casts by international financial organizations and           protected and state enterprises still play a large
private entities, these studies give an empirical base       (though declining) role in these economies.
to perceptions that the economic policy reforms of               The mismatch between predictions and results,
the 1990s yielded results below expectations.                and the successes of China, India, and Vietnam
10                                                                 E C O N O M I C G ROW T H I N T H E 1 9 9 0 s



where there were substantial deviations from the        intermediaries channeled to state enterprises and
full package of reforms, suggest several possible       related borrowers, contributing to the massive crises.
explanations. First, sufficient time may not have yet   In some cases, lack of competitive political forces
elapsed for results to emerge in all countries. Over    and such institutions as a free press allowed those
time, market-oriented reforms may ultimately yield      who were politically well connected to take advan-
the results expected. Growth rates in African and       tage of privatizations and to take control of natural
other developing countries have rebounded since         resources while enabling corruption to flourish.
1997; Argentina is experiencing its second year of          These explanations are not mutually exclusive;
rapid growth after the collapse of 2001–02; and         one or more may apply to specific country circum-
growth rates in Eastern Europe have increased. Sec-     stances.The experience also holds some deeper les-
ond, perhaps the reforms implemented in the 1990s       sons. For example, while at one level Argentina’s
were not sufficiently ambitious. Insufficient fiscal    experience teaches that fixed exchange regimes
adjustment in Latin America, very partial privatiza-    require a very demanding set of conditions,a deeper
tion in Africa, and insufficient openness to interna-   lesson is that rigid rules are no substitute for credi-
tional trade in the Middle East and Northern Africa     bility, and that government’s discretion needs to be
may explain performance below expectations in           checked, not replaced with rules. Another deeper
these regions. A third possible explanation is that     lesson is that the reforms of the 1990s did not focus
there were incoherencies in the implementation of       on the binding constraints. For example, they
policies.Argentina introduced a rigid exchange rate     reduced fiscal deficits when perhaps the binding
without the fiscal and financial conditions needed      constraints were lack of public capital and aggregate
to sustain it. Fiscal adjustments in some African       demand. Or they reduced tariffs on imports when
countries were achieved at the cost of reducing pro-    perhaps the binding constraint was the workings of
ductive public spending. Open capital accounts          the financial sector. Or they focused on correcting
encouraged pro-cyclical flows. Correction of these      government failures, when the binding constraints
incoherencies may enable growth to resume.              were market failures.
    Perhaps most important, while reforms in the
1990s focused on increasing the role of markets and
decreasing the role of the state, they tended to neg-   3. Lessons from the 1990s
lect the role of institutions. Francisco Gil Diaz,
Mexico’s Minister of Finance (as quoted in Krueger
2004), recently suggested that                          Promote Growth, Not Just Efficiency
                                                        Reforms need to go beyond the generation of effi-
     The policies that have been undertaken are         ciency gains to promote growth. The policy focus
     not even a pale imitation of what market           of reforms in the 1990s enabled better use of exist-
     economics ought to be, if we understand            ing capacity but did not provide sufficient incen-
     market economics as the necessary institu-         tives for expanding that capacity. While this
     tional framework for a sound economy to            emphasis on efficiency was warranted at a time of
     operate and flourish. What has been imple-         extremely large distortions and waste, it also
     mented throughout our continent is a               explains the frequent instances of stabilization with-
     grotesque caricature of market economics.          out growth or liberalization without growth. The
                                                        experience highlights the importance of the invest-
    State enterprises were privatized without much      ment climate, and of providing predictable condi-
attention to the operation of the markets in which      tions for investors and other economic agents.
they would function. Financial liberalization swelled        It also highlights that growth entails more than
the resources, foreign and domestic, that ineffective   the efficient use of resources. Growth entails struc-
I N T RO D U C T I O N                                                                                               11



tural transformation, diversification of production,             purposes resources and abilities that are hid-
change, risk taking by producers, correction of both             den, scattered, or badly utilized.
government and market failures,and changes in poli-
cies and institutions.It is also a process of social trans-        In retrospect, it is clear that in the 1990s we often
formation: people will change activities and live in          mistook efficiency gains for growth.The “one-size
different places. Social relations will change, and the       fits all” policy reform approach to economic growth
informal networks of rural life will be lost as other         and the belief in “best practices” exaggerated the
more formal networks and organizations are estab-             gains from improved resource allocation and their
lished. Entrepreneurs will invest in new machinery            dynamic repercussions, and proved to be both theo-
to produce new products and adopt new organiza-               retically incomplete and contradicted by the evi-
tional forms. Farmers will adopt new farming meth-            dence. Expectations that gains in growth would be
ods and change their product mix.The economy will             won entirely through policy improvements were
produce and demand different goods and services.              unrealistic. Means were often mistaken for goals—
These changes take place over time, alongside                 that is, improvements in policies were mistaken for
changes in institutions that render them possible.Any         growth strategies, as if improvements in policies
growth strategy needs to include actions, both on             were an end in themselves. Going forward, the pur-
the policy and the institutional front,that address and       suit of policy reforms for reform’s sake should be
support this process of change.                               replaced by a more comprehensive understanding of
    Better policies can bring efficiency gains, and           the forces underlying growth. Removing obstacles
may increase incentives for investment, but without           that make growth impossible may not be enough:
amounting to a growth strategy.They will not nec-             growth-oriented action, for example on technolog-
essarily induce the behavior by private investors             ical catch-up, or encouragement of risk taking for
and the public sector that is needed to put an econ-          faster accumulation, may be needed.
omy on a sustained growth path. For this, faster
accumulation of physical and human capital by
                                                              Common Principles and Diverse Ways to
both the private and the public sector are essential,
as are gains in productivity.                                 Implement Them
    This may explain why the growth impact of the             Another mistake often made in the 1990s has been
reforms of the 1990s was smaller than expected.               the translation of general policy principles into a
The incentives needed to expand productive capac-             unique set of actions. The principles of the 1991
ity (“expanding the frontier” in economists’ parl-            World Development Report, “macroeconomic stabil-
ance) differ from those that are needed to use                ity; domestic liberalization, and openness,” have
existing capacity better (“movements toward the               been interpreted narrowly to mean “minimize fiscal
frontier”). What matters for growth is less the               deficits, minimize inflation, minimize tariffs, maxi-
degree to which policies approximate the ideal than           mize privatization, maximize liberalization of
“the extent to which a given development strategy             finance,” with the assumption that the more of these
is able to mobilize the creative forces of society and        changes the better, at all times and in all places—
achieve ever-higher levels of productivity” (Alejan-          overlooking the fact that these expedients are just
dro Foxley, in World Bank 2005b). And, in Albert              some of the ways in which these principles can be
Hirschman’s words (1958):                                     implemented.
                                                                  There are many ways of achieving macroeco-
    Development depends not so much on find-                  nomic stability, openness, and domestic liberaliza-
    ing optimal combinations for given                        tion. As seen above, for example, the goal of
    resources and factors of production as on                 achieving macroeconomic stability does not imply a
    calling forth and enlisting for development               need to minimize fiscal deficits at all times. A lower
12                                                                     E C O N O M I C G ROW T H I N T H E 1 9 9 0 s



