iC 2: = A_ jH c z <^ii ~~~~~~~~~~~~.9 - --(-x | rr _ N)v k f ,_ - Transition The First Ten Years Analysis and Lessons for Eastern Europe and the Former Soviet Union Transition The First Ten Years Analysis and Lessons for Eastern Europe and the Former Soviet Union THE WORLD BANK Washington, D.C. © 2002 The International Bank for Reconstruction and Development/The World Bank 1818 H Street, NW Washington, DC 20433 All rights reserved. 1 2345050403 02 The findings, interpretations, and conclusions expressed here are those of the author(s) and do not necessarily reflect the views of the Board of Executive Directors of the World Bank or the governments they represent. The World Bank cannot guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply on the part of the World Bank any judgment of the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is copyrighted. No part of this work may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or inclusion in any information storage and retrieval system, without the prior written permission of the World Bank. The World Bank encourages dissemination of its work and will normally grant permission promptly. For permission to photocopy or reprint, please send a request with complete information to the Copyright Clearance Center, Inc, 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978- 750-8400, fax 978-750-4470, www.copyright.com All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org ISBN 0-8213-5038-2 Libraty of Congress Cataloging-in-Publication Data has been applied for. Contents Foreword ....................................................... ix Acknowledgments ....................................................... xi Overview ................................... ................................................................. xiii The Quest for Growth: Promoting Discipline and Encouragement .............. ................... xv The Flipside: Protection and Discouragement ............................................................ xvii Shading the Classification ............................................................ xvii New Enterprises Spur Economic Growth ....................................................... ..... xviii When Is Transition Over? ............................................................. xix Do Central Europe and the Baltics Point the Way Forward? ..................... ..................... xix Can We Have It Both Ways: Encouragement without Discipline? ................................... xx Learning from China? ............................................................. xxi Institutions Are Important, but So Are Policies ............................................................ xxi The Political Economy of Discipline and Encouragement ......................... .................... xxii Shifting Policy Priorities to Account for Experience and New Conditions .......... ......... xxiv Conclusions ............................................................ xxviii Annex 1. Discipline and Encouragement: The Reform Agenda ............... .................. xxix Part 1. The First Decade in Transition .......................... .......................... 1 1. How Did Transition Economies Perform? .3 Output Fell Sharply .3 Industry Shrank-Services Grew .5 Private Enterprises Overtook the State Sector .6 Exports Rose-Moving Toward Industrial Countries .6 Poverty Increased Sharply .6 Contents 2. Explaining Variation in Output Performance ........................................................ 11 Did Initial Conditions Affect Performance? ............................................................... 11 External Economic Shocks Delayed Recovery ............................................................... 13 Policies-Do They Matter? .............................................................. 13 What Initial Conditions Matter and When Do They Matter? ......................................... 15 What If Policies Themselves Are Endogenous? .............................................................. 16 Does the Speed of Reform Matter? .............................................................. 16 Annex 2.1. Summary of Cross-Country Empirical Literature on Growth in Transition Economies .............................................................. 16 Annex 2.2. Additional Empirical Analysis .............................................................. 19 Part 2. Policy and Institutional Challenges Ahead .21 3. The Quest for Growth ........................ 23 A Tale of Two Approaches ........................... 26 The Associated Fiscal Adjustment ............................ 29 ... And the Role of Labor Markets ...................... 30 4. Discipline and Encouragement ........................................................ 33 New Enterprises Drive the Transition ............................ .................................. 39 A Small Number of High-Productivity Small Enterprises Is Not Enough .......... .............. 41 Can We Have It Both Ways, That Is, Protecting Old Inherited Enterprises and Encouraging New Enterprises? ...................... ........................................ 45 Annex 4.1. Assumptions for Small and New Enterprises ....................... ....................... 48 Annex 4.2. Implications of the Higher Productivity of SMEs ................. ...................... 50 5. Imposing Discipline ........................................................ 53 Soft Budget Constraints Can Create Macroeconomic Crises .................... ....................... 53 Nonpayments Weaken the Incentives for Efficiency and Restructuring ........... ................ 54 Exit Mechanisms-Implement Now, Revise Later .......................................................... 56 Competition Is Linked to Innovation and Growth .......................................................... 56 6. Extending Encouragement ........................................................ 59 Corruption and Anticompetitive Practices Mar the Investment Climate ........... .............. 59 Enterprises Lack Confidence in Legal and Judicial Institutions ................. ...................... 61 Financial Deepening Is Slow but Progressing .............................................................. 62 Privatization Attracts Foreign Direct Investment, and Positive Spillovers Follow ........ .... 67 Tax Reform: Broadening the Base and Lowering Rates ................................ ................... 67 Intergovernmental Fiscal Relations Supporting Discipline and Encouragement ........ ...... 68 7. Privatization: Lessons and Agenda for the Future .................................................... 71 Traditional Privatization or Rapid Privatization? ............................................................ 73 Why Countries Did What They Did ......................... ..................................... 73 Were Vouchers a Mistake and Other "What Ifs" .............................................................. 78 Summarizing Lessons .............................................................. 79 8. Supportive Social Policies ........................................................ 81 Reforming Pension Systems .............................................................. 81 vi Contents Social Assistance Should Protect Children and the Most Destitute, Adding More as Budgets Allow ............................................................. 83 Severe Cuts Have Compromised the Quality of Education ..................... ........................ 84 Containing Costs Will Make Health Care Affordable for Those Who Need It Most ...... 86 Part 3. The Political Economy of Discipline and Encouragement .. 89 9. The Winners and Losers from Discipline and Encouragement ................................. 91 Who Wins and Who Loses? .................. .................................................. 92 The Government Must Be Credible and Able to Constrain Oligarchs and Insiders ......... 94 10. Classifying Political Systems in Transition .............................................................. 97 Competitive Democracies Have High Political Contestability ... ..................................... 99 ... and High Government Turnover. 101 11. Political Systems Influence the Choice of Economic Reforms ................................. 103 Political Systems Create Rent-Seeking Opportunities .................................................... 105 How Do Political Systems Affect Economic Reform? ................................................... 107 12. Confronting the Political Challenge ............................................................ 1111 For Concentrated Political Regimes, Mobilizing Potential Winners .............................. 112 For War-Torn Political Systems, Restoring Stability and Reducing Uncertainty ............. 114 For Noncompetitive Political Systems, Taking Advantage of State Capacity ................. 115 For Competitive Democracies, Using Momentum to Build Coalitions for Reform ........ 115 Conclusion ................................................................... 1'16 Selected Bibliographic Guide to the Political Economy of Transition ............................. 117 References ............................................................ 125 Boxes 1 Increased Inequality .................................................................. xiv 1.1 Limits of GDP Statistics for Transition Economies .................................................................. 8 2.1 The Regional Impact of the Global Financial Crisis and Recovery ........................................ 13 3.1 The Business Environment and Enterprise Performance Survey .................. ........................... 24 3.2 The Problem of Tunneling .................................................................. 27 4.1 Can the CIS Learn from China's Reform Experience? ............................................................ 35 4.2 The German Experience .................................................................. 37 4.3 Belarus .................................................................. 46 5.1 External Debt and Fiscal Sustainability in the Low-Income CIS Countries ............................. 55 6.1 Reducing the Cost of Entry and Doing Business in Armenia ..................... ............................ 60 7.1 Historical Counterfactuals: Mass Privatization in Russia ....................................................... 76 Figures 1 Winners and Losers from Reform .xxii 1.1 Changes in Real Output, 1990-2001 .4 1.2 Output Growth Rates, 1990-2001. 4 2.1 Progress in Policy Reform, 1990s .14 vii Contents 3.1 Productivity Distribution of Old, Restructured, and New Enterprises ................................... 24 3.2 Performance of Old and New Enterprises, 1996-99 .26 4.1 Private Sector Share in GDP, 1999 .40 4.2 Share of Employment in Small Enterprises, 1989-98 .41 4.3 Share of Value Added in Small Enterprises, 1989-98 .41 4.4 Value Added per Employee in Small Enterprises, 1998 .42 4.5 Index of GDP and Shares of Value Added and Employment Accounted for by Small Enterprises, 1989-98 .43 4.6 Employment and GDP, 1990-98 .............................................................. 44 4.7 Soft Budget Constraints and Employment in Small Enterprises, 2000 ................................... 48 A4.1 Factor Price Frontier: SMEs and the Rest of the Economy .................................................... 50 A4.2 Efficiency Gain from 10 Percent Factor Reallocation to the SME Sector ............ ................... 51 6.1 Insecurity of Property Rights in Transition Economies, 1999 ................................................ 61 6.2 Quality of Legal Drafting in Transition Economies, 1999 ..................................................... 62 6.3 Quality of Judiciary in Transition Economies .............................................................. 63 6.4 Domestic Credit by Deposit Money Banks to the Private Sector, 1998 .................................. 64 6.5 Stock Market Capitalization and Per Capita Income, 1998 ................................................... 65 6.6 Operating Costs to Total Assets in the Banking Sector, 1997 ................................................ 65 6.7 Interest Rate Spreads, 1998 .............................................................. 66 6.8 Cumulative Foreign Direct Investment Per Capita and Employment in Small Enterprises, 1998 ............................................................... 68 8.1 Public Expenditures on Education in Transition Economies, 1998 ........................................ 85 8.2 Health Expenditures, 1998 .............................................................. 86 9.1 Winners and Losers from Reform .............................................................. 93 10.1 Classifying Political Systems in Transition Economies, 1990-99 ........................................... 98 10.2 Veto Points Index, 1989-99 ............................................................... 100 10.3 Main Political Executive Turnovers, 1989-99 .............................................................. 102 11.1 Political Systems and Economic Reform Outcomes, 2000 ................................................... 104 11.2 State Capture Index, 1999 .............................................................. 106 Tables 1.1 The Transition Recession ............................................................... 5 1.2 Composition of Output, 1990-91 and 1997-98 .6 1.3 Private Sector Growth, 1990s .6 1.4 Export Growth and Destination, 1990s .7 1.5 Main Recipients of Foreign Direct Investment, 1992-99. 7 1.6 Average Poverty Rates, 1990 and 1998 .8 1.7 Changes in Inequality during the Transition, Various Years .................................................... 9 A2.1 Regression of Average Growth on Initial Conditions, 1990-99 ............................................. 20 A2.2 Regression of Annual Output Growth on Policies and Initial Conditions Allowing for Differential Effects Early in Transition, 1990-99 ...................................................... 20 4.1 SMEs Have Higher Labor Productivity, 1998 ................................................................... 42 A4.1 Differences between New Enterprises and Small Enterprises, 1995 and 1998 ....................... 49 7.1 Methods of Privatization of Medium-Sized and Large Enterprises ........................................ 75 8.1 Reform Options for Social Protection Programs ................................................................... 83 8.2 Student-Teacher Ratios in Basic Education, 1990 and 1997 .................................................. 