fiscal deficit achieved today through off-budget con-       successful growth experiences in eight East Asian
tingent liabilities, or through cutting back public         economies, reported in the World Bank’s East Asian
investments and thus reducing long-run growth and           Miracle (World Bank 1993), resulted from diverse
the future tax base, may mean a higher fiscal deficit       policy and institutional paths, but common func-
in the future. A lower fiscal deficit does not even         tions were fulfilled along these paths.4 This perspec-
guarantee greater macro-stability if it is based on         tive has several implications.
external borrowing in which interest rates are                  First, different policies can yield the same result,
reduced at the cost of greater vulnerability to             and the same policy can yield different results,
exchange rate fluctuations, or if it is based on build-     depending on country institutional contexts and
ing off-budget liabilities through the banking sys-         underlying growth strategies. This nonformulaic
tem, which eventually translate into an increase in         result holds not only for the eight East Asian
public debt—as Latin American countries and                 economies featured in the 1993 study, but also for a
Turkey found to their cost in the 1980s and 1990s.          larger set of countries 10 years later. Countries with
Similarly, trade integration can be achieved through        remarkably different policy and institutional frame-
various means that offset the effect of tariffs and         works—Bangladesh, Botswana, Chile, China,
reduce the implicit tax on exports. Duty rebate             Egypt, India, Lao PDR, Mauritius, Sri Lanka,
schemes, subsidized credit to exporters, and other          Tunisia, and Vietnam—have all sustained growth in
forms of export promotion, export processing                per capita income at rates above the U.S. long-term
zones, infrastructure, and transport corridors have all     growth rate of close to 2 percent a year.
helped China, India, Korea, and Mauritius to inte-              Second,common to all successes is that four func-
grate into the world economy while keeping their            tions have been fulfilled: rapid accumulation of capi-
tariffs relatively high in the initial phases of integra-   tal, efficient resource allocation, technological
tion and reducing them gradually over time.Thai-            progress, and sharing of the benefits of growth. Rates
land’s and Indonesia’s foreign domestic investment          of progress in these four functions have not always
regimes had few restrictions, whereas those of Korea        been uniform, but successful countries have achieved
and India had many until very recently—but both             a balance among them over time, and disruptions
Korea and India found alternative instruments to            have ensued when the balance was not achieved.
access and adopt modern technologies. Financial             While there can be substitution temporarily, the bal-
intermediation can be increased by relaxing entry           ance will need to be reestablished at some point.
restrictions in the banking system, or by improving             For example, Korea’s policies in the 1960s and
the workings of the legal system, particularly those        1970s sought to encourage risk taking by the pri-
parts that deal with the repossession of collateral.        vate sector. Import protection and priority lending
    To sum up,“getting the policies right” mistakes         contributed to higher levels of capital accumula-
means for ends. Clearly not everything can be right         tion, at the cost of efficient allocation, which
at once, and not everything needs to be “right” for         became a more important priority in the 1980s.
growth to take place—as witnessed in examples               The Soviet Union, well into the 1960s, grew rap-
from Bangladesh, China, India, Indonesia, and many          idly on the basis of sacrificing consumption, accu-
other countries.                                            mulating capital, and maintaining a relatively
                                                            equitable income distribution. But its considerable
                                                            progress in science and technology was not effec-
Common Functions and Diverse Ways to
                                                            tively deployed in production and, more important,
Achieve Them                                                resource allocation was enormously wasteful. Even-
To sustain growth requires key functions to be ful-         tually, the costs of this inefficiency and the political
filled, but there is no unique combination of poli-         reforms of the late 1980s combined to bring growth
cies and institutions for fulfilling them. The              to collapse. In India, a “big push” in capital forma-
I N T RO D U C T I O N                                                                                         13



tion in the decades following independence was                  Different policies can have the same effect, and
complemented in the 1980s—when evidence of                  the same policy can have different effects, depending
misallocation and low productivity growth began             on the context. In large economies, with access to
to emerge—by policies that gradually freed market           foreign technology and equipment, competition and
forces and increased efficiency in resource alloca-         economies of scale lessen the efficiency cost of trade
tion (Virmani 2004), thus ensuring not only the             restrictions and markedly widen the scope for suc-
sustainability of growth but also its acceleration.         cessful inward-oriented industrialization. Brazil,
    Factoring these four functions into analyses of         China, and India were able to develop manufactur-
growth makes it easier to understand why both               ing, many segments of which became internationally
policies and institutions play a role. For example,         competitive, whereas in small countries such as
capital accumulation by the public sector requires          Jamaica and Uruguay, or Sri Lanka in the 1960s and
sound tax policies and administration, sustainable          1970s, the market was too small; the benefits of
macro policies, and a bureaucracy that is capable of        inward-looking industrialization were negligible and
formulating and managing public expenditure                 did not justify its costs. Sri Lanka became successful
programs effectively and of choosing programs               only after it began to liberalize imports in late 1977
with high returns. Accumulation by the private              and follow export-oriented policies.Thus, the same
sector requires at least reasonably secure private          inward-industrialization policy produced different
property rights, stable expectations about the              outcomes because country characteristics differed.
future, a stable macroeconomy, and access to                    Conversely, a given policy can yield different
finance. One country might strengthen private               results because of institutional variation. In Japan
investment by, say, improving expectations, whereas         during the Meiji industrialization and, more
another country could achieve the same result by,           recently, in Korea, public institutions were able to
say, reforming the financial sector.5 Similarly, effi-      resist pressures from narrow interest groups. Public
ciency in allocation requires not only reasonably           enterprises were run efficiently, and state ownership
sound policies—such as competitive exchange                 built capacity in sectors that the private sector had
rates and an open trade regime—but also institu-            not entered because of perceived high risks. The
tions that can enforce contracts and enable markets         same policy in Bolivia, however, where public
to function (World Bank, World Development Report           enterprises were run for the benefit of narrow
2001). Technological catch-up requires not only             interest groups, did not play a strategic role in the
investment and trade policies that enable a country         industrialization process, and most of the enter-
to attract foreign direct investment (FDI) and              prises were liquidated when Bolivia had to stabilize
import equipment, but also institutions that,               its economy in the 1980s. In the case of India, it has
depending on the country’s development stage,               been shown that in the presence of poor institu-
promote adaptive research or a patent regime.               tions, liberalization can lead to less growth than
Indeed, in some instances, it is institutions and           expected (Virmani 2004).
political realities that define the set of feasible poli-       Like that of policies, the effect of institutions
cies, as testified by Russia’s former Minister of           depends on the context. Security of ownership
Finance Yegor Gaidar (World Bank 2005b):                    rights has been achieved in different ways and to
                                                            different extents in different country contexts. In
    If I were the tsar of Russia, I would have              Soeharto’s Indonesia, securing returns depended on
    done everything differently… But if I were              connections with the ruling elite. In contemporary
    deputy prime minister and finance minister,             China, the definition and enforcement of property
    in a government without a parliamentary                 rights depend on party and local government sup-
    majority and under many pressures, I would              port—and only recently have initiatives been taken
    have done more or less what we did.                     in this direction. And in India, success depends on
14                                                                   E C O N O M I C G ROW T H I N T H E 1 9 9 0 s



the functioning of a judiciary modeled after west-        ways in which it can be exerted effectively. What
ern legal systems.                                        was learned in the 1990s is not only that sound
    Sharing the benefits of growth has been impor-        policies do not necessarily engender the institutions
tant in all sustained growth experiences, and partic-     of a modern economy—that institutions are not
ularly in countries with authoritarian forms of           entirely endogenous—but also that institutions can
government, where it has helped to legitimize             prevent the adoption of growth-oriented policies
regimes that often were neither fully representative      or offset their impact. Experience showed how
nor democratic.Various policies have been used to         much institutions matter, and how hard it is to work
promote the sharing of the benefits of growth.They        around their absence or to improve their quality.
include land reform and redistribution of other           Above all, the experience showed that government
assets; public expenditures on infrastructure (the 8-     discretion cannot be bypassed. It is needed for a
7 program in China); social spending (Tunisia); poli-     wide range of activities that are essential for sustain-
cies to increase opportunities to economically            ing growth, ranging from regulating utilities and
underprivileged groups (affirmative actions for           supervising banks to providing infrastructure and
bumiputra in Malaysia); and poverty-targeted pro-         social services. Improving institutions that support
grams (food stamps in Sri Lanka or employment             the implementation of policies, and strengthening
programs in India and Bangladesh).                        checks on the use of discretion, are more promising
    To sum up, diversity in the form of successful        guiding principles than seeking to eliminate gov-
growth experiences should be no surprise. Each            ernment discretion.
successful country was successful in its own way.             Much of the complexity encountered in the
                                                          realm of economic institutions is also found in the
Government Discretion Needs to Be                         institutions governing political life. The formal
                                                          institutions of democracy, for example, do not nec-
Managed and Checked, Not Replaced
                                                          essarily ensure appropriate checks on discretion,
by Rules                                                  nor are those checks always absent in authoritarian
Because developing countries’ societies resolve           regimes. Mechanisms and levels of accountability
agency, predation, and collective decision-making         can take very different forms, rarely amenable to
problems less effectively than do those of industri-      the simplicity of formal political institutions. Much
alized countries, much of the reform effort in the        of the growth success of East Asian countries can be
1990s sought to introduce policies that would limit       attributed to these countries’ ability to allow discre-
the discretion of national authorities in growth          tion by different government agencies, alongside
strategies and minimize demands on institutions.          checks on this discretion that made them account-
Privatization, financial liberalization, and removal of   able.The forms of these checks varied: an authori-
quantitative restrictions on imports are examples of      tarian development-oriented political leader in
policy reforms meant not only to improve incen-           some cases (Soeharto’s Indonesia, Korea in the first
tives for more efficient allocation but also to reduce    decades of its take-off), the checks and balances
the need for government discretion. Dollarization,        inherent in complex one-party systems (China), or
fiscal rules, or integration in larger economic unions    the normal checks and balances of a democratic
are examples of institutional reforms meant to            regime (India, Sri Lanka).
replace government discretion by rigid rules; they
are consistent with the sense that, on balance, the
                                                          Prudent Macroeconomic Management Is at
costs of failures outweigh the benefits of discretion
in the workings of an activist, developmental state.      the Heart of Successful Growth Strategies
    However, government discretion cannot be dis-         Avoidance of busts usually requires avoidance of
pensed with altogether, so it is important to find        booms.The costs of the crises of the 1990s in terms
I N T RO D U C T I O N                                                                                          15



of forgone growth, social distress, and public debt         ance of payments increased; private-sector and gov-
highlight once again the importance of prudent              ernment debt exposures fed resistance to letting the
macroeconomic management. They also stress the              exchange rate adjust; governments sought to sustain
importance of avoiding macroeconomic vulnera-               the rate by drawing down reserves, but the policy
bilities, and the risks associated with indiscriminate      lacked credibility, or reserves were insufficient to
opening of the capital account. Last but not least,         sustain it; a large devaluation followed; and the
they stress the importance of responding quickly to         tightening effect of the devaluation was amplified
downturns. One difference between successful and            by the consequences of currency mismatches for
less successful growth experiences is the frequency         the balance sheets of banks or those of their bor-
of downturns: virtually nonexistent for China,              rowers and of firms. In Indonesia and Korea, for
Korea, or Malaysia, but numerous for Argentina,             example, where current account deficits were rela-
Brazil, and Turkey.                                         tively small, the trigger was the need for a large debt
    In addition to dealing with crises effectively, it is   rollover at a time when investors were retreating
also important to reduce financial fragilities and          from emerging markets and when risk perceptions
hence vulnerability to shocks.The financial crises of       were on the rise.This was accentuated, in Indone-
the 1990s differed from the many that preceded              sia, by the uncertainty of the political transition.
them because of their cost and their suddenness, and            The booms and busts of the 1990s are reminis-
they were much harder to predict.The risks of finan-        cent of some of the crises of the 1980s.They teach
cial integration had been underestimated and its            several important lessons. First, as with most liquid-
gains overestimated. In the 1990s in emerging mar-          ity surges, busts inevitably follow booms: avoiding
ket economies, the opening of the capital account to        the bust requires avoiding the boom and strengthen-
financial inflows triggered large surges that lowered       ing the fundamentals.Countries such as Chile,India,
the costs of sovereign and private borrowing and            or Malaysia that managed inflows,including through
helped reduce inflation. For those reasons, govern-         the imposition of restrictions, were able to weather
ments (with exceptions such as those of Chile, India,       the crises much better than countries that took no
Korea in the early and mid-1990s, and Malaysia)             such precautions. Can a boom be distinguished from
encouraged these inflows. Of the 10 economies that          a favorable lasting trend, ex ante? In most cases the
received the largest inflows, however, 7 suffered           distinguishing factors are the volume of the surge,
severe crises that took the form of large output            the pressure it puts on the exchange rate, and its
declines, higher incidence of poverty, and large            impact on bank credit. Second, the crises of the
exchange rate devaluations. The three exceptions            1990s highlight the extent to which banks can
were China, India, and Hong Kong (China).                   amplify the consequences of a crisis, and the risks
    Each crisis was preceded by a large surge in            associated with currency mismatches,including mis-
inflows that either led to appreciation of the real         matches on the borrowers’ balance sheets. Third,
exchange rate and increased current account                 sovereign borrowing in foreign currencies is risky.
deficits or, as in some East Asian countries, created       While sovereign borrowing should, in theory, help a
an external debt maturity profile excessively biased        country to access external resources on better terms,
toward the short term, and exposed unhedged                 in practice it has encouraged governments and pri-
commercial banks to currency and maturity risks.            vate firms to take excessive risks.
In current account crises—many of which arose in
the context of stabilization programs anchored on a
                                                            Move Away from Formulaic Policy Making
nominal exchange rate—the sequence of events
followed a remarkably similar pattern: a surge in           and Focus on the Binding Constraint(s)
capital inflows put pressure on the exchange rate to        A vital lesson for policy formulation and policy
appreciate; the current account deficit of the bal-         advice is the need to be cognizant of the shadow
16                                                                 E C O N O M I C G ROW T H I N T H E 1 9 9 0 s