85 viii Foreword T his study looks at lessons to be drawn from the ten-year experience of the transition countries in Eastern Europe and the former Soviet Union in the period 1991 to 2000. The World Bank's World Development Report 1996: From Plan to Market focused on the transition process during the first half of this period. It recognized that while initial conditions are critical, decisive and sustained reforms are important for recovery of growth and should be accompanied by social policies designed to protect the most vulnerable groups until growth takes hold. It highlighted the need to create institutions in support of markets, and it emphasized that investing in people is a key to growth. Today we have more evidence and data on the transition experience. The variability in growth performance across countries has intensified. Poverty and income inequality in some countries have increased to levels not foreseen earlier. In many countries, reform efforts that started in the early 1990s have been interrupted and in some cases even stalled. As a result, output recovery in some of the transition countries was sharply reversed during the second half of the 1990s. Many of the prescriptions of the 1996 World Development Report continue to be valid today. At the most general level, the present study confirms that while initial conditions were critical for ex- plaining the output decline at the start of transition, the intensity of reform policies explains the variability in the recovery of output thereafter. Beyond this, important new lessons highlight some key tradeoffs facing countries in transition that can be translated into priorities for policy. First, this study highlights the key role of the entry and growth of new firms, particularly small- and medium-size enterprises, in generating economic growth and in creating employment. The growth of new firms depends in part on direct policies to encourage entry-what this report calls the encouragre- ment strategy. Does this mean that policymakers can focus on encouraging the new sector while post- poning the pain of liquidating and restructuring the old sector to a later time when a cushion has been put in place? Not so. The report shows that to succeed the encouragement strategy needs to be accompa- nied by a strategy of discipline; that is, policies that impose hard budget constraints on the old-large enterprises that remain from pretransition days. Soft budget constraints that allow these enterprises to not pay their taxes, social security contributions, and bank debts undermine the level playing field be- tween different kinds of enterprises and have also been at the root of explosive fiscal and banking crises. ix Foreword The combined encouragement-and-discipline A third lesson involves the recognition that strategy that this study proposes is crucial to re- winners from the early stages of reforms such as allocate assets to more productive uses and to liberalization and privatization may oppose sub- provide economic space for new emerging firms. sequent reform steps when these reduce their This perspective of encouragement and discipline initially substantial, but potentially temporary allows the report to shed light on issues such as benefits or rents. These winners will tend to re- privatization and fiscal policy, which have been sist reforms such as further trade liberalization; major elements of economic reform in transition entry of new competitors, including foreign di- countries. Privatization is important to the ex- rect investment; and legislation protecting mi- tent that it facilitates hard budget constraints on nority shareholders and creditors to the extent old enterprises and creates incentives for produc- they reduce such rents. Furthermore, if the rents tion and innovation rather than asset stripping are large as a fraction of total gross domestic and rent seeking in new enterprises. product, which is usually the case in natural A second lesson concerns the need to develop resource- and energy-rich countries, these early or strengthen legal and regulatory institutions to winners may capture the state and force the oversee the management and governance of en- economy into a trap of a low-level reform equi- terprises, both those in the new private sector and librium. Understanding how such reform traps those remaining under state ownership. In coun- arise and how to break out of them is an impor- tries where direct sales of assets to strategic inves- tant area of inquiry addressed by this study. To tors-a preferred method of privatization-was the extent underlying political economy consid- unavailable, policymakers were confronted with erations permit a reform-minded team room to the often difficult choice between privatization to maneuver, fiscal policy can play an important ineffective owners in a context of weak corporate role here, by redirecting support away from ail- governance on the one hand and continued state ing enterprises and toward worker training and ownership until strategic investors could be found severance payments; by divesting social assets on the other. However, continued state ownership such as housing, child care, and health facilities does not lead to efficient stewardship of enter- from enterprises to governments; and by main- prise assets unless there is a political commitment taining the high levels of human capital with to transparent privatization outcomes and a mini- which countries entered transition. mum institutional capacity to prevent asset strip- The experience of transition from centrally ping by managers during the intervening period. planned to a market economy is an historically In either event, rules to protect minority share- unprecedented process, and one that is by no holders; rules against insider deals and conflicts means finished in many countries in Eastern of interest; adequate accounting, auditing, and dis- Europe and the former Soviet Union. We hope closure standards; and takeover, insolvency, and that this study will encourage further discussion collateral legislation, together with development among policymakers and think tanks in the tran- of enforcement capacity, are key to preventing sition countries and will assist their dialogue with asset stripping that reduces the true long-term external donors and advisers on how to better value and competitiveness of a firm. support the transition process. Johannes E Linn Vice President Europe and Central Asia Region x Acknowledgments T his report was prepared by a team led by Pradeep Mitra and Marcelo Selowsky that com- prised Joel Hellman, Ricardo Martin, Christof Ruehl, and Asad Alam. Important contribu- tions to specific sections were made both from within the World Bank and by external part- ners. Contributors from the World Bank include Harry Broadman, Alan Gelb, Christine Jones, Ira Lieberman, John Nellis, Samuel Otoo, Ana Revenga, Roberto Rocha, Randi Ryterman, Sergei Shatalov, Mark Sundberg, and Marina Wes. External contributors include Barry Eichengreen of the University of California at Berkeley, Simon Johnson of the Massachusetts Institute of Technology, and Christian Mumssen and Thomas Richardson, both at the International Monetary Fund. Bruce Ross-Larson was the principal editor. The selected bibliography was prepared by Graeme Robertson, Columbia University. The team is also indebted to many country economists in the World Bank's Poverty Reduction and Economic Management Unit in the Europe and Central Asia region for background notes and analy- ses of specific transition economies: Sebnem Akkaya, Ritu Anand, Robert J. Anderson, Carlos Cavalcanti, Mark Davis, Ilker Domac, Mansour Farsad, Lev Freinkman, Reza Ghasimi, Ardo Hansson, Erika Jorgensen, Brian Pinto, Rosanna Polastri, Carlos Silva-Jauregui, Andriy Storozhuk, Jos Verbeek, Dusan Vujovic, and Leila Zlaoui. An earlier version of the report was first discussed at a seminar at the World Bank chaired by Johamnes Linn. Peer reviewers were Nicholas Stern, Alan Gelb, and Marek Dabrowski. The team thanks all of them for their comments and suggestions. The team also thanks the Bertelsmann Foundation for spon- soring a seminar for policymakers in transition economies in May 2000 in Kronberg, Germany that originally helped to motivate and sharpen many of the issues addressed in this report. Participants in- cluded Vahram Avanessian, Marek Belka, Lajos Bokros, Aleksandar Bozhkov, Oraz A. Jandosov, Vladimnir A. Mau, Ivan Miklos, Viktor Pinzenik, Ion Sturza, and Marat Sultanov. Earlier versions of the report were presented at seminars at the Institute for Economies in Transition in Moscow, at the United Na- tions Commission for Europe in Geneva, at the Center for Economic Research and Graduate Education of Charles University and the Economic Institute of Sciences of the Czech Republic in Prague, and at a conference on "Trade, Integration, and Transition" held in Bela Balassa's memory in Budapest in Octo- ber 2001, where it benefited from the comments of Jinos Kornai and Andras Nagy. Book design, production, and dissemination were coordinated by the World Bank Publications team. xi Overview A dozen years have elapsed since the world witnessed the euphoria greeting the fall of the Berlin wall. Ten years ago last summer, Boris Yeltsin claimed his place in history by climbing atop a tank in the streets of Moscow in defiance of the last defenders of an imploding Soviet Union. Thus did the march from plan to market capture the world's imagination. In those heady days, famously dubbed a time of "extraordinary politics" by a leading reformer from the region, everything seemed possible (Balcerowicz 1995). At the beginning of the new millennium, a profound divide lies between Central and Southeastern Europe and the Baltics (CSB) and the Commonwealth of Independent States (CIS).' In the CSB, of fi- cially measured gross domestic product (GDP) bounced back from a transition recession, recovered to its 1990 level by 1998, and exceeded that level by 6 percent in 2000. However, in the CIS GDP in 2000 stood at only 63 percent of its 1990 level. While GDP in Poland, the most populous country in the CSB, increased by more than 40 percent between 1990 and 1999, it shrank by 40 percent during the same period in the Russian Federation, the most populous country in the CIS.2'3 In 1998 one in five people in the region survived on less than US$2.15 a day, a standard poverty line.4 A decade before fewer than one in 25 lived in such absolute poverty. While absolute incolne deprivation at those levels is virtually nonexistent in many Central European countries, its incidence is as high as 68 percent in Tajikistan, 50 percent in the Kyrgyz Republic, and 40 percent in Armenia. Inequality, which has just barely increased in Central Europe since the onset of transition, has in- creased so much in CIS countries such as Armenia, the Kyrgyz Republic, and Russia that they have come to rival the most unequal countries in the world (box 1). A similar divide runs across the political landscape as well. Competitive democracies-underpinned by widespread political rights to participate in multiparty elections and an extensive range of civil liber- ties-have taken root in nearly all of Central Europe and the Baltics. In contrast, limitations on rights to participate in elections and constraints on civil liberties during at least some period of the transition have concentrated political power in many countries in the CIS and in Southeastern Europe. Nevertheless, this concentration has been associated with diminished state capacity to provide public goods needed lfor the market economy as a result of corruption, weak public sector management, and in some cases war xiii Overview Box 1. Increased Inequality The countries of Europe and Central Asia started the transition with some of the lowest levels of inequality in the world. Since then, however, inequality has increased steadily in all transiton economies and dramatically in some of them (see figure A). Countries such as Armenia, the Kyrgyz Republic, Moldova, and Russia are now among the most unequal in the world, with Gini coefficients (a standard measure of inequality) nearly twice their pretransition levels. It is tempting to attribute increasing inequality to reforms and liberalization. But this is only part of the story, While in- FIGURE A. equaiity has increased almost everywhere, the more advanced reformers show much more equal, rather than more unequal, Changes in Income Inequaliy in Selected outcomes, compared with less advanced reformers. This differ- Transiton Economies ence cannot be solely explained by different conditions across the countries at the start of transition. Czech Republic Rather, a recent World Bank study (2000b) shows that Hungary positive developments largely explain the rise in inequality in Slovenia the CSB: rising retums to education, decompressing wages, and Belarus emerging returns to risktaking and entrepreneurship. These Latvia forces are welcome despite the increase in inequality, because Ukraine they signal that the market is now rewarding skills and effort, Estonia as in more mature market economies. In the CSB, moreover, Poland strong social transfers and redistribution mechanisms have Lithuania dampened the rise in education premiums and wage disper- Croatia sion, in line with the demands these societies have placed on Moldova their governments for such measures. Georgia The experience of the CIS is very different Rising educa- Kyrgyz Republic tion premiums and wage dispersion explain very little of the Russian Federation rise in inequality. In Armenia, Georgia, the Kyrgyz Republic, Tajikistan Moldova, and Russia income differences linked to educational Armenia achievement explain less than 5 percent of inequality, com- pared with 20 percent in Slovenia and 15 percent in Hungary 0 0.2 0.4 0.6 0.8 and Poland. The causes of the huge rise in inequality lie: Gini coefficient * In the prevalence of widespread corruption and rent seek- ing. There is a strong correlation between higher corruption * 1987-90 * 1996-99 and higher inequality (and higher poverty) in the region. Source: World Bank (2000b). The poor are disproportionately affected by corruption (World Bank 2000c). * In the capture of the state by narrow vested interests, which have modified policy to their advantage, often at a high social cost These interests have been able to limit competition and concentrate their economic power through such mechanisms as special licenses and monopolies. They have undermined state institutions and blocked reforms that would serve the pubilic good. * In the resulting collapse of formal wages and income opportunities. Wages at old jobs have collapsed or are not paid, while new formal job opportunities are stifled by the lack of competitive markets and by the pervasiveness of corruption. People, except for a privileged few, are largely stuck in their low-paying (and sometimes nonpaying) jobs. To make ends meet they supplement their incomes with diverse forms of self-employment. much of it subsistence agricuiture in small household plots. Throughout the CIS, eamings from such small plots account for 40-70 percent of total household eamings. Access to connections and informal networks and an ability to pay are key to finding a job and getting ahead. This has led to highly unequal outcomes. xiv Overview and civil strife. Outside Central Europe and the economies have some common characteris- Baltics the optimism pervading the beginning of tics that make those lessons applicable today? transition has been tempered by the harsh eco- * If the advantages of economic reform are so nomic realities of its first decade. obvious, why do countries mired in a no However, the starkness of this binary picture man's land between centrally planned and needs to be softened; growth outcomes have var- market economies not adopt them? How ied significantly even within the two broad might political support for reform be built in groups of countries. Furthermore, all countries those countries? went through the transitional recession, which * In what key respects should policy advice t.o caused real GDP to dip from its 1990 levels by transition economies be modified to reflect nearly 15 percent in the CSB and by more than experience from the first decade and the new 40 percent in the CIS. conditions prevailing today? Of the CSB countries, Hungary, Latvia, Po- land, Slovenia, and to some extent Estonia and This report seeks answers to these questions. Lithuania have enjoyed several years of uninter- rupted growth. By contrast, growth in Bulgaria and omana wa shaply nterupte by en- The Quest for Growth: Promoting Discipline and Romania was sharply interrupted by seri- anEcorgm t ous macroeconomic crises brought on by insuf- ficient structural reform in the mid-1990s, and T he focus on economic growth needs to be GDP in 2000 stood at four-fifths its 1990 level. T put into a broader perspective. For much The Czech Republic had a similar but less severe of the CIS and Southeastern Europe the restora- experience; GDP declined during 1997-99 be- tion of sustained growth is a key priority. Wit]h- cause of a macroeconomic crisis with structural out it these countries will not generate income- origins. In fact, the Czech Republic was the only earning opportunities for households. Nor will country in Central Europe that had not reached they have the resources needed to provide basic its 1990 GDP level by 2000. public goods such as legal and judicial systems, In the CIS such early reformers as Armenia, secure property rights, and basic infrastructure; Georgia, and the Kyrgyz Republic, whose GDP maintain essential investments in education and fell steeply, and such nonreformers as Belarus health; or set up social safety nets targeted to and Uzbekistan, where the decline in GDP was the most vulnerable. In this respect transition smaller, have been growing in the past five years. economies are no different from other economies. But Russia, barring a short-lived upturn in 1997, Continued growth is also important for the did not begin to grow until 1999, while Ukraine leading reformers in Central Europe and the did not return to growth until 2000. Baltics. While all the Central European cou]- The wide variation in transition across the tries cxccpt the Czech Republic had surpassed region raises questions: their 1990 GDP by 2000, per capita incomes in the three wealthiest countries aspiring to Euro- * Why has the growth of some transition pean Union accession were still only 68 percent economies been better than that of others? of the European Union average for Slovenia,59 To what extent can these differences be as- percent for the Czech Republic, and 49 percent cribed to economic policy choices rather than for Hungary. However, an exclusive focus on circumstances at the start of transition or growth, while providing basic public goods and external economic shocks? protecting the most vulnerable, is not enough * Do the policy lessons from the countries that for them. They need to consolidate the gains of enjoyed several years of rapid growth con- the first decade of transition and address "sec- tinue to be relevant for the CIS and South- ond generation" reform issues. They have tlo eastern Europe, which have made less secure control over quasi-fiscal and contingent progress with the transition? Do transition liabilities. They have to undertake reforms in xv Overview labor and financial markets to allow the ben- resources than they produce, and restructured efits of growth to be more widely shared. They and new enterprises, which increase growth. The also have to restructure social expenditures to fall in growth is initially dominated by the drag make them fiscally more affordable without im- of old enterprises, which leads to a period of de- pairing the effectiveness of the social safety net. cline. With time, if the business environment fa- Several of these reforms overlap with those re- vors production and innovation rather than rent quired to join the European Union.5 seeking, restructured and new enterprises gain This report is primarily about economic the critical mass to overcome the negative effects growth. But the focus is not meant to be exclu- of old enterprises, leading to recovery and sive. Two companion reports on poverty and economywide growth. inequality and on anticorruption deal with is- The initial conditions of geography, history, sues particularly important in the transition and price and output distortions at the start of (World Bank 2000b, c). transition and the external economic shocks aris- The common heritage of socialism implied ing from the breakup of the Soviet Union, war, that all countries in the region began their tran- and civil strife were of course important. How- sition with a production system adapted not to ever, analysis done for this report shows that ini- a competitive environment but to the exigen- tial conditions were significant factors during the cies of a command economy. External liberal- initial period of output decline (1990-94), rather ization at the beginning of transition generated than throughout the full ten years of transition, significant productivity differences across sec- even after accounting for differences across coun- tors and enterprises in a production system tries in policy reform and the impact of external based on cheap energy and subsidized transport. economic shocks. That analysis also demon- For example, energy intensity, measured as the strates that policy reforms have been significant amount of energy used per unit of GDP, was factors in differences among countries in the 0.95 tons of oil equivalent per US$1,000 of GDP speed of economic recovery, once due account is in the Soviet Union in 1985, compared with taken of differences among countries in initial 0.50 tons of oil per US$1,000 of GDP in OECD conditions and the impact of external economic countries (IMF and others 1991). In April 1992, shocks. Furthermore, market-oriented policy re- after Russia had adjusted the price of oil sev- form not only speeded up economic recovery and eral times, its domestic price was still only 3 promoted growth in the medium term, but it also percent of the world price (Tarr 1994). Many mitigated the effects of the transitional recession sectors and enterprises were not viable after in the short term. How effective policies have price liberalization. been in disciplining the old sector and encour- Two challenges had to be confronted: aging the new therefore holds the key to under- First, the imposiionofmarketdistanding why growth has been better in some * First, the imposition of market discipline on trnito ecnme.hni tes transition economies than in others. inherited enterprises so that they would face Discipline forces old enterprises to release the incentive to restructure and, in so doing, assets and labor, which are then potentially become more productive and able to com- petatote nrew prics.Filue tndo seo shou available to restructured and new enterprises. lead to closure. It does this by hardening budget constraints, l Second, nlosuragementof.tecratioo introducing competition in product markets, * Second, encouragement of the creation of prvdn extmca'm.admntrn . . ~~~~~providing exit mechanisms, and monitoring new enterprises willing and able to compete in the marketplace without seeking special managerial behavior to generate hcenthves for faor fro th,tae production and innovation (rather than for as- favors from the state. set stripping and theft).6 Discipline also pushes Economic growth reflects the interplay be- old enterprises to divest themselves of such so- tween old enterprises in need of state support, cial assets as housing, health clinics, and kin- which reduce growth by absorbing more dergartens to local governments, shifting the xvi Overview locus of social protection away from enterprises Support for the old sector is ultimately fi- to governments. The social safety net then needs nanced through taxes on households, new enter- to be strengthened to ensure that labor shed by prises, and enterprises that have restructured suc- contracting enterprises and other losers from cessfully to survive the market test. The social reform do not fall into poverty, while not erod- safety net is unable to prevent people from mov- ing these workers' incentives to find employ- ing into subsistence and low-productivity activi- ment in new enterprises. ties to ensure their survival. Such a protect-and- Encouragement entails policies to create an discourage strategy creates an environment where attractive and competitive investment climate in resource transfers tend to flow in a direction op- which restructured and new enterprises have in- posite to that in a discipline-and-encourage envi- centives to absorb labor and assets rendered in- ronment. Transfers from efficient to inefficient expensive by the downsizing and to invest in enterprises and sectors undermine the credibility expansion. These policies include reducing ex- of government policy, with detrimental conse- cessively high marginal tax rates, simplifying quences for the economy. regulatory procedures, establishing secure prop- The logic of discipline and encouragement erty rights, and providing basic infrastructure, is intended to apply broadly to the production while maintaining a level playing field among of goods. But it may also be applied, with some old, restructured, and new enterprises. At the modification, to the banking system, an impor- same time the policy environment must provide tant part of the investment climate required to incentives for wealth creation rather than rent attract new enterprises. Hard budget con- seeking and asset stripping by new enterprises. straints on state banks and a credible threat of This mode of adjustment broadly corresponds exit for failed banks are essential to discipline to the experience to date of economies in Cen- banks and enterprises alike. However, encour- tral Europe and the Baltics. agement does not always imply free entry of new banks. Free entry could help make the banking sector competitive, but potential en- trants must satisfy prudential norms, such as T he disposition of assets among old, restruc- those for minimum capital requirements and T tured, and new enterprises also provides a capital adequacy. It is also important that ex- useful perspective on the experience of many of pansion of the banking system not outpace the the CIS countries and, to some extent, South- capacity for effective supervision and growth eastern Europe. These countries have tended to in the number of creditworthy borrowers. protect rather than discipline old enterprises through subsidies granted through the budget, Shading the Classification energy consumption, and the banking sectors. Where institutions of public and corporate gov- T he juxtaposition of discipline-and- ernance are not strong enough, asset stripping, T encouragement and protection-and- theft, and other violations of property and share- discouragement highlights two contrasting holder rights become widespread. Entry of new modes of adjustment. In reality, country out- enterprises is discouraged-or, at best, only se- comes span a range of intermediate possibilities, lectively encouraged-because of opposition depending on whether liberalization was imple- from entrenched interests that would lose from mented, hard budget constraints imposed, and further liberalization of entry. Furthermore, be- an enabling business environment promoted and cause of a poor investment climate where tax in what order and how vigorously. rates are high, licensing and registration proce- dures are open to abuse, and the legal and judi- * The discipline of hard budget constraints and cial system is weak, corruption becomes a seri- institutions of corporate governance to moni- ous obstacle to the growth of new enterprises. tor managerial behavior and encouragement xvii Overview through liberalization and a climate hospi- leading and lagging reformers in the region, drawn table to domestic and foreign investment are from the World Bank's database on small and perhaps seen most clearly in Estonia, Hun- medium-size enterprises (SMEs). It is also sup- gary, and Poland. ported by a comparison between old and new en- * Even in the broad category of discipline and terprises in the Business Environment and Enter- encouragement, however, softer budget con- prise Performance Survey, conducted jointly by straints and hence less discipline prevailed for the European Bank for Reconstruction and De- a long time in the Czech Republic, Lithuania, velopment and the World Bank in 1999 (see box and the Slovak Republic. 3.1). The survey finds that new enterprises out- * Bulgaria, the Kyrgyz Republic, Moldova, perform old enterprises in sales, exports, invest- Romania, Russia, and Ukraine liberalized ment, and employment (chapter 3). Thus a trans- their economies, but for a long time failed to fer of resources from old enterprises to new can maintain discipline through hard budget con- be a source of growth. Whether that potential is straints. They were also unable to contain realized depends on the discipline imposed on old tunneling, the expropriation of assets and enterprises to shed resources and the encourage- income belonging to minority shareholders, ment extended to new enterprises to absorb them. and theft through either rule of law or ad- The interaction between old and new enter- ministrative control. Though many of these prises is key to economic growth. The share of countries did encourage new entry early in total employment and value added accounted for the transition, the capture of the state by a by small enterprises (defined as employing fewer narrow set of vested enterprises-old enter- than 50 workers) as a proxy for new enterprises prises and well-connected early entrants-dis- divides transition economies into two groups.7 couraged further entry and created a poor In the Czech Republic, Hungary, Lithuania, and investment climate, resulting in a pattern of Poland, new enterprises grew very rapidly. They protection and selective encouragement. now account for 50 percent or more of employ- * Belarus, Turkmenistan, and Uzbekistan, ment, the average for the European Union, and which have undertaken some liberalization, for between 55 and 65 percent of value added. but have not imposed hard budget constraints, But in Kazakhstan, Russia, and Ukraine, which strongly discourage new entry. Policies such have seen modest or no growth in new enter- as access to foreign exchange and credit on prises, the share of employment has stayed at or special terms soften budget constraints for below 20 percent and the share of value added state enterprises. But continuing reliance on has stayed between 20 and 30 percent. centralized political power and mechanisms Hungary and Poland saw a sharp and early of administrative control inherited from the decline in employment and a rapid demise of the command economy did limit extensive asset old sector, which initially made resources avail- stripping and other forms of theft at the en- able cheaply to the new sector. Such discipline is terprise level. That led to a situation incorpo- important but insufficient; encouragement is also rating some elements of discipline in an oth- needed. Growth takes off only when the new erwise strongly protective stance, together sector evolves from a passive receptacle for ab- with discouragement of entry. sorbing resources into an active competitor, rap- idly increasing its share of employment and at- New Enterprises Spur Economic Growth tracting the most qualified workers. The evidence suggests that new enterprises must reach a thres- T he growth-enhancing effects of new enter- hold of around 40 percent in their contribution T prises and the growth-restraining effects of to employment before they can become an en- the old broadly suggest that new enterprises in gine of growth. In Russia and Ukraine, where transition economies are more productive than the contribution of the new sector to employ- old enterprises. This is supported by data from ment is well below the threshold, a large pro- 10 transition economies covering both the portion of the labor force remains mired in old, xviii Overview unrestructured enterprises not generating in- restructured, or old? They can also be distin- creases in productivity. The new sector has not guished by economic performance: are they pro- emerged as a source of growth. ductive? Furthermore, history and performance New enterprises are more productive than old are related. New enterprises are expected to Ibe ones, but productivity differences diminish with more productive than restructured enterprises, transition. The difference is greater in Kazakhstan, which are expected to be more productive than Russia, and Ukraine than in the Czech Republic, old enterprises. As markets develop and resources Hungary, Latvia, Lithuania, and Poland. Why? are allowed to flow to their most valued uses, It is because closure and restructuring can raise the role of history progressively weakens, and the productivity of factors in the old sector and differences in productivity arising from member- because fast growth of enterprises and employ- ship in any of the categories tends to disappear, ment can reduce the productivity of factors in the consistent with the evidence on the behavior of new sectors. Thus a comparison of labor produc- differences in productivity. tivity shows a difference in favor of new enter- This is not to suggest that differences in pro- prises of more than 100 percent in Ukraine, where ductivity across enterprises will disappear alto- the contribution of the new sector to total em- gether. These differences always exist as a result ployment is 17 percent. That difference narrows of technical innovation and new export market to just more than 40 percent in Hungary, where penetration, among other factors. However, the the contribution of the new sector to total