prices of constraints, and to address whatever is       at any point in time, the constraint that most
the binding constraint on growth, in the right man-     severely limited growth, and the right sequence of
ner and in the right sequence. This requires rec-       policies needed in each situation.There may be sit-
ognizing country specificities, and more                uations in which a country needs to address many
economic analysis and rigor than does a formu-          constraints at once, as during the transition of East-
laic approach to policy making. Policy makers           ern European countries. These situations are rare,
face the practical problem that no scientific           however. In most cases, countries can deal with
method permits ex ante identification of the most       constraints sequentially, a few at a time. Success in
important constraint(s) binding growth in specific      addressing one or a set of constraints makes it eas-
country circumstances, and hence the specific           ier to deal with the others, and may help establish
measures that are needed address it (them). Dur-        virtuous circles.
ing the 1980s and 1990s, China’s approach was to
“cross the stream by groping for the stones.” Con-
straints were identified and dealt with as the          4. Lessons from Policy and
growth process unfolded, through experimenta-              Institutional Reform
tion and trial and error (chapters by Lim and              Experiences in the 1990s
Huang, in World Bank 2005a).
    Which policy should be introduced, and when,        The economic policy reforms of the 1990s focused
varies considerably from case to case depending         on improving efficiency in the allocation of
on initial conditions and institutional endow-          resources through macroeconomic stabilization,lib-
ments. For example, one can generally assume that       eralization of trade and the financial sector, privati-
where hyperinflation is raging, or public debt          zation, and deregulation. Deregulation and
demands high real interest rates—as it does in          reduction in the role of government were expected
Argentina, Brazil, Jamaica, and Turkey, for exam-       to improve the governance of the public sector
ple— macroeconomic stabilization is the first pri-      through improvements in incentives for perform-
ority. Where trade restrictions are extreme and         ance, more transparency, and fewer opportunities
hinder utilization of existing capacity, as in many     for rent seeking. Institutional reforms focused on
countries of the Middle East and North Africa,          improving collective decision making and solving
reducing them will be essential. Where there is         agency problems through democratization, decen-
uncertainty regarding the future course of eco-         tralization, and public sector reforms aimed at
nomic policies, as in Bolivia, Democratic Repub-        enhancing the efficiency, transparency, and account-
lic of Congo, and Nigeria, financial sector             ability of government activities. From 60 countries
liberalization will do little to channel resources to   choosing their leaders through competitive elec-
private investment. Where property rights are           tions in 1989, the number rose to 100 by 2000. Del-
poorly defined and enforced, and regulation pre-        egation to subnational levels of government of
vents the movement of domestic resources across         political, administrative, and financial powers has
sectors, as is still the case in some Central Asian     taken place not only in federated states such as
and some African countries, trade liberalization        India, Brazil, and Russia but also in smaller states
will be of little effect.                               and centralized states such as Bolivia and the Czech
    Experimentation and learning is hence an            Republic. Deregulation and privatization have been
important part of the growth process.The East Asia      trends virtually everywhere, even though the inten-
Miracle study highlighted that behind the miracle       sity of the reforms has varied significantly from
was the East Asian countries’ willingness to exper-     region to region.
iment, and ability to learn from, not to persist in,        What have we learned from a decade of reforms
their mistakes.This approach helped them identify,      in these areas?
I N T RO D U C T I O N                                                                                       17



Macro-stability Needs to Be Achieved in                  to destabilize the economy, and they maintained the
a Manner That Is Sustainable and                         competitiveness of the real exchange rate. These
                                                         countries fared much better than those that opened
Pro-Growth
                                                         themselves to external liquidity surges. Notwith-
The rise in real interest rates in the late 1970s and    standing the theoretical arguments in favor of capi-
early 1980s, combined with a variety of commod-          tal account openness, the evidence on growth is
ity price shocks, had rendered unsustainable the fis-    inconclusive and volatility clearly increased. A
cal stances, debt levels, and exchange rate regimes      major lesson of the decade is that restrictions should
of most countries in Latin America, East and South       be placed not so much on outflows as on inflows.
Asia,Africa, and the Middle East. Performance dif-       Obviously, differentiating an unsustainable boom
fered sharply between countries that rapidly             from a positive sustainable trend can be difficult, but
adjusted to these shocks (Korea and East Asian           standard indicators of vulnerability such as indebt-
countries in general) and those that did not (Brazil,    edness, evolution of the real exchange rate, and cur-
Nigeria, and many other countries in Latin Amer-         rent account deficits have proven to be reliable, if
ica and Africa).                                         imprecise, tools.
    As a result, the Structuralist view that inflation       Macroeconomic stabilization programs often
and macro instability were inevitable companions of      suffered from other design flaws, which created
structural transformation and growth was replaced        serious macroeconomic fragilities. While primary
in the 1990s by the strong belief that macroeco-         deficits did decline over the 1990s, public debt
nomic stability was needed for growth.The 1990s          increased in most countries, whether because of the
indeed saw considerable progress in this area: fiscal    bank recapitalization costs of financial crises (as in
deficits declined in most countries, exchange rates      Indonesia,Turkey), or because of the cost of contin-
were adjusted to reflect market realities, black mar-    gent liabilities being shifted to the public sector
kets for foreign exchange disappeared, and inflation     (pensions in Argentina), or because of high real
declined virtually everywhere. However, while            interest rates on the public debt (as in Brazil and
macroeconomic policies as conventionally meas-           Jamaica). Other design flaws help explain why the
ured improved in a majority of countries, the            search for macro-stability may in some cases have
growth benefits failed to materialize. In addition,      actually been inimical to growth. A preoccupation
financial crises were numerous, with severe adverse      with reducing inflation led some countries to adopt
effects on economic growth and poverty.                  exchange rate regimes that ultimately proved desta-
    The openness of the capital account was a key        bilizing—price stabilization was achieved at the cost
source of fragility which, combined with unsound         of appreciating exchange rates. Fiscal adjustment
policies in the financial sector (such as currency       was often based on highly distortionary taxes (for
mismatches on banks’ or final borrowers’ balance         example on external trade or on domestic financial
sheets in the absence of hedging instruments) and        transactions); or on cuts in spending on productive
appreciation of the real exchange rate, helps to         infrastructure or human capital that proved detri-
explain many of the crises of the 1990s. Countries       mental to sustained growth; or on borrowing
such as India have avoided appreciation of the real      abroad where interest rates were lower but currency
exchange rate, and made the opening of the capital       exposure increased risks. Hence a single-minded
account a medium-term goal, to be realized con-          pursuit of macro-stability sometimes came at the
tingent on strengthened economic performance             cost of public spending that might have both
(including fiscal adjustment), export diversification,   increased growth and made stability more durable.
and achievement of a sound banking system. Chile             There are two lessons to draw from this experi-
and Malaysia, among others, did not hesitate to tax      ence. First, even with macroeconomic stability,
capital inflows when excessive liquidity threatened      macroeconomic vulnerabilities induced by policy
18                                                                  E C O N O M I C G ROW T H I N T H E 1 9 9 0 s