em- variation in productivity could not be systern- ployment is 55 percent.8 atically attributed to the enterprises' historically Creating a policy environment that disciplines determined categories: old, restructured, and low-productivity old enterprises into releasing new. When that distinguishing characteristic is resources and encourages high-productivity new lost in a country, the transition can be taken to enterprises to absorb those resources and to un- be over. At that point, the economic issues and dertake new investment, without tilting the play- problems policymakers must deal with are rio ing field in favor of any particular type of enter- longer specific to transition. At what level of per prise and while strengthening the social safety capita income will this occur? The answer de- net to protect the most vulnerable, is central to pends on the success of disciplining the old sec- economic growth in transition economies. This tor and encouraging the new one. It also dependls is the main lesson from the successful reformers on the success of the business environment in in Central Europe and the Baltics. attracting investment. When Is Transition Over? Do Central Europe and the Baltics Point the D o transition economies at different lev- Way Forward? els of gross national income per capita T he striking diversity in challenges and cir- (ranging in purchasing power parity lcumstances among countries that have not terms from US$2,100 in Moldova in 1999 to proceeded far in the transition-in particular, US$16,050inSloveniain2000 [WorldBank2001]) countries in the CIS and Southeastern Europe-- have anything in common that would make les- raises the question of whether 10 years after the sons from economies in the region leading in re- dissolution of the Soviet Union these countries form applicable to those lagging in reform? The can learn from the successful reforms in Central wide dispersion in the productivity of labor and Europe and the Baltics. capital across types of enterprises at the onset of Countries in the CIS and Southeastern Eu- transition and the erosion of those differences be- rope continue to face significant productivity tween old and new sectors during reform provide differences across old, restructured, and new a natural definition of the end of transition. enterprises characteristic of the transition and Enterprises in a typical transition economy are therefore a long way from the end of transi- can be distinguished by history: are they new, tion. Hence, the framework of discipline and xix Overview encouragement and its associated policy and in- nonperforming loans as a share of total bank- stitutional implications remain relevant for un- ing sector loans in the 1990s (in Romania, 34 derstanding what needs to be done to restore percent in 1998). These loans prevented the growth and protect the most vulnerable in these expansion of bank credit to new, small, and economies (the policy and institutional reforms politically less-connected enterprises. They also associated with discipline and encouragement triggered banking and macroeconomic crises are summarized in annex 1). that called for stabilization and a tightening However, the political context for pursuing of credit, hurting new enterprises. reform policies has changed greatly in a decade. * Protection of the old industrial sector in At the beginning of transition, implementing re- Belarus and Uzbekistan through specially fa- forms focused on overcoming the resistance of vorable foreign exchange regimes, directed the nomenklatura, whose political and economic credit, and high trade protection has meant privileges fueled support for the prevailing eco- that whatever credit and foreign exchange re- nomic system, and on building support for re- main are available only to new smaller enter- form among the newly mobilized public. But in prises at prices several times higher than what the past decade power over economic resources would have been paid in unified markets. In has shifted, often in a highly concentrated pat- Uzbekistan small enterprises have to pay three tern, from state bureaucrats to the private sector, times more for foreign exchange to finance even in much of the CIS and in Southeastern Eu- their imports than do large state enterprises. rope. That has made it easier for narrow special * Protection through tax and utility arrears in interests to capture the state and block further countries such as Georgia, the Kyrgyz Repub- reforms that may undermine short-term rents. lic, Moldova, Russia, Romania, and Ukraine Other problems, specific to individual coun- meant that new and more energy-efficient tries or subgroups of countries, have little to do enterprises were charged more to compen- with the transition but demand urgent resolu- sate for revenue losses from nonpayment by tion. Securing peace and inaugurating the pains- old, less energy-efficient enterprises to utili- taking task of nation building in the South ties in the energy sector. Tax exemptions for Caucasus and the Balkans, wracked by war and large enterprises and agricultural collectives civil strife, are priorities. So is controlling the in Ukraine, negotiated offsets to pay taxes in spread of tuberculosis and HIV/AIDS, which Russia, and tax avoidance in exchange for threaten millions of lives. But these challenges bribes by large enterprises in Georgia typi- are outside the scope of this report. cally worked to the disadvantage of new and smaller enterprises, which ended up paying Can We Have It Both Ways: Encouragement higher prices and bribes as a proportion of without Discipline? their annual revenue. T he implementation of policies associated The lack of a vibrant emerging private sec- with both discipline and encouragement tor, because of a policy of discouragement, lim- presents a challenge. Is it possible to downsize the its the outside options available to those in old old sector slowly while encouraging the new sec- enterprises. These limited private sector job op- tor and thus to avoid the pain of liquidation and tions increase the social cost of restructuring the restructuring until a cushion has been put in place? old enterprises, resulting in the need for addi- Encouragement without discipline will not work tional protection for the old sector. The comple- if old enterprises absorb resources that would oth- mentary relationship between discipline and en- erwise flow to new enterprises. For example: couragement also sheds light on why a relaxation in discipline-brought about, for example, by * Protection of state-owned enterprises and farm special treatment for powerful lobbies-is asso- collectives through the banking sector in Bul- ciated with selective rather than complete encour- garia and Romania led to a sharp increase in agement. It also helps explain why policy reform xx Overview must cover both discipline and encouragement, state enterprises remain to be fully recognized. thus proceeding along an ambitiously broad The share of nonperforming loans in the bank- front, and therefore why there are no magic bul- ing system, which served as a conduit for assis- lets in transition. tance to state enterprises, is between 30 and 40 percent of annual GDP. Addressing this prob- Learning from China? lem is likely to pose a major fiscal challenge. In sum, the transition economies of Eastern Europe T he success of encouraging entry of new en- and the former Soviet Union did not have the T terprises in China (where GDP per capita resources for a phased transition for state enter- grew 8 percent per year from 1978 to 1995, lift- prises, but they would be well advised to draw ing 200 million people out of absolute poverty) from China's experience the importance of en- without imposing significant discipline on state couraging new enterprises as a basis for wealth enterprises raises a question of the applicability creation and economic growth. of China's reform to the transition economies of Eastern Europe and the former Soviet Union. EaThroEughdifrpen channelse Chinar modulated t Institutions Are Important, but So Are Policies Through different channels China modulated the tradeoff between encouraging new enterprises arly on, Fischer and Gelb (1991) flagged and not imposing hard budget constraints on Ethe role of institutional reforms in the transi- existing state enterprises. The country reaped tion. Given that it is now argued that a key de- spectacular gains from liberalizing repressed sec- ficiency of the transition has been insufficient tors such as agriculture, which had surplus la- attention to building a market-friendly institu- bor, and rural industries and from a massive in- tional framework, reflecting on the experience flow of foreign direct investment. of East Germany's unification with West Ger- Part of these gains, helped by a high savings many in 1990 is instructive. It illustrates, among rate, could be transferred through the banking other things, how inappropriate policy choices system to finance loss-making state enterprises, can undermine performance even in the most which were far less important in China than in favorable of institutional environments. East most countries in Eastern Europe and the former Germany was able to adopt all the institutions Soviet Union. For example, only 19 percent of of West Germany without delay. It also received the Chinese labor force worked in the state sec- massive financial transfers, which averaged be- tor and were thus entitled to a range of social tween 40 and 60 percent of East Germany's benefits, compared with 90 percent in Russia. GDP over the period 1991 to 1997, and which Furthermore, tight political control over asset continue at levels exceeding 4 percent of West stripping, arbitraging between controlled and Germany's GDP per year. Furthermore, German market prices for private gain and corruption, reunification conferred on East Germany auto- together with some state capacity to manage pub- matic membership in the European Union. lic assets allowed China to move its loss-making Despite these considerable advantages, East state enterprises more slowly to market condi- Germany suffered an initial decline in GDP that tions at the same time that explosive growth of was deeper than its transition economy neigh- new enterprises took place. If a substantial in- bors' declines, and it has experienced one of the flow of resources, for example, from liberaliz- slowest GDP growth rates in Europe, disappoinr- ing a previously repressed sector or from foreign ing expectations that it would catch up rapidly direct investment, and state capacity to manage with West Germany. This occurred, first, because the process allow a country to follow such a the one-to-one conversion rate between West phased transition, it is less likely to experience a German and East German marks led to a sub- period of contraction in output. stantial overvaluation of the East German cur- These conditions were largely absent in most rency and, second, because attempts to bring East transition economies covered by this report. But German wages in line with West German wages, the costs of soft budget constraints on China's in the face of much lower labor productivity in xxi Overview the East, undermined East Germany's competi- cruing to three different constituencies at dif- tiveness. As a result, a much larger part of the ferent doses of reform in a typical transition inherited capital stock was rendered unproduc- economy. tive than would have been the case had more appropriate macroeconomic policies been cho- * State sector workers, employed in state en- sen. Thus, while institutional change is impor- terprises and lacking the skills to become tant, so too is policy reform, and it is essential new entrants in the competitive market, face that they proceed hand in hand. a sharp drop in income as discipline calls for downsizing the sector, with little hope The Political Economy of Discipline and of any substantial recovery with the inten- The Political Economy of Discipline and siiato ofrfom Encouragement sification of reform. * Potential new entrants, workers in state en- M /[ any transition economies outside Central terprises and new entrepreneurs with skills to lVl Europe and the Baltics are stuck in a no become new entrants in the competitive mar- man's land between plan and market. If the ad- ket, have a classic J-curve pattern of income. vantages of economic reforms are so obvious, They face significant adjustment costs at low why doesn't every country adopt them? Can eco- levels of reform as they exit the state sector. In nomic policy choices be systematically related addition, they realize gains only when enough to particular institutional characteristics of po- progress has been made with policy and insti- litical systems in transition? tutional reforms to promote and support new The political economy of reform within the entry into the competitive market. framework of discipline and encouragement can * Oligarchs and insiders begin the transition be expressed graphically by tracing the paths with substantial de facto control rights over of winners and losers from the transition. Fig- state assets and close ties with the political ure 1 depicts the gains and losses in income ac- elite inherited from the previous command FIGURE 1. Winners and Losers from Reform Income gains + a-.--- New entrants Oligarchs and insiders O R = Extent of reforms State sector workers Note: Ro = no reforms; R1 = point at which income gains of oligarchs and insiders are maximized; R2 = level of reforms that allows the winners of reforms beyond R1 (new entrants) to compensate for or exercise enough political pressure to neutralize the resistance of oligarchs, insiders, and state sector workers. Source: Authors. xxii Overview system. However, because of limited skills to lead to barriers to entry. Public support for to compete in the market economy, they face radical reforms therefore depends on perceptions an inverted U-curve of income gains. They of government credibility in its commitment to are the immediate beneficiaries of liberal- follow through with such reforms. ization and privatization, as de facto con- The risk of "getting stuck" at a low level of trol rights over state assets can be converted reform (R1) characterized by liberalization with- into de jure control and cash flow rights. out discipline and limited encouragement of new They reap concentrated gains in the early entry is high. As both insiders and state sector stages of reform from the opportunities for workers face declining incomes after R,, these arbitrage, rent seeking, and tunneling that groups have a strong incentive to join forces to arise if liberalization and privatization are oppose further economic reforms. It is only not combined with discipline and encour- when reforms reach a critical threshold (R2) that agement. But these gains dissipate as further the added gains to new entrants are enough to reforms lead to increasing competition and allow these winners to either compensate the market entry. losses of the other groups or to generate enough political pressure to neutralize opposition to Given these patterns of gains and losses, each continued reform. constituency prefers a different combination of By recognizing that different combinations reforms. State sector workers prefer the status of reforms produce different configurations of quo Ro and reject all reforms. Oligarchs and in- winners and losers, the framework of discipline siders prefer a partial reform and sustain the re- and encouragement suggests two political chal- form process through Rp, the point where their lenges in promoting economic reform: gains are maximized and beyond which further .. ... ~~~~* Securing the support of potential new en- implementation of policies of discipline and en- couragement threaten to undermine gains from trants for comprehensive reforms until wider rent seeking and tunneling. For potential new efficiency gains from disciplne and encour entrants, the reform process offers sacrifices at agement are realized. * Preventing the early winners from liberaliza- the beginning for the promise of gains when the tion and privatization from undermining fur- reforms are further