flaws can be serious, and these can have tremendous      takes the form of a publicly announced inflation
costs. However, indicators of sustainable macro-sta-     target—has succeeded among emerging market
bility are less self-evident than common indicators      economies during the past decade (Brazil, Chile,
of fiscal and external stance suggest.The inability of   Colombia, Korea, Mexico, Peru, South Africa, and
financial markets (as measured by country risk pre-      Thailand). This institutional arrangement has the
mia) to predict most of the financial crises of the      important advantages of flexibility (since the central
1990s provides further evidence that unambiguous         bank is not constrained in how it attains its inflation
indicators of risk are difficult to find.                target) and commitment (since the central bank’s
    Second, the institutions underlying macroeco-        prestige is publicly put on the line).
nomic outcomes and stability matter as much as sta-
bility itself.There is ample evidence that budgetary     Trade Openness, a Key Element of
processes influence fiscal outcomes and that coun-
                                                         Successful Growth Strategies,
tercyclical fiscal policy rules strengthen macroeco-
nomic stability. Centralized budget processes lead to
                                                         Can Be Achieved in Many Ways
better balanced fiscal outcomes over time, and           During the 1980s, the performance of countries
countercyclical fiscal policies shorten cycles and       that responded to shocks by increasing their out-
narrow their amplitude. Few governments find it          ward orientation (East Asian countries) contrasted
politically appealing to run fiscal surpluses during     sharply with that of countries that did not (Latin
good times, however. Transparent fiscal rules, with      America, Africa, most countries of the Middle East
stipulated penalties for noncompliance, may be           and North Africa). Most policy makers concluded
effective in some contexts. In others, the creation of   that openness mattered for growth and, as a result,
institutions such as oil stabilization funds may be      during the 1990s, most developing countries signif-
needed to save windfalls. One promising example is       icantly reduced tariffs on imports and dismantled
Chile’s Structural Surplus rule, which establishes       other forms of trade restrictions. As in the case of
fiscal policy targets adjusted for the variation in      macroeconomic reforms, however, the results var-
growth over the cycle. Other proposals, yet to be        ied and, in general, fell short of expectations.
adopted, have focused on creating an independent         Whereas openness helped efficiency and growth in
fiscal policy council, modeled along lines similar to    many cases (East and South Asian countries,
an independent central bank, that would set annual       Botswana, Chile, Mauritius,Tunisia), it failed to do
deficit limits.Another institutional dimension of fis-   so in many others. Several lessons emerge.
cal policy is transparency. Uncertainty about the            First, openness to trade has been a central ele-
state of the fiscal accounts probably played a large     ment of successful growth strategies. Although the
role in generating the volatility of the risk premi-     paths taken toward greater integration with the
ums that developing-country borrowers faced dur-         world economy were far from uniform during the
ing the 1990s. There is also evidence that more          1990s, the most successful developing countries
transparent budgetary processes brought down             reduced barriers to international trade and foreign
deficits and debt.                                       investment during the decade.
    For monetary policy, institutional arrangements          Second, trade is an opportunity, not a guarantee.
are equally important to ensure that low and stable      Trade reforms in some countries yielded few gains
rates of inflation are achieved and maintained, and      in terms of export expansion or increased eco-
that they last. However, there are no magic institu-     nomic growth, while creating social and economic
tional shortcuts to monetary credibility, which has      adjustment costs. Liberalization of trade in
to be earned through anti-inflationary perform-          Argentina in the 1980s and 1990s, and in Chile in
ance. The institution of an independent central          the early 1980s, for example, was accompanied by
bank—with a commitment to price stability that           an appreciation of the real exchange rate that
I N T RO D U C T I O N                                                                                       19



reduced the competitiveness of domestic industries,      tionship between trade reforms and poverty is to
and incentives to exports—with adverse conse-            date mostly indirect. Even where trade policy has
quences for the balance of payments and the real         reduced poverty, there are still distributive issues.An
economy. In some countries of Eastern Europe in          important policy lesson is that countries need to
the 1990s, trade was liberalized while property          help the affected workers move out of shrinking
rights were not well defined, and the institutional      (import-competing) sectors into expanding
base for a market economy was not well developed.        (exporting) sectors.
These, and other institutional issues preventing the         Fifth, the preservation and expansion of the
free movement of resources, often meant that trade       world trade system hinges on its ability to strike a
reforms did not expand economic opportunities            better balance between the interests of industrial-
but restricted them instead (Bolaky and Freund           ized and developing nations. Though more sup-
2004). Such experiences do not imply that less           portive of development than at the beginning of
trade reform would have been desirable, but that         the 1990s, the world trade system is still biased
trade reform must be done sensibly, as part of an        against the poor. Notwithstanding a decade of sig-
effective growth strategy.                               nificant expansion of international trade, global
    Third, countries that have successfully opened       markets are most hostile to the products the world’s
their economies have done so following a striking        poor produce—agricultural products, textiles, and
variety of policy approaches.They have opened up         labor-intensive manufactures—and problems of
different sectors at different speeds (for example       escalating tariffs, tariff peaks, and quota arrange-
Bangladesh and India). Some, such as China and           ments systematically deny the poor market access
Mauritius, have achieved partial liberalization          and skew incentives against adding value in poor
through the establishment of export processing           countries. These problems are embedded in the
zones, and some have combined unilateral trade           remaining structure of protection in both industrial
reforms with participation in regional trade agree-      countries and developing countries (the latter
ments (Mexico and countries in Central and East-         owing to their own anti-export biases and also to
ern Europe that have now joined the European             higher barriers to trade in developing-country mar-
Union). These differences, and differences in the        kets), and they can be addressed through collective
range of complementary policies adopted, make it         actions.Those actions are best achieved through the
difficult to pin down the statistical relationship       Doha Round and the World Trade Organization
between trade integration and growth. The aca-           (WTO). Although there is a role for nonreciprocal
demic debates on whether openness to trade causes        preferences and for reciprocal regional approaches,
higher growth are riddled with problems of meas-         such preferential arrangements are economically
urement, reverse causation (faster-growing coun-         arbitrary; they come at a cost to excluded countries
tries tend to open their markets more quickly), and      and are not the best way to generate the right
omitted variable bias (countries that successfully       incentives for investment.
lower tariffs and increase growth also adopt other
complementary policies).                                 Design Privatization and Deregulation
    Fourth, the distributive effects of trade liberal-
                                                         with regard to Institutional Strengths and
ization are diverse, and not always pro-poor. Trade
reforms were expected to be pro-poor because in          Weaknesses
most societies the relatively wealthy and urban          Privatization and deregulation have a potentially
classes have been more successful at using protec-       large efficiency impact and can benefit the popula-
tion for their own benefit.The expectation was that      tion at large, including the poor. But there is a need
trade reform would increase the incomes of the           to keep expectations realistic as to what they can
unskilled.Yet evidence from the 1990s on the rela-       achieve, to establish the institutions that are key to
20                                                                   E C O N O M I C G ROW T H I N T H E 1 9 9 0 s



success, and to design privatization strategies taking    cally ensure this separation.The well-publicized dif-
into account institutional strengths and weaknesses.      ficulties of doing business in countries such as Rus-
    Privatization and deregulation were key areas of      sia show that a government can use a wide range of
reform in the 1990s. Commercial public enter-             laws to influence a firm’s decisions without owner-
prises, development banks, and other forms of pub-        ship. Separating the commercial from the political
lic interventions in the economy, even when meant         requires institutions that define and limit govern-
to address market failures, had become discredited        ment powers. Notwithstanding the claims of some
because in many instances they had failed to work         privatization advocates, institutions to support a
well in practice. State activist policies using discre-   well-functioning market economy will not spring
tion, combined with weak accountability in public         up quickly in response to demand. The lack of
sector organizations and weak political accountabil-      effective institutions permits predation through sev-
ity of states to citizens, were producing costs that      eral avenues, not just the government: in extractive
were just too high.The end of communism, and the          industries, for example, the mining or petroleum
deregulation revolution in the United States and          firm and the government are beholden to each
the United Kingdom, added further impetus for the         other, and either could act or collude at public
wave of privatization that swept across the industri-     expense.Vested interests could act through either
alized and developing world in the 1990s.                 the public or private sector, and poor shareholder
    The results varied. In most countries, privatiza-     oversight over a firm, as well as poor public over-
tion brought unambiguous gains in terms of more           sight over governments, permits misappropriation.
efficient use of resources, more investment, and              Turning to utilities, the second major area of
enhanced welfare for consumers.At the same time,          privatization during the 1990s, there are three main
however, privatization itself failed to bring about all   lessons. First, expectations of private investment in
the gains for investment and growth that were             infrastructure have been overly optimistic—because
expected of it. It is also clear that too much was        (1) underpricing continued to be a problem that
expected from privatization, particularly in some         governments did not fully address, (2) the risks of
areas of infrastructure, but also in terms of the gov-    infrastructure investment were not appreciated, and
ernance improvements it would bring. In cases             (3) governments could not credibly to commit to a
where the overall package of reform failed to bring       policy and regulatory regime. At the beginning of
about the expected growth, even the efficiency            the 1990s, the private sector was expected to enter
gains of privatization were put in question—a prob-       virtually all areas of infrastructure, including roads,
lem that is particularly serious in Latin America and     but experience has since shown that the risks
Africa, where it has in some cases derailed the pri-      involved in infrastructure investment are often too
vatization process.                                       large to be taken up by the private sector.
    In addition, the process of privatization has often       Second, if privatization is overstated as a means
been less than fully transparent and competitive, and     of severing the link between economics and poli-
this has left sequels that in some cases can be costly    tics, regulation as a means of restoring the link is
to repair—particularly where privatization has led        underappreciated. The clearer the separation
to concentration of economic power, as it has in          between economics and politics, the better it is for
many parts of Africa and Eastern Europe.                  each: commerce will be more efficient and politics
    Privatization is not just finding “better owners”     less corrupt. But the more complex the regulatory
than the government but about changing gover-             issue, the more likely are mistakes, and the less likely
nance to separate the commercial from the politi-         that bad regulation (and capture of the regulators by
cal. As is now widely accepted, government                vested interests) will be detected. Even if detected,
ownership of a commercial firm makes this separa-         the poorer the institutions, the less likely it is that
tion difficult. But privatization does not automati-      bad regulation will be corrected.
I N T RO D U C T I O N                                                                                      21