advanced. thorms thatld from dermmmg fur Where the risk of oligarchs and insiders therreforms thatwould imposedscipline and blocking anything more than partial reform is encourage new entry and competition and high, potential new entrants and state workers will either reject reform or support only partial To meet these challenges, governments must reform, because the latter, by limiting the appear credible to potential new entranrs in its downsizing of the state sector and maintaining commitments to follow through with the long the flow of subsidies, imposes lower adjustment and difficult process of economic reform. Gov- costs. Yet it is precisely such partial reforms- ernments must also be able to constrain oli- liberalization without discipline and with selec- garchs and insiders from using their initial ad- tive encouragement-that make capture of the vantages in the reform process to derail further state by oligarchs and insiders a self-fulfilling reforms that would create a more competitive prophecy. This has led to a so-called partial re- market economy. form paradox in many transition economies in Credibility and constraint are rooted in po- which governments lack credibility and are highly litical institutions shaped by the cultural and his- susceptible to state capture. This leads potential torical legacies that guided the exit from com- new entrants at the outset of transition to dis- munism. In many countries in the CIS ancd count substantially the potential gains from any Southeastern Europe, where the state has been proposed radical reforms and instead support captured by narrow private interests, the collapse partial reforms that offer lower costs early in the of communism was rooted in a contest among reform process, even though they are more likely competing elites rather than in any broad social niii Overview movement. The new political arrangements in strong public administration allowed for greater these "concentrated political regimes" were de- security of property and contract rights and bet- signed by incumbent leaders, often as a way to ter public infrastructure, important preconditions consolidate their power. They lacked the cred- for promoting new entry. As reform progressed ibility to build and sustain broad popular sup- to promote entry and improve the enabling en- port for a comprehensive reform program. vironment, constituencies with a stake in advanc- As a result, these countries embarked on tran- ing reform grew stronger, and the emergence of sition without a broad social consensus on the powerful insiders and oligarchs diminished. This goals of reform and without a way of organiz- combination allowed these countries to imple- ing the public behind these goals. Instead, in- ment and sustain comprehensive reforms. cumbent politicians sought alliances with pow- Political developments and economic reforms erful incumbent enterprises. In addition, the are closely interrelated. Political systems affect politicians continued partial liberalization and the incentives of politicians to make certain eco- privatization in the context of soft budgets and nomic policy choices; reform choices shape the barriers to entry that created tremendous oppor- configuration of social groups and the distribu- tunities for rent seeking by old and new enter- tion of power, which affects the structure and prises, especially in economies rich in natural functioning of the political system. For example, resources. Countervailing pressures from com- economic reforms that facilitate new entry also peting groups were weak and the disaffection and strengthen the constituency of SMEs, which build apathy of the "losers" minimized the direct costs support for increasing political competition. to politicians of poor policy choices. As a result, Nevertheless, given the sharp break with com- countries with concentrated political regimes munism and the disintegration of the Soviet have tended to languish in an equilibrium trap Union, choices about the structure of political of partial economic reforms. Political and eco- systems in the transition economies were gener- nomic power has been used to preserve market ally made before decisions about the nature and distortions that benefit narrow vested interests pace of economic reform. Moreover, in all but a at considerable social cost. couple of countries-Croatia and the Slovak This partial reform equilibrium can be con- Republic-the nature of the political regime has trasted with the situation in the "competitive de- not changed much since the start of transition. mocracies" prevailing in Central Europe and the This suggests that while the pace and direction Baltics. In the aftermath of popular revolutions of economic reforms may have reinforced initial against communist rule, political institutions in choices about the structure of the political sys- most of these countries emerged from roundtable tem, they do not appear to have decisively shifted negotiations among broadly representative popu- the course of political transition. As a result, a lar fronts and a wide range of other organized stronger case can be made for identifying the interests. This, together with the close ties of these direction of causation from political choices to countries to Western and Northern Europe and economic choices, thus providing part of the the pull of potential European Union accession, explanation for why some countries have been contributed to a wider social consensus on the unable to move beyond partial reform. main directions of reform and broad public sup- port for comprehensive reform programs in the Shifting Policy Priorities to Account for early stages of transition. Experience and New Conditions New governments in competitive democra- cies tended to focus first on promoting new con- N 4 uch economic policymaking is endog- stituencies of "winners" by removing entry bar- Mi enous from the broader perspective of riers, quickly tackling severe macroeconomic political economy. Designing effective reform instability (with its high costs to the public), and strategies must therefore take into account the using social protection to support the "losers" political incentives and constraints that block from the dislocations of reform. A legacy of progress in transition. But although initial xxiv Overview conditions and political institutions influence re- sector and to provide a major share of employ- form paths, these factors cannot wholly predeter- ment (50 percent) and value added (55-65 per- mine outcomes in a process as complex and mul- cent) in the economy. By contrast, in countries tifaceted as transition. Experience from across the where restoring sustained growth has proved world demonstrates that talented political lead- more elusive, new enterprises account for a low ers can maneuver countries out of so-called re- share of employment (10-20 percent) and value form traps. Critical elections or external shocks added (10-20 percent). can break long-term stalemates on reform. New That is why encouraging an investment cli- leaders can mobilize alternative coalitions and mate attractive for new entrants and meeting spark collective action that tips the balance of the policy and institutional challenges of en- power between the potential winners and losers couragement should be the highest priorities from further economic reforms. Clever winners for policymakers in transition economies. Re- can devise win-win strategies that co-opt their member, though, that encouragement cannot opponents to build support for reform. go very far without discipline. Therefore, the In what ways should policy advice during emphasis on encouragement is more effective these extraordinary opportunities for reform re- the more it is accompanied by hard budget con- flect the experience of the past decade and today's straints, exit mechanisms, product market comi- conditions? Three broad areas can be identified. petition, and stronger institutions for monitor- ing managerial behavior. From Privatization and Restructuring to Promoting Entry From Depoliticizing Enterprises to Monitoring Policy needs to shift its emphasis from Managers privatization and restructuring of assets to creat- The lack of restructuring in privatized enterprises ing wealth through new enterprises. The early in the CIS, together with tunneling and theft, emphasis on rapid privatization entailed remov- renew interest in the question of what institt- ing ownership of enterprises from the state, cre- tions are needed to encourage managers to be- ating a constituency for private ownership to help come effective stewards of enterprise assets. Al- guarantee the irreversibility of reform, and stop- though developing these institutions of corporate ping abuses such as asset stripping by enterprise governance has been on the reform agenda since managers and other forms of "spontaneous the start of transition, the difficulty of doing so privatization." Although much remains to be in countries without recent market experience done, particularly in privatizing medium-sized and was probably underestimated. large enterprises, these past concerns weigh less International experience suggests that with- heavily on policymakers today. To this must be out effective legal protection, suppliers of finance added the suggestion from the empirical litera- do not enter into contracts with enterprises to ture that privatization has promoted restructur- ensure that they get a return on their invesit- ing in the CSB, but not in the CIS (see Djankov ment-the essence of the corporate governance and Murrell 2000 and chapter 7 in this volume). problem-even if such arrangements are in the New enterprises are important to promoting interest of both parties. Concentrated ownership, growth. In both the leading and lagging reform- by providing enhanced monitoring of managers ers, new enterprises enjoy a productivity advan- by shareholders, can overcome some of the cor- tage over old enterprises. So transferring re- porate governance problems that plague transi- sources from old enterprises to new ones is a tion economies lacking such legal protection. But source of growth. Although causality cannot be the type of concentrated ownership matters aLs inferred from the evidence, countries that have well. Enterprises controlled by strategic inves- returned to sustained growth have relied on a tors, particularly if they are foreign, have per- vibrant new sector to absorb labor and other formed much better than those controlled by resources released by the downsizing of the old holding companies or other financial institutions. xxv Overview The selection of strategic investors matters too. accruing to the early winners. In Russia, for Enterprises sold through transparent tenders or example, powerful insiders with a stake in weak auctions have generally attracted better owners, corporate governance have frequently hampered outperforming enterprises sold directly to politi- the work and enforcement efforts of the Secu- cally connected parties, frequently at highly sub- rities and Exchange Commission. In environ- sidized prices. Without such safeguards, concen- ments of high state capture, privatization has trated ownership does not avoid the risk of not created enough demand for the enforcement expropriation of assets and income belonging to of property rights. Indeed, quite the opposite is minority shareholders. true. While privatization is positively associated In countries where the preferred method of with public governance (the latter summariz- privatization-direct sales through transparent ing the state's capacity to provide key public tenders or auctions to strategic investors-was goods) in low-capture environments, the asso- unavailable, the relevant comparison in assess- ciation is negative in high-capture environments ing privatization is not between the actual (EBRD 1999). method chosen and the ideal method, but be- Several of these issues were not foreseen at tween the actual method and continued state the beginning of the transition. One was the ap- ownership until strategic investors were found. parent stability of such partial reform equilibri- Countries where mass privatization using ums. Another was the unexpectedly perverse re- vouchers originally dispersed ownership and lationship between privatization and the quality where secondary trading has not led to trans- of governance in such environments. That in- parent consolidation of shares witnessed expro- creased the challenge of enhancing creditors' and priation of assets and income of minority share- shareholders' rights; promoting internationally holders by those that were able to gain control recognized accounting and auditing standards; over the enterprise in the first stages of the and enforcing takeover, insolvency, and collat- privatization program. But the success of con- eral legislation in the face of opposition from a tinued state ownership is not assured unless narrow set of entrenched private interests. there is a political commitment to transparent The need to strengthen corporate governance, privatization outcomes and a minimum insti- despite opposition from oligarchs and insiders, tutional capacity to prevent asset stripping by is an important lesson from the first decade of enterprise managers in the interim period. In- transition. This is, however, a time-consuming deed, navigating between continued state own- process, during which policymakers still need to ership with eroding control rights on the one make choices about the appropriate stewardship hand and a transfer to ineffective new private of state assets, including privatization. The fol- owners with an inadequate institutional frame- lowing broad principles should guide a program work on the other hand was one of the most of privatization: difficult challenges confronting policymakers in charge of privatization. Irrespective of the al- * Privatization should be part of an overall ternative chosen, governments need to enshrine strategy of discipline and encouragement. investor protection in the legal system and * Small enterprises still owned by the state supplement it with a system of regulation for should be sold directly to new owners financial intermediaries, such as investment through an open and competitive auction and funds and brokers (Johnson and Shleifer 2001). without restrictions on who may bid for the Developing laws and institutions to protect shares. investors and monitor managerial behavior and * In general, medium-size and large enterprises thus facilitate the development of bank and non- should be privatized to strategic outside in- bank financial intermediation in countries with vestors, who, with a concentrated controlling no recent market experience is far more diffi- stake, will best use enterprise assets. Although cult when opposed by early winners from tran- several transaction methods may be used, in- sition. Further reforms would dissipate the rents cluding negotiated sales, the evidence suggests xvi Overview this can be brought about most effectively by powerful vested interests in the new private through competitive case-by-case methods, sector, granting extraordinary decree-making which are more deliberative than voucher powers to the executive branch to dissipate rents schemes or rapid, small auctions. They use and level the playing field has not won against independent financial advisors who both pre- strong opposition from insiders and oligarchs. pare the enterprise for sale and act as sales What is needed instead is to mobilize through agents on behalf of the state. Rapid privati- greater political inclusion and coordination all zation to insiders or through mass privatiza- constituencies that lose from partial reform and tion should be avoided. In countries such as that stand to gain from further progress toward Belarus, Turkmenistan, and Uzbekistan, each a more competitive market economy. For ex- of which has a substantially unfinished agenda ample, given the wide and generally regressive of privatization of medium-sized and large en- impact of high inflation, political parties in sev- terprises, but where the state retains the ca- eral transition economies mobilized enough elec- pacity to provide