    Third, the reform experiences of network utili-      almost universal pension coverage trimmed the
ties clearly show that there is no universally appro-    benefits of their defined benefit schemes, out of fis-
priate reform model, and that privatization is not       cal necessity. Many Latin American countries
necessary or indispensable for every country. Every      sought to phase out their defined benefit schemes
restructuring and privatization program needs to         and replace them with mandatory coverage by pri-
explicitly consider the specific features of each sec-   vate providers through defined contribution
tor (its economic attributes and technology) as well     schemes. Few of these schemes have lived up to
as the country’s institutional, social, and political    their billing: despite favorable demography their
characteristics. Important lessons in this respect are   coverage remains low because of the small size of
as follows:                                              the private formal labor market; their administrative
                                                         costs have been high, partly because insurance costs
• Regulatory reform should promote competi-
                                                         are included and partly because of start-up costs,
  tion, not control; competition is the most effec-
                                                         and they remain dependent on government
  tive regulator.
                                                         finances because there are few securities besides
• Getting the economics right is key. Understand-        government paper to invest in.There was really no
  ing the source of benefits helps in structuring        way to isolate these countries’ social security and
  the reform.A pricing policy that does not allow        pension schemes from their governments without
  adequate revenue cannot improve the situation          allowing them a greater range of investment in
  even if a utility is privatized or an independent      external markets.
  regulator is established. For example, as of 2000,
  in almost all Commonwealth of Independent              The Impact of Financial Liberalization on
  States (CIS) countries, household electricity
                                                         Growth Depends on Underlying Institutions
  prices covered less than 50 percent and indus-
  trial prices were less than 70 percent of the long-    and on Macroeconomic Management
  run marginal costs of supply.                          Over the 1980s and 1990s, as part of the general
                                                         shift to a more market-oriented economy, the
• Institutions differ, and hence regulatory agencies
                                                         approach to finance shifted away from holding
  cannot be easily transplanted. Countries differ
                                                         down interest rates, limiting competition, and rely-
  greatly in their economic structures and in their
                                                         ing on governments to allocate credit and toward
  institutionsæthe whole chain that includes
                                                         more market-based, internationally open systems.
  courts (where appeals are made), legislatures
                                                         Financial liberalization reflected the reaction to the
  (where laws are passed), the press (which informs
                                                         costs, corruption, and inefficiencies of financial
  the electorate), an engaged public (which
                                                         repression; the demands of government and the
  demands more from governments), and acade-
                                                         public for more financial resources and services; and
  mia (which trains regulators and encourages
                                                         the pressures from greater trade, travel, and migra-
  studies of problems). These institutional differ-
                                                         tion, and better telecommunications.
  ences across countries determine why what is
                                                             Contrary to expectations, financial liberaliza-
  sound regulation in one country is ineffectual in
                                                         tion did not add much to growth, and it appears to
  another.They are analogous to the differences in
                                                         have augmented the number of crises.As expected,
  performance of state-owned firms: they are dis-
                                                         deposits and capital inflows rose sharply as a result
  appointing in some countries (India, Mexico)
                                                         of liberalization. But, other than in a few East Asian
  but not in others (Sweden or France).
                                                         and South Asian countries, capital markets did not
    Pensions are an area in which the private sec-       provide resources for new firms. Numbers of stock
tor’s contribution has most clearly fallen short of      market listings declined, even in the newly created
expectations. Eastern European countries with            markets in the transition countries that were some-
22                                                                   E C O N O M I C G ROW T H I N T H E 1 9 9 0 s



times used for privatizations. Also, although rele-       tribute to capital flight and devaluation), bank clo-
vant time-series data on access are weak, and con-        sure, and handling of explicit and implicit guaran-
trary to expectations, it appears that access to          tees to depositors; experience in the 1990s suggests
financial services did not improve substantially after    that it is difficult to avoid socializing the losses and
liberalization.                                           a fall in output. Further, open capital accounts and
     The explanation for these disappointing out-         volatile international capital flows place a large pre-
comes lies largely in weak institutions, concentrated     mium on sound macroeconomic management.
economic and political power, and macroeconomic           Internationally, few attempts have been made to
shocks. The implicit and explicit guarantees that         reduce the volatility of capital inflows (reducing
were extended to depositors and investors weak-           volatility depends on limiting the upside, not just
ened the market discipline that might have limited        trying to stop outflows when a crisis develops).
the activities of weak lenders. By the end of the         Chile’s implicit taxes on short-term inflows appear
1990s, much of the deposit growth had been                to have had some success in extending maturities,
absorbed by central bank debt and government              reducing inflows, and limiting volatility against
deficits.The state banks, which remained important        small shocks, albeit at the cost of reducing credit
during the 1990s, and financial industrial conglom-       availability to the private sector (Edwards 1999;
erates used their increased deposits to expand lend-      Forbes 2003). Part of India’s success in avoiding a
ing to state enterprises, well-connected borrowers,       1997 crisis stemmed from its limits on banks’ (and
and other parts of financial-industrial conglomer-        firms’) offshore borrowing, even as it allowed
ates. Regulation and supervision were weak, reflect-      inflows into the stock market and eased direct for-
ing not just technical problems but also political        eign investment. Indonesia’s limits on state banks’
pressures for leniency. Eventually the poor quality       external borrowing did reduce their growth, but
of lending was exposed in crises, as were the weak-       excessive inflows to private banks and corporations
nesses of the bank privatizations in the context of       were a major factor in the 1997 crisis. Except for
the weak institutional environment and the exclu-         Chile’s taxes on short-term flows and some
sion, in many cases, of international banks.              attempts to hold down interest rates, countries have
     The lack of improved credit access reflected not     made few attempts to remove the incentives to
only the preemptive borrowing by the public sector        banks for increasing their offshore borrowings. All
and central banks but also weak informational and         attempts at limiting excessive inflows depend on
legal frameworks. Lack of information on borrowers        political will to restrict them during a boom. In
hindered lenders and gave borrowers no incentive          practice, countries often have eased restrictions on
to maintain a good credit record. Weak legal and          capital inflows to prolong a boom with negative
judicial frameworks (designed to protect borrowers        consequences when the flows necessarily slowed.
and often responsive to economic and political                Improvements are being made in regulation and
elites) reduced the incentives to service debts; they     supervision in an attempt to limit financial crises,
made it difficult for new borrowers to gain access to     but experience in the industrial countries, where
finance by pledging collateral effectively and made it    political and economic power is more diffuse than
difficult for lenders to execute collateral.              developing countries, suggests that this will not be
     The 1990s reinforced the old lesson that suc-        easy. In the United States, for example, financial
cessful financial liberalization depends on macro-        economists have raised concerns about some U.S.
economic management. No banking system,                   banks being too big to fail. Also in the United
however sound in principle, can withstand a serious       States, political forces and regulatory forbearance
macroeconomic crisis. Dealing with a banking cri-         are often cited as contributory factors in the savings
sis is quite complex, involving highly political issues   and loan crisis. In many developing countries a few
of liquidity support to banks (which can easily con-      large banks, often state-owned, dominate the sys-
I N T RO D U C T I O N                                                                                     23



tem. Bankers and major borrowers are often one          tutions, defined broadly as the “rules of the game”
and the same. Limits on connected lending are a         that shape the behavior of organizations and indi-
problem because the industrial-financial groups are     viduals in a society (North 1990, 3).6 A crisis of
also the main entrepreneurs in many countries,          governance of varying intensity pervades much of
even large ones. If problems of loan quality develop,   the developing world, with the poor paying the
the political strength of the economic and political    heaviest price for it.
elite will likely lead to regulatory forbearance in         Public sector reforms in the 1990s sought to
loan classification and provisioning standards.These    change the structure of organs of the state, and
kinds of problems suggest that attempts to improve      incentives within them, in the hope of improving
the regulatory and supervisory framework need to        government efficiency and responsiveness. From
include a substantial effort to improve market disci-   mega-reforms such as decentralization to less
pline, through better information on the banks and      sweeping reforms in budget or personnel manage-
credible limits on deposit guarantees. Increased        ment, the aim was to find a balance between the
entry of well-known foreign banks, which have a         discretion of politicians and bureaucrats over policy
reputation to protect, can also improve the func-       making and policy implementation and their
tioning of the system.                                  accountability for decisions and actions.The fall of
    Thus, in finance, the 1990s may best be regarded    authoritarian regimes and the consequent spread of
as a transition period.The high expectations for lib-   democratic processes constrained the previously
eralization were met only in resource mobilization.     wide discretion of many governments. Decentral-
Resource allocation, which makes a key contribu-        ization sought to further limit central government
tion to development, did not generally improve.         discretion while granting local governments more
However, much of the debris of the old financial        managerial autonomy. Legal, judicial, and legislative
system was removed by the crises, albeit by govern-     reforms were initiated to establish institutional
ment recapitalization bonds that now represent          checks on executive power. Public management
much of the system’s assets.                            reforms sought to give public managers more flex-
    As the connecting link between savers and           ibility in decision making while demanding greater
investors, the contribution of finance to growth        accountability from them for their decisions. Per-
depends not only on macroeconomic stability and         haps partly because of the immense difficulty of
reasonable interest rates, but also on the quality of   addressing problems in political institutions, many
financial intermediaries and information and of the     countries and donors in the 1990s focused largely
legal and regulatory framework. Improving the           on reforming legal and judicial systems—a channel
contribution of finance to development will             of political accountability that seemed more
depend not only on market-based finance but also        amenable to technocratic solutions, often using
on sound institutions, appropriate incentives for       models directly transplanted from industrialized
lenders, further improvements in informational and      countries.
legal frameworks and, ultimately, on a more com-            Most of the reforms had little effect on behav-
petitive political system that is able to reduce the    ior.The ills that they sought to treat—nonmerito-
power of political-economic elites and their ability    cratic civil services, weak financial controls, opaque
to tap the financial system.                            or incoherent budget processes—are deeply rooted
                                                        in local political and institutional arrangements that
                                                        favor the status quo.
Pragmatic, Incremental Approaches to Public
                                                            The decade was not all discouraging, however.
Sector Governance Are More Effective                    Homegrown initiatives gave hope for improving
Economic performance depends partly on gover-           government performance. In some instances, civil
nance, which in turn is shaped by underlying insti-     society engagement and participation and innova-
24                                                                  E C O N O M I C G ROW T H I N T H E 1 9 9 0 s



tive applications of information technology led to        Politics: Checks and Balances Are Central to
improvements in transparency and accountability in        Accountability and Results, but There Is No
public decision making and consequently to some
                                                          Single Way to Achieve Them
increase in government responsiveness, efficiency,
and effectiveness.The challenge is scaling up these       Institutions resolve a number of problems in soci-
initiatives, given political constraints and historical   ety, of which two are particularly important: collec-
inertia.                                                  tive decision-making processes, and principal-agent
    The designs of governance reform strategies in        problems.7 Not all preferences can be represented
the 1990s typically fell into two broad types: “big       in collective decision making, and principal-agent
bang” or ad hoc incrementalism. Big bang                  problems can be reduced but never resolved.
approaches proved to be largely inconsistent with             Both theory and evidence suggest that the for-
capacity constraints and political realities. Their       mal rules of democracy do not ensure efficient,
main results were major changes in formal rules:          accountable, and credible government, and con-
new or amended constitutions, new legislation,            versely that nonelected governments are not inca-
ostensibly independent courts and audit institu-          pable of responding to citizens or of acting
tions, and so forth. Meanwhile, the informal rules        accountably. Though the number of elected gov-
shaping the incentives that face politicians, bureau-     ernments grew significantly in the 1990s, the
crats, and citizens remained in place.                    decade produced no clear evidence that elected
    Ad hoc incrementalism has also been problem-          governments perform better in delivering policies
atic. Many of the ad hoc reforms were symbolic,           benefiting average citizens than do nonelected
intended to preserve the old informal rules while         ones. The experience did confirm, however, that
pretending to reform. Some represented well-moti-         relative to the situation in richer democracies, pri-
vated attempts of individual or small groups of           vate investors in most developing-country democ-
reformers who,for lack of support,were undermined         racies receive less enforcement of their contractual
by jealousy, intrigue, or fatigue. More important, they   and property rights, and average citizens are not as
tended to be unrelated to a more coherent reform          well treated by the state as special interests.
strategy and thus over time many lost their steam.            By the close of the 1990s, we had begun to
    An important general lesson is that technocratic      understand the complicated interaction of formal
responses to the governance crisis work only in very      political institutions with informal rules and norms.
auspicious settings—where there is committed              Elected governments are most likely to make poli-
leadership, a broad-based coalition in support of         cies at the expense of the majority and in favor of
reform, and sufficient capacity to carry the reform       narrow segments of the population when citizens
process forward. Clearly, these conditions exist in       are badly informed about what government does,
only a few developing countries, and rarely in those      when political competitors cannot make credible
that most need governance reform.                         promises to voters, and when society is polarized.
    State building is a complex process that requires     Evidence shows that uninformed or polarized citi-
time, leadership, and social capital. Governance          zens and noncredible politicians undermine the
reforms have to find a delicate balance consistent        connection between voters and politicians in
with the country’s politics, history, and culture.What    democracies. Long-run economic growth and the
may be needed are highly focused, pragmatic inter-        provision of public goods are significantly higher in
ventions that may be termed “strategic incremental-       democracies with more credible politicians, better
ism.”These interventions are opportunistic because        informed citizens, and less social polarization. Non-
they exploit the willingness to reform, but they are      democracies vary substantially as well: those that
grounded in political realities and consistent with       have internal checks on the exercise of discretion
the capacity constraints of the country concerned.        by the executive seem to perform better, both in
I N T RO D U C T I O N                                                                                     25