public goods, the state should toral support for macroeconomic stabilization use its administrative capacity to control the to overcome the opposition of powerful com- disposition of public assets as transparently mercial banks and other actors that gained from as possible, while developing institutions of economic volatility. Similarly, banking crises in corporate governance. Transparency would be Bulgaria, the Czech Republic, and Hungary enhanced, for example, if decisions regarding sparked electoral appeals to disgruntled savers public assets were to be reviewed by indepen- that helped break the stalemate over such issues dent boards of directors and accompanied by as banking privatization and regulatory reform. public disclosure. Business associations could serve as vehicles * Privatization should be accompanied by in- of collective action by SMEs, new enterprises, creasing competition in the market for the and "second-tier" enterprises that suffer from products sold by the enterprise in question and an unlevel playing field, discretionary taxation vigorously enforced by the competition policy and regulation, and anticompetitive barriers. In authority. This can help discipline managers countries with concentrated political regimes, when corporate governance is weak. such associations are weaker than those in the * Divestiture of enterprises in sectors charac- competitive democracies of Central Europe, terized by a natural monopoly or oligopoly which have more voice. So political parties have (becoming rarer with advances in technology) yet to seek strategic alliances with such actors as must proceed with caution, if at all. Estab- an alternative base of support and funding. lishing an efficient regulatory regime is a pre- Overcoming the coordination dilemmas of requisite to protect the public interest, lest mobilizing the highly dispersed winners of fur- divestiture transform an inefficient public ther reform is not easy. A major challenge for monopoly into a poorly regulated or unregu- the reformist team that comes to power during lated private monopoly. a period of extraordinary politics in countries * The state's property and cash flow rights with concentrated political regimes is to make should be clarified and strengthened in en- clear the links between rents from partial re- terprises in which the state continues to hold form and the direct costs to society. Tax arrears, a stake. tax and duty exemptions for high-profile con- glomerates, and nonpayments need to be linked Mobilizing the Winnemsfrom Further Reform in the public mind to delayed public sector wages and pensions and the poor provision of Breaking the political economy equilibrium un- social services. The complex web of derlying partial reforms is the most important and nontransparent subsidies to powerful businesses difficult challenge in advancing transition in many needs to be uncovered, revealing that such sub- countries of the region, particularly in the CIS. sidies tend to benefit incumbent managers rathesr Where the state is already susceptible to influence than workers. xvii Overview To advance reforms, governments should fo- Conclusions cus on smoothing the curves of the winners and nalysis of the first ten years of transition in losers at the initial stages of reform (see figure A Eastern Europe and the former Soviet 1). This means lowering the adjustment costs for Union highlights the following lessons, which potential new entrants and reducing the high could be applied in future to economies that have concentration of gains to oligarchs and insiders. made limied progres wt reform. One way to do this is by strengthening the pro- made lmite progress with reform. vision of basic public goods, such as secure prop- * While the initial conditions that prevailed at erty rights and a legal and judicial system. An- the beginning of transition were critical for other way is by reducing excessively high explaining the output decline that occurred marginal tax rates and broadening the tax base initially in all countries, market-oriented that promotes entry of enterprises from the un- policy reforms have played a significant role official to the official economy. This can break in promoting subsequent economic growth. the vicious cycle of informalization, lower tax Creating an environment that disciplines old revenue, and further intensification of tax rates enterprises into releasing assets and labor and on a shrinking base. Developing a rule-based tax encourages new enterprises to absorb those administration to enforce efficient taxation of resources and undertake new investment, the new private sector is also important. without tilting the playing field in favor of To align the incentives of local governments any particular type of enterprise, is central to identify with small business and increase en- to economic growth. try, taxes on small enterprises should be allocated * Policymakers cannot postpone the pain of liq- to local government. Simplifying entry and li- uidating and restructuring the old sector until censing arrangements for new enterprises is criti- the cushion provided by new enterprises is in cal. These measures, by encouraging the emer- place. The success of the encouragement strat- gence of new enterprises, offer a stable outside egy requires simultaneous application of dis- option to state workers, creating opportunities cipline, because a lack of discipline undermines for them to become potential entrants into the the level playing field between different kinds burgeoning sector of new enterprises and less- of enterprises. Furthermore, the practice of ening their opposition to reform. allowing old and large enterprises to avoid As entry occurs gradually at the margin, these paying taxes and social security contributions actors become an effective constituency demand- and to avoid repaying bank debts has been at ing reforms to remove weaknesses in the invest- the root of macroeconomic crises. ment climate over the long run. Furthermore, a * Developing legal and regulatory institutions gradual reallocation of public expenditures from to oversee enterprise management, though nontransparent and discretionary subsidies to time-consuming, is important. In the mean- worker training, severance payments, and grants while, where direct sales of state assets to stra- for improving services in communities affected tegic investors-a preferred method of by downsizing can create further momentum for privatization-is not feasible, policymakers reform. More broadly, strengthening the social face a difficult choice between (i) privatization safety net and divesting such social assets as hous- to ineffective owners in a context of weak ing, child care, and health facilities shifts the corporate governance, with the risk of ex- burden of social protection from enterprises to propriation of assets and income of minority governments, thus facilitating the restructuring shareholders by those who gained control that will foster a return to growth. Fiscal policy over the enterprise, and (ii) continued state therefore has the potential to smooth the curves ownership in the face of inadequate political described in figure 1 and redistribute a part of commitment to transparent privatization the reform dividend to those who would other- outcomes and limited institutional capacity wise bear its costs. It is thus a key element in to prevent asset stripping by incumbent en- supporting comprehensive reform. terprise managers. fliyi Overview * Breaking out of low-level equilibrium traps * Controlling fiscal risks arising from implicit in which the immediate beneficiaries of lib- and contingent liabilities on account of state- eralization and privatization have captured owned enterprises, banks and pension sys- the state and oppose measures of encourage- tems, guarantees for projects and balance ment such as competition and free entry that sheets of special purpose agencies, and thLe would reduce their rents requires mobilizing fiscal stance of subnational governments. small, medium-sized, and new enterprises and * Implementing bankruptcy laws to facilitate exit second-tier businesses that suffer as a result through a formal process and to create incen- of the uneven playing field and stand to gain tives for closure through informal mechanisms. from further reform. Fiscal policy has an im- * Reforming the budget process so the state can portant role to play here, by redirecting sup- meet its expenditure obligations in cash and port away from ailing enterprises and toward on time, with a view to eliminating arrears worker training and severance payments, and and noncash payments. by divesting social assets such as housing, child care, and health facilities from enter- Monitoring and influencing managerial be- prises to governments. havior requires: * In reforming economies, where the bulk of Annex 1. Discipline and Encouragement: assets are in private hands, improving insti- The Refom Agenda tutions of corporate governance by strength- XV! hat policy and institutional reforms are ening legal protection for minority sharehold- needed to create an environment favor- ers and creditors, bringing in management abl needed to create an enco ronment favo by concentrated owners or strategic investors, able to discipline and encouragement? While no promoting internationally recognized ac- individual policy can be assigned to a single out- countmg itingtany da gnized ac- come in an interrelated system, it helps to think counting and auditing standards, and work- of policy packages as those primarily directed at in, as capital markets grow, to develop a disciplining the old sector and those primarily * market n corporate control. ., * ~In nonreforming economies, where assets re- directed at encouraging the new sector without a . . 1 ,. 1 1 . , , . . ~main largely in the public sector and where tilting the playing field in favor of any particular the state retains the capacity to provide pub- type of enterprise. lic goods, the state should use its administra- tive capacity to control asset disposition as Discipline transparently as possible, while developing institutions of corporate governance. In an environment of price and trade liberaliza- tion, discipline requires imposing hard budget Promoting competition in product markets constraints on enterprises, providing exit mecha- is an important ingredient of discipline, especially nisms for insolvent enterprises, monitoring and because of competition's effect on the behavior influencing managerial behavior to reward effi- of enterprise managers. In the CIS, for example, cient stewardship of assets and to discourage competition can compensate to some extent for tunneling and theft, increasing product market weak monitoring of managers by shareholders competition, transferring social assets from en- and creditors. This requires: terprises to local governments, and using the social safety net as a cushion for displaced work- including trade liberalization ers and other losers from reform. Imposind otheralosersdfrom budgetfconstras r Enforcing competition laws through a gov- Imposing hard budget constraints requires: ernment agency vested with the requisite * Eliminating tax exemptions, fiscal and finan- authority. cial subsidies, budget and tax offsets, and Transferring responsibility for social assets such directed credits. as housing, utilities, clinics, and kindergartens fromn xxix Overview enterprises to local governments will allow enter- * Improving targeting of limited cash benefits prise restructuring to proceed in countries that have in the low-income CIS countries through relied heavily on enterprises as instruments of so- geographical targeting, community-based cial policy. This requires: identification, or self-targeting through some * Clarifying the roles and responsibilities of form of public works scheme. subnational governments * Giving those governments resources to ful- Encouragement fill their assigned responsibilities In addition to liberalizing prices and trade, im- * Reforming communal services (including cen- proving the investment climate for domestic and tral heating and gas) and moving tariffs to foreign investors is key to encouraging new en- cost-recovery levels terprises. This requires establishing secure prop- * Replacing across-the-board housing and util- ity subsidies with targeted social assistance erty and contract rights and providing basic in- to the poorest households, sometimes com- frastructure, reducing excessive marginal tax bined with lifeline tariffs. rates, simplifying regulatory and licensing pro- bined withlifelietariffscedures, and developing a competitive and effi- Social insurance programs that cover pen- cient banking system. sions adn py niEnsuring the adoption of laws best suited to sions and unemployment insurance and social seuigportrghsadcnatseqrs assistance programs make it more feasible to dis- securmig property rghts and contracts requires cipline old enterprises into shedding labor and emphasis on two areas: to help create a constituency for discipline. Re- * The process for drafting laws: enterprises form of social insurance programs requires: should have input in their design and be in- Movingpmin Central Europe formed beforehand of changes in rules that *Moving pension reform wil affectl therrpeaton and the Baltics to multipillar systems, with a will affect their operations minimum poverty-based benef. Bec e aThe effectiveness of the judiciary: its fairness minimum poverty-bDased benefit. Because structural unemployment is falling in these (bias, honesty, and consistency) and the like- structural unemploymentioodo ISnfallmemethes countries, they can implement unemployment ihood of enforcement. insurance programs. Tax reform should: * Reforming pay-as-you-go systems in Bul- Raise the turnover threshold for becoming a garia, Romania, Russia, and Ukraine to put value-added taxpayer high enough to exclude them on a firmer fiscal footing, with a mini- small enterprises, which should instead be mum poverty-based benefit. Unemployment subject to a small-enterprise tax regime. This insurance might involve a flat benefit or sev- regime should be simple enough to lighten erance payment. erance payment. ~~~~~the administrative and reporting burden on * Moving to a flat benefit structure in the thea s and re porting burden resource-constrained, low-income countries taxpayers and reduce iteractions between of the CIS, to protect the poorest elderly. thecte salndsh taxeauthority. ' ' * ~~~~~~~Allocate small business taxes and property The reform of social assistance programs re- taxes to subnational governments to help quires: them identify with new emerging enterprises, * Moving to a means-tested cash benefit assis- which are typically a source of job creation, tance program in Central Europe and the Baltics rather than with large, bankrupt enterprises * Moving to categorical cash benefits, either to save jobs that should be cut. universal or targeted by category, in Bulgaria, Streamlining the business licensing and reg- Romania, Russia, and Ukraine and using istration requirements that govern entry of new means-tested benefits only where local insti- enterprises is a high priority. Addressing these tutions are strong issues requires: oom Overview * Simplifying and making transparent entry can help upgrade managerial and supervi- and licensing procedures sory standards. The alternative-privatizing * Reducing the scope for arbitrary decision- banks to concentrated owners-should oc- making and abuse of power. cur only if there is a clear separation between shareholders and borrowers. The pace of Insecure property rights are more of a con- atize sod notroutrun The devef privaiaion sol o urntedvl straint on investment for new enterprises than . . the availability of bank finance, particularly in opment of adequate supervision authority. the CIS, which is less far along in the transition (Johnson, McMillan, and Woodruff 2000). But Notes as the transition proceeds and legal and judi- cial reforms strengthen property rights, finan- 1. The CSB comprises Albania, Bosnia and cial deepening and development will be essen- Herzegovina, Bulgaria, Croatia, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, the