terms of growth and public policy performance,           implemented within a specific institutional, social,
than others.The lesson here is that governments of       and historic context. Recent economic and sector
all kinds, elected or not, are most credible and most    work at the World Bank already seeks to achieve a
likely to respect property rights when they face         better balance between country specificities and the
checks and balances on their decision making.            lessons from country experiences, but more is
    Another lesson of the 1990s is that policies fail    needed fully to recognize that country-specific
when citizens cannot hold politicians accountable        market structures and institutions have a strong
for poor performance and when governments can-           influence on policy outcomes. In particular, this
not make credible commitments. Credible, sustain-        recognition calls for harder and more rigorous eco-
able reform depends on the checks and balances           nomic, institutional, and social analysis.
provided through political institutions. In democ-           Third, analytical work needs to change its orien-
racies, checks and balances and elections prevent        tation, away from seeking to assess how far policies
arbitrary policy reversals by governments. But they      diverge from optimality, to seeking to assess what
are not the only means to hold governments               policy and institutional conditions—for capital
accountable: broad-based political parties can in        accumulation, shared growth, productivity growth,
some circumstances substitute for democratic             and risk taking in a country-specific context—are
checks and balances in one-party states.                 needed to set the growth process in motion.


5. Operational Implications                              For Strategy
                                                         There is a need to rethink the focus of growth
The complete operational implications of this study      strategies and of development assistance. Up to
still need to be fully developed. Some preliminary       now, that focus has been on the nation state with
ideas are outlined below.                                the implicit assumptions that (1) development
                                                         outcomes within the boundaries of a nation state
                                                         are homogeneous, and (2) all developing coun-
For Analysis
                                                         tries’ per capita incomes could and should con-
On the analytical front, the first implication is the    verge with those of industrialized countries.There
need to redress the balance between analysis of pol-     is now greater evidence and acknowledgment that
icy instruments and analysis of strategies—understand-   these two assumptions do not always hold. Con-
ing strategies as coherent sets of actions that are      vergence is much less a force now than anticipated
intended to initiate and sustain growth. Over the        a decade or more ago. Within countries such as
years, in institutions such as the World Bank, the       Brazil, China, and India, income differences across
focus of research gradually has shifted away from        regions are as large as income differences across
country-specific growth experiences to focus             countries, and even in relatively small Bolivia,
increasingly on policies—trade, finance, macro, pri-     income differences between the lowlands and the
vatization to name a few—with secondary impor-           highlands are large.This recognition implies a need
tance given to country contexts.8 At the same time,      to pay much greater attention to the forces driving
outside the World Bank there has been increasing         agglomeration and migration, both within and
emphasis on individual country experiences (for          across countries.
example, Rodrik 2003b, and the research programs
sponsored by the Global Development Network).
     The second implication is the need to recognize     For Research
country specificities in country economic analysis,      On the research front, two issues in particular war-
acknowledging that policies are conceived and            rant further examination.The first relates to devel-
26                                                                   E C O N O M I C G ROW T H I N T H E 1 9 9 0 s



opment agencies’ role in aid-dependent countries.        For Behavior
The agencies’ large role in financing the budget
                                                         On the behavioral front, if solutions must be found
has forced them to be involved in budget
                                                         in specific-country contexts, rather than applied
processes, weakening national decision making
                                                         from blueprints, those who advise or finance devel-
and rendering the concept of “ownership elusive
                                                         oping countries will need more humility in their
in practice” (Kwesi Botchwey,World Bank 2005a),
                                                         approaches, implying more openness on the range
particularly in aid-dependent Africa. Clearly, forms
                                                         of solutions possible, more empathy with the coun-
of engagement developed for project finance do
                                                         try’s perspectives, and more inquisitiveness in assess-
not apply to budget finance.There may be a need
                                                         ing the costs and benefits of different possible
to explore new approaches to the transfer of
                                                         solutions.
resources to these countries, rooted in public
finance, such as those typically used in federated
nations that have chosen rule-based, arms-length
systems of transfers.
                                                         Notes
    Second, the unit of analysis for economic and         1. See Country Note 3, “Poverty and Inequality: What
social development has traditionally been the nation         Have We Learned from the 1990s?”
state, reflecting the assumptions (outlined above)        2. Countries successful at “converging” include most
that nations are homogeneous and that all nations            South Asian countries (Bangladesh, Bhutan, India,
                                                             Nepal, Sri Lanka); many East Asian countries (China,
would be able to catch up to the income levels of            Indonesia, the Republic of Korea, Lao PDR, Malaysia,
industrialized countries. There is a rich research           Thailand, Vietnam); and Botswana, Chile, the Arab
agenda on these assumptions that needs to be artic-          Republic of Egypt, Lesotho, Mauritius, and Tunisia. See
ulated. All nations may not succeed in reaching              Country Note 2, “Lessons from Countries That Have
industrialized countries’ income levels within a rea-        Sustained Their Growth.”
                                                          3. This point of view was reinforced by the fiasco of the
sonable time frame—partly because institutions can
                                                             collapse of the Southern Cone stabilization programs
take such a long time to develop, but also because           of the late 1970s, in which strong currency apprecia-
the economics of agglomeration and poles of devel-           tions combined with rapid reductions in tariffs to cre-
opment do not necessarily follow national bound-             ate an adverse shock to industry, ultimately derailing
aries. Research in this area may yield important             the stabilization programs. Most analysts soon blamed
implications for the role of nations and migration,          the collapse on excessive speed, leading to faulty
                                                             sequencing of the reform program: they argued that
and also for the optimal degree of discretion regard-        capital accounts had been liberalized too soon, without
ing national policies.                                       waiting until fiscal probity had been established and
                                                             both trade and the domestic financial system had been
                                                             successfully liberalized.
For Operations                                            4. This study reviewed the growth experience of eight
                                                             economies: China, Hong Kong (China), Indonesia,
On the operational front, the recognition that not
                                                             Japan, Korea, Malaysia,Taiwan,Thailand, and Singapore.
everything needs to be right for growth to succeed,          The report highlighted the variations in policy and
and that partial success may sometimes be a more             institutional environments under which these
pragmatic goal than optimal policies, has obvious            economies reached unprecedented rates of growth. It
consequences for the type and extent of condition-           emphasized that, with few exceptions, the state in the
ality associated with development lending.Again in           economies studied had taken an activist role to stimu-
                                                             late risk taking in both the private and the public sector.
this case, more rigorous economic analysis should            It concluded that while highly successful in East Asia,
help to distinguish what are binding constraints,            the institutions needed for replicating this activist role
and thus to inform decisions. The record suggests            may not be present in other contexts.
that forecasts need to be realistic and mindful of the    5. It has been argued, for example, that the increase in
forces driving growth.                                       India’s growth rate in the early 1980s was less the result
I N T RO D U C T I O N                                                                                                   27



    of the reforms introduced at that time than of the pri-          mirror image of governance: bad governance invariably
    vate sector’s changing expectations regarding the                leads to corruption; but corruption can likewise perpe-
    future—where the government was credible in ensur-               trate bad governance.
    ing reduced expropriation risks and a more welcoming          7. The principal delegates the implementation of a task to
    environment (Rodrik and Subramanian 2004).                       an agent but must monitor the agent efficiently to
 6. Public sector governance refers to how the state                 ensure that the task is accomplished.
    acquires and exercises the authority to provide and           8. A recent World Bank research project focusing on indi-
    manage public goods and services. Corruption, which              vidual country experiences is Aid and Reform in Africa
    refers to the use of public office for private gain, is the      (2001).