former Yugoslav Republic of Macedonia, Poland, Romania, led by new enterprises. the Slovak Republic, and Slovenia. The Federal Re- Developing a competitive banking sector re- public of Yugoslavia is not included in the aggregates quires a strategy to deal with the exit of failed for the CSB countries because no data are available banks, the entry of new banks, and bank before 1998. The CIS comprises Armenia, Azerbaijan, Belarus, Georgia, Kazakhstan, the Kyrgyz Republic, privatization and restructuring. Strengthening the Moldova, the Russian Federation, Tajikistan, regulatory authority for bank supervision is ex- Turkmenistan, Ukraine, and Uzbekistan. tremely important for performing these activi- 2. The terms "Russia" and "Russian Federation" are ties effectively. This will require: used interchangeably and refer to the same country. * The resolution of failed banks, which domi- 3. Conceptual and measurement problems plague thLe nate the bnigsseinGDP data (see box 1.1). However, these problems do nate the banking system in transition econo- not modify the qualitative thrust of the statements mies and have large interbank exposures and presented here. loans to loss-making enterprises, therefore 4. These estimates are based on 1993 purchasing poses special problems. Liquidation and re- power parity rates (World Bank 2000a). structuring options should be assessed care- 5. The agenda of accession to the European Union fully in the event of systemic risks to the fi- looms large for countries in Central and Eastern Eu- nancial sector and the need to impose hard rope. The World Bank has been engaged in a major budget constraints on banks and enterprises. project that examines country-specific and subregional * The fostering of a competitive and efficient issues arising from European Union accession. The principal outputs of the project are listed in the bib- banking system requires encouraging the entry liographic guide; therefore, that subject will not be of new banks that satisfy prudential criteria for covered in this report. minimum capital requirements and capital ad- 6. Kornai (1986) introduced the term soft budget equacy. Because supervisory responsibility re- constraint (the opposite of a hard budget constraint) sides mainly with the home country regulatory to characterize the environment faced by state enter- authority, entry by foreign banks and acquisi- prises in Hungary in the 1980s. tion of stakes in existing banks by foreign banks 7. The analysis is based on the assumption that data is a quick way of.importing. managerial and for small enterprises can be used to approximate new iS a quick way of importing managerialand enterprises. The approximation is not accurate inas- govemance expertise and improving bank regu- much as the set of small enterprises includes small old lation in a transition economy. In any event, enterprises. Annex 4.1 presents an upper bound for the expansion of the banking system should the error committed by this assumption in the CSB occur only in line with the growing capacity countries. While the estimate for the upper bound is for bank regulation and growth in the number substantial in 1995, it shrinks significantly under rea- sonable assumptions about the mortality rate of en- of creditworthy borrowers. terprises by 1998, the year for most of the data in this * The privatization of banks to strategic in- report. The terms "new enterprises" and "small en- vestors whenever possible. If foreign, they terprises" are henceforth used interchangeably. xxxi Overview 8. Annex 4.2 describes the circumstances under which comparisons based on labor productivity (for which data are available) correspond to comparisons based on total factor productivity, which is the rel- evant concept for this report. mud Part 1 The First Decade in Transition How Did Transitionr Economies Performl? T he countries of the Europe and Central Asia region, in the first 10 years of transition, dis- played some common trends and some significant variations. These variations were most evident between the Central and Southeastern Europe and the Baltics (CSB) and those coun- tries in the Commonwealth of Independent States (CIS)-and within these two groups. Output Fell Sharply The trend in real gross domestic product (GDP) for the CSB matches, at least qualitatively, what was expected at the onset of the transition (figure 1.1). There was a sharp initial fall, followed by a fast recovery and then sustained growth at levels determined by factor accumulation and increases in productivity (figure 1.2). Even for these countries, though, the initial fall was larger than anticipated.' The output decline was far deeper and longer in the newly independent countries of the CIS, particularly with the incipient recovery in 1997 derailed by the fiscal-financial crisis in the Russian Federation the next year. Only now is there evidence of growth being restored in this group of countries. The magnitude and duration of the transition recession was, for all countries, comparable to that for developed countries during the Great Depression, and for most of them it was much worse (table 1.1). The CIS had an average of 6.5 years of declining output, resulting in the loss of half the initial level of measured output. Even at the end of the decade, the CIS had recovered only 63 percent of its starting GDP values (but see also Aslund 2001). Poland had the shortest and mildest recession: a 6 percent drop in production over two years. The three Baltic countries had the longest (5-6 years) and deepest (35-51 percent) recessions among the CSB. In this, they are much closer to the average of the CIS than to other CSB countries. In the CIS Armenia, Georgia, and Moldova saw the steepest declines-Georgia, an astonishing 80 per- cent fall in output, largely a result of the long internal turmoil-while Belarus and Uzbekistan had mild declines. 3 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Former Soviet Union FIGURE 1.1. Changes in Real Output, 1990-2001 Index (1990 = 100) 120 110 _ 100 Central and Southeastern 90 Europe and the Baltics 80,- Commonwealth of 80 > Independent States 70 - Europe and Central Asia 60 50 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Note: Europe and Central Asia is the average of the CSB and the CIS. All aggregates are population-weighted. Values for 2001 are projected. Source. World Bank country office data. FIGURE 1.2. Output Growth Rates, 1990-2001 Annual rate (percent) 8 4- 8------------------- ...................... .......................... /- Central and Southeastern Europe and the Baltics -4 /- Commonwealth of Independent States -8 -- l - Europe and Central Asia -12 -16 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Note: Europe and Central Asia is the average of the CSB and the CIS. All aggregates are population-weighted. Values for 2001 are projected. Source: World Bank country office data. The transition recession is now over-Ukraine, output decline after their initial recovery. Four the only country with 10 consecutive years of others had one-year recessions: Albania in 1997, output decline, registered growth in 2000. How- as the collapse of a pyramid scheme led to a fi- ever, the recovery has not been smooth in all coun- nancial crisis and civil disturbances, and Croatia, tries. Three CSB countries (Bulgaria, the Czech Estonia, and Lithuania in 1999, largely because Republic, and Romania) had at least two years of of the Russia crisis in August 1998. -4 How Did Transition Economies Perfonn? TABLE 1.1. The Transition Recession Consecutive years Cumulative output Real GDP, 2000 Countries of output decline decline (percent) (1990 = 100) CSBs 3.8 22.6 106.5 Albania 3 33 110 Bulgaria 4 16 81 Croatia 4 36 87 Czech Republic 3 12 99 Estonia 5 35 85 Hungary 4 15 109 Latvia 6 51 61 Lithuania 5 44 67 Poland 2 6 112 Romania 3 21 144 Slovak Republic 4 23 82 Slovenia 3 14 105 CISa 6.5 50.5 62.7 Armenia 4 63 67 Azerbaijan 6 60 55 Belarus 6 35 88 Georgia 5 78 29 Kazakhstan 6 41 90 Kyrgyz Republic 6 50 66 Moldova 7 63 35 Russian Federation 7 40 64 Tajikistan 7 50 48 Turkmenistan 8 48 76 Ukraine 10 59 43 Uzbekistan 6 18 95 Output decline during the Great Depression, 1930-34 France 3 11 n.a. Germany 3 16 n.a. United Kingdom 2 6 n.a. United States 4 27 n.a. n.a. Not applicable. a. Simple average, except for the index of 1990 GDP, which shows population-weighted averages. Source: World Bank country office data; Maddison (1982). Industry Shrank-Services Grew economies with comparable per capita incomes. During the transition, the industrial sector B oth the CSB and CIS started the transition shrank, falling to about a third of the economy, B with a much larger industrial sector and a and the share of services grew to about half (table much smaller service sector than market 1.2). Perhaps less expected, the increase in 5 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Forner Soviet Union services in the CIS took place at the expense of 70 percent, for such advanced reformers as the industry and agriculture, which both declined by Czech Republic, Estonia, and Hungary. In view about 9 percent of GDP. These sectoral shifts and of the rapid growth of GDP in the CSB in recent the decline in total output mean that output in years and the slow privatization of large state agriculture and manufacturing is now 40-45 per- enterprises, the increase in share testifies to a dra- cent of its pretransition level. matic increase in new private sector activity. (The role of new enterprises in promoting economic Private Enterprises Overtook the State Sector growth is discussed in part 2.) rhe most basic transformation was moving Exports Rose-Moving Toward Industrial T resources from the state to the private sec- C tor. In 1999 the private sector in most countries in the region produced more than half of GDP ountries more advanced in their recovery (table 1.3). The share was much larger, more than Chave been more successful in increasing their exports and reorienting them to the indus- trial countries. However, even in countries whose TABLE 1.2. output did not grow between 1993 and 1998, such Composition of Output, 1990-91 and 1997-98 as Moldova, Russia, and Ukraine, real exports increased (table 1.4). Percentage of GDP Important in the recovery of output was di- rect investment from abroad. These flows are Regions and periods Agriculture Industry Servzices important not only as a source of capital and new CSB technology to modernize industries and extract 1990-91 13.7 45.1 41.2 natural resources, but also as a signal of confi- 1997-98 13.9 33.0 53.1 dence in the transition to a market economy. Dur- ing 1996-99 more than US$70 billion in direct CIS investment came to the region, most of it to the 1990-91 27.5 39.7 32.8 CSB (table 1.5). In the CIS foreign direct invest- 1997-98 18.7 31.2 50.1 ment was largely confined to the energy-rich coun- Source: World Bank country office data. tries, with Azerbaijan, Kazakhstan, and Russia receiving 75 percent of the total. Russia's share TABLE 1.3. was even lower than several of the CIS countries' shares, despite the considerably greater size of its Private Sector Growth, 1990s economy and resource endowment. Percentage of GDP ______ Poverty Increased Sharply Countries 1990 1994 1999 The foregoing discussion suggests that the CSB 1 1 50 68 T initial fall in output may overestimate the Czech Republic 12 65 80 decline in living standards during the transition Estonia 10 55 75 (box 1.1). However, the period was still one of Hungary 18 55 80 extreme hardship for many people (World Bank Romania 17 40 60 2000b). Although extreme poverty is still lower CIS 10 20 50 in the transition economies than in other devel- Armenia 12 40 60 oping countries, it increased sharply during the Belarus 5 15 20 decade (table 1.6).2 In 1998 one of every 20 Russian Federation 5 50 70 people in the transition economies had per capita incomes below US$1 a day, up from fewer than Source: EBRD (2000). one in 60 a decade before. Moreovcr, the increase 6 How Did Transition Economies Perfonn? TABLE 1.4. Export Growth and Destination, 1990s (percent) Real export growth Share of exports to industrial countries Countries 1993-98 1992-93 1998-99 CSB 8.8 35.8 67.5 Albania 22.0 62.9 94.1 Bulgaria 4.3 55.1 59.0 Czech Republic 10.4 29.9 69.3 Estonia 10.8 25.9 71.3 Hungary 11.1 67.4 81.5 Macedonia, FYR 7.3 22.2 65.5 Poland 12.9 71.6 75.5 Romania 8.7 44.3 71.0 Slovak Republic 6.9 15.9 59.2 Slovenia 5.7 33.7 70.7 Cis 3.2 28.0 29.0 Armenia -8.6 9.4 34.9 Azerbaijana 14.0 4.2 20.0 Belarus -3.2 15.3 11.0 Georgiaa 10.3 2.3 25.9 Kazakhstan 3.4 43.8 29.6 Kyrgyz Republic -2.4 24.7 44.0 Moldova 4.8 6.2 31.3 Russian Federation 4.7 59.3 49.4 Ukraine 5.8 18.1 23.3 a. 1995-98. Source: World Bank and International Monetary Fund databases. TABLE 1.5. Main Recipients of Foreign Direct Investment, 1992-99 1992-95 1996-99 Countries USS millions Percentage of GDP US$ millions Percentage of GDP CSB 21,091 0.5 50,558 3.3 Czech Republic 4,821 2.9 10,104 4.6 Estonia 647 3.9 1,050 5.2 Hungary 9,399 5.7 6,979 3.8 Poland 2,540 0.6 17,096 2.9 CIS 8,272 1.0 22,001 2.5 Azerbaijan 237 4.2 3,222 20.9 Kazakhstan 2,357 2.7 4,971 6.4 Russian Federation 3,965 0.3 8,412 0.7 Turkmenistan 427 3.5 334 3.0 Note: Shares of GDP are period averages of medians for the group. Source: World Bank staff estimates and country statistical office data. 7 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Former Soviet Union Box 1.1. UmHs of GDP Statistics for Transtion Economies Estimating GDP and using it as an indicator of living standards have some well-recognized problems. These range from data collection and aggregation issues to the omission of nonmarketed goods (pollution and family services) and depletions of exhaustible resources. Official GDP also has special limits as a welfare indicator for transition economies, particularly when comparing output performance to the pretransition period. These limits fall into three groups: index number problems, omission of informal activities, and effects of the changes on the composition of outpuL Real GDP is an aggregate index constructed by weighting the outputs of individual products according to their respective prices. To be precise, the aggregation is done using value added (that is, the output price minus the cost of intermediate inputs) rather than prices. When relative prices change greatly-from transition economies opening to external trade, liberalizing domestic prices, and suffering the initial hyper- inflation-the estimated weights based on these prices can differ greatly between periods, making the calculated change in 'real' output very sensitive to the base period for the aggregation. Moreover, the standard relations among different indexes do not hold when the initial prices are not good indicators of either the opportunity cost of production (say, because of artificially low energy prices) or the value to consumers (say, because of generalized shortages, rationing, and queuing). The net effect of these factors is not clear, but they do reduce confidence that the initial changes in output are accurate measures of changes in welfare. Eliminating queuing, for example, with the same amount of real output clearly improves welfare. However, the increase in the relative price of consumer goods (say, for housing and utilities) means that the same overall 'real output' could be associated with lower real aggregate consumption and welfare. The collapse of central planning meant that its statistical system became inadequate to measure real economic activity, particularly that coming from the prvate sector. This, compounded by the need to go from net materal product to GDP, which includes services, meant that official output statistics did not capture the rapid growth of the informal sector. Subsequently, tax evasion and pressure from regulations and public sector bureaucracy provided incentives to operate business in the informal economy. Estimates of the size of the informal sector, using varous methods, suggest that its share of GDP varies enormously across the region, from 6 to 60 percent of GDP. In addition, durng the transition there were sharp changes in the composition of output, reducing the proportion of goods from which consumers derved little (current or future) value, such as military output, low-productivity capital goods, and poor quality consumer goods. These qualitative factors place additional limits to the usefulness of aggregate output as a measure of changes in the population's standard of living. TABLE 1.6. in poverty was much larger and more persistent Average Poverty Rates, 1990 and 1998 than many expected at the start of the process. (percent) Even in the most successful countries-such as Poland, where poverty came down steadily from Population living on less its peak in 1994-poverty rates were still higher than US$1 a day in 1998 than in 1991 (World Bank 2000b). Poverty increased not just because of the fall Regions 1990 1998 in output, but because of greater inequality in Eastern Europe and Central Asia 1.5 5.1 the distribution of income. Inequality increased East Asia and Pacific 28.2 15.3 in all transition economies, with great variation Latin America and the Caribbean 16.8 15.6 across the region (table 1.7). In some cases, the Middle East and North Africa 2.4 1.9 increase was modest-as in Hungary, where the South Asia 43.8 40.0 Gini coefficient for per capita income rose from Sub-Saharan Africa 47.0 46.4 0.21 in 1987 to only 0.25 a decade later. Even in Total 20.0 17.1 the Czech Republic and Slovenia, where inequal- ity rose more, the distribution of income remains Source: World Bank data. fairly egalitarian. Yet in the CIS and elsewhere, 8 How Did Transition Economies Perfonn? the increases in inequality have been unprec- TABLE 1.7. edented. In Armenia, Russia, Tajikistan, and Changes in Inequality during the Transition, Various Ukraine, the level of inequality as measured by Years Gini coefficients has nearly doubled. Gini coefficient of income per capita Notes Countries 1987-90 1993-9'4 1996-98 1. For example, one of the most authoritative early reports on the state of the socialist economies, A Study CSB 0.23 0.29 0.33 of the Soviet Economy (IMF and others 1991), antici- Bulgaria 0.23 0.38 0.41 pated an early recession that would end by mid- Croatia 0.36 - 0.35 decade at the latest (or earlier, under a "radical reforms" Czech Republic 0.19 0.23 0.25 scenario). The report projected an annual average Estonia 0.24 0.35 0.37 growth rate for the 1990s of 1.1 percent (3.3 percent under the radical reform scenario), which only two Hungary 0.21 0.23 0.25 countries, Poland and Slovenia, were able to achieve. Latvia 0.24 0.31 0.32 Lithuania 0.23 0.37 0.34 2. International comparisons of poverty rates are Poland 0.28 0.28 0.33 fraught with problems of estimation and interpreta- Romania 0.23 0.29 0.30 tion. In transition economies in Europe and Central Asia, harsh winters, deteriorating housing stocks, utili- S ties designed for pretransition conditions, subsidized CiS, 0.28 0.36 0.46 energy prices, and similar conditions mean that fami- Armenia 0.27 - 0.61 lies with US$2 a day are likely to have a lower stan- dard of living than families with the same income in Belarus 0.23 0.28 0.26 other regions. Furthermore, incomes do not always Georgia 0.29 - 0.43 match people's assessment of their economic welfare, Kazakhstan 0.30 0.33 0.35 which tends to depend also on their deprivation rela- Kyrgyz Republic 0.31 0.55 0.47 tive to other people and their previous condition. Moldova 0.27 - 0.42 Russian Federation 0.26 0.48 0.47 Tajikistan 0.28 - 0.47 Turkmenistan 0.28 0.36 0.45 Ukraine 0.24 - 0.47 - Not available. a. Median of countries with data. Source: World Bank (2000b). 9 Explaining Variation in Output Performance T he search for explanations of economic outcomes-causes, differences in magnitude, varia- tions in speed and sustainability-has spawned a large literature. The explanations focus on the characteristics of countries at the beginning of transition, the shocks emanating from the breakdown of the central planning system, the dissolution of the Soviet Union, wars and civil strife, and the policies to facilitate the transition. The political economy of post-socialist transition has also been examined to explain why economies may be trapped in situations of partial reform-in a no man's land between plan and market, where the early gainers from reform vigorously oppose further progress toward a market economy. Did Initial Conditions Affect Performance? S everal characteristics of the countries at the start of transition may have affected economic perfor- mance over the past decade: geography (such as endowment of natural resources and the proxim- ity to Western markets), years spent under central planning, and the nature of economic development under socialism (such as the extent of overindustrialization, military output, and repressed inflation). In testing for the influence of initial conditions on the economic performance of transition econo- mies, this report used the indicators developed by de Melo, Denizer, and Gelb (1996) aggregated into three categories: structure, distortions, and institutions.1 Structure encompasses such variables as the share of industry, the degree of urbanization, the share of trade with the socialist block, the richness of the natural resource endowment, and the initial income. These can be described as follows: * Sbare of industry in GDP. This share was high across the region because trade, financial services, and business and consumer services were repressed in the centrally planned economies. * Degree of urbanization. This indicator is related to the level of development as higher-income countries are generally more urbanized. The proportion of people in urban areas in 1990 ranged Transition-The hrst Ten Years: Analysis and Lessons for Eastern Europe and the Former Soviet Union from around 70 percent (Estonia, Latvia, exchange as well as a subsidy to imports and Russia) to around 30 percent (Albania, tax on exports. For the Soviet Union, the black Kyrgyz Republic, Tajikistan). market exchange rate reached as high as 1,800 * Trade dependence. High trade dependence on percent in 1990. other communist countries (measured by the * Terms of trade loss for the CIS. Trade flows ratio of Council of Mutual Economic Assis- within the Council of Mutual Economic As- tance exports and imports to GDP) reflected sistance took place at administrative prices the level of industrialization under central that were not directly linked for the most part planning, which favored large plants and re- to world prices. This meant large changes in gional interdependence.2 Inter-republic flows the terms of trade after trade was liberalized. were especially large for the smaller repub- The indicator measures the terms-of-trade lics of the Soviet Union, which had little trade loss as a share of GDP, as calculated by David outside the area. Tarr (1994). Small energy importers, such as * Natural resource endowment. Several coun- Moldova and the Baltic states, suffered the tries in the region-Azerbaijan, Kazakhstan, largest proportional losses (more than 10 Russia, and Turkmenistan-have rich but percent of GDP). Countries that were net underdeveloped deposits of oil and gas. This energy exporters generally had gains. gave them the potential for rapid growth, but * Reform history. Some countries (Hungary, required large investments to make produc- Poland, Federal Republic of Yugoslavia, and tion and transportation possible. Some energy to a lesser extent, Bulgaria) introduced some exporters have tended to delay reforms-with elements of market-based reforms before the deleterious effects on growth. collapse of the Soviet Union. The indicator * Income. Incomes (in 1989 dollars adjusted capturing reform history is the World Bank's for purchasing power parity) were generally Index of Liberalization in 1989 (de Melo, higher in Central Europe and the European Denizer, and Gelb 1996). part of the Soviet Union, ranging from * Pretransition growth rate. Looking at the US$1,400 per capita in Albania to US$9,200 average growth in 1985-89, the more ma- in Slovenia. ture countries stagnated in the late 1980s, while the poorest countries on average had Distortions in the economy refer to such fac- highe growth. tors as repressed inflation, black market ex- higher growth. change rates, trade shocks arising from the dis- Institutions encompass such variables as solution of the Soviet Union, the extent of prior years under central planning, location in rela- economic reform within the centrally planned tion to Western markets, and experience with system, and the degree of economic stagnation nationhood. prior to the transition. prior to the transition. Market memory. Some countries could draw * Repressed inflation. Most transition econo- on their market experience before the Soviet mies had repressed inflation, measured here period in the design of an institutional-legal as the difference between the increase in real framework supporting markets at the start wages and real GDP from 1987 to 1990. This of the transition. index was high for the Soviet Union, pro- * Location. Countries in Central Europe, par- pelled by the federal government's gradual ticularly those bordering on the West, had weakening of control over wages and regional more extensive trade links with market budgets associated with the partial liberal- economies and enterprises and institutions ization of the Gorbachev reforms. were more exposed to competitive pressures. * Black market exchange rates. This variable is Individuals had more freedom of travel, al- defined as the difference between the black lowing more exposure to Western markets. market exchange rate and the official exchange * New states. Countries with little experience rate, indicating the rationing of foreign as independent nation states may have had 12 Explaining Variation in Output Performance more difficulty creating efficient political in- The various financial crises of the 1990s-- stitutions and achieving political consensus. Mexico, East Asia, and particularly Russia-also contributed to delaying or interrupting the recov- ery of output (box 2.1). War and civil strife-iin Armenia, Azerbaijan, and Tajikistan in 1992-94, T he onset of transition was accompanied by in Georgia and Moldova in 1992, and in Croatia T severe shocks. The collapse of the institu- and FYR Macedonia in 1991-94-took a major tional and technological links of the Soviet cen- toll on lives, infrastructure, and the state, under- trally planned system disrupted the supply of in- mining the political consensus on reforms needed puts for production and the delivery of outputs, for successful transition. (The political economy posing new challenges for enterprises. The loss of of reform in war-torn countries is examined in budget transfers from the center and the elimina- part 3 of this volume.) tion of subsidized energy imports were severe blows, particularly to some of the newly indepen- Policies-Do They Matter? dent states of the CIS. The broader external eco- nomic environment was also less favorable in the he shift from planned to market economies 1990s, and the transition thus coincided with T is a social and economic transformation of lower growth rates in other developing countries.3 unprecedented scale. History offers no time-tested Box 2.1. The Regional Impact of the Global Financial Crisis and Recovery The global financial crisis, and particularly its spread to Russia in mid-1998, had a big effect on the other transition economies in the region. For many countries the first effect was disrupted trade with Russia-as demand contracted and trade finance and payments system arrangements were interrupted. CIS countries, such as Armenia, Azerbaijan, Kazakhstan, Tajikistan, and Ukraine, saw their exports to Russia decline up to 70 percent in the nine months after the crisis. The global crisis also deepened the recession in Western Europe, hurting the export performance of countries in Central and Eastern Europe. Mounting unemployment in Russia and administrative controls to stem capital flight severely curtailed workers' remittances from Russia, whose real value also declined as the ruble depreciated. This was particularly damaging for Russia's smaller neighbors. Between 1997 and 1999 Russian transfers abroad fell from US$771 million to US$493 million. Russian foreign direct investment abroad shrank from US$2.6 billion in 1997 to US$1 billion in 1998. The crisis made foreign financing scarce and expensive. Net inflows of private debt finance to the region fell by 50 percent between 1998 and 1999 to US$5.8 billion. This was only partly offset by the increase in net official lending, from US$1.3 billion to US$2.2 billion. For many countries privatization prospects were dampened. Despite initial fears, however, net foreign direct investment and portfolio equity flows actually increased in 1998 and 1999 over 1997. The crisis exposed vulnerabilities in the financial sector of most countries, even those implementing strong adjustment programs before the crisis. As the crisis hit exporters and industrialists across the region, local banks and other lending institutions saw their portfolios deteriorate. In Belarus and Latvia, with a high share of Russian assets in the balance sheets of banks, there was a direct impact on bank portfolios because the market value of those assets collapsed. Across the region, many banks (for example in Ukraine) held large amounts of nonindexed govern- ment debt and suffered capital losses as market interest rates rose with the flight of foreign capital. Some banks also had foreign exchange exposures that became more difficult to handle as local currencies depreciated. The severity of the crisis depended on the extent of economic interdependence with Russia and the policy response to the crisis. In general, the contractions and disruptions quickly translated into less economic activity and more unemployment. In many countries unemployment was already high before the crisis, and social safety nets were incapable of dealing with more unemployment and falling real incomes. As a result, poverty generally worsened. In Moldova, for example, the poverty rate increased from 35 percent of the population in mid-1997 to 46 percent in end-1998, and to 56 percent in mid-1999. Since mid-1999 the massive real devaluaton of the ruble and higher oil prices have fueled a rapid recovery of exports and economic activity in Russia, helped by world demand growth. Oil prices are expected to remain firm in the nearterm, as are prices for metals and raw materials. However, exporters of food and beverages are likely to face continued deterioration in their terms of trade. 13 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Former Soviet Union blueprints. Although the specific design of poli- * Legal and judicial reform cies, their sequencing, and their speed of imple- * Reform of public sector institutions. mentation are still subject to debate, there was a broad consensus that reforms should include: The question of whether these reform poli- * Macroeconomic stabilization cies matter can be tested against the alternative Price and trade liberallization hypothesis that output levels and annual changes in those levels were determined primarily by ini- banmpositiondof hatrdri budgetconstraintson tial conditions and external economic shocks, as banks and enterprises * Enabling environment for private sector de- described earlier. velopment the The extent of economic policy reform has been Reformenf the tax system and restructuring measured in a liberalization index developed by Reform of taxlicsystnditurem adrsucrig the World Bank to quantify progress in the tran- sition to a market economy.4 This index measures reforms needed to make markets the main mecha- nism for allocating resources, such as eliminating FIGURE 2.1. central planning and mandated allocations Progress in Policy Reform, 1990s through government orders and creating condi- tions to allow private production. The index also Hungary I * covers reforms to ensure the efficient functioning Poland of markets, such as stabilizing the macroeconomic Czech Republic environment, liberalizing the trade regime, and Estonia 1I pursuing procompetition policies. The index Slovak Republic _s ranges from zero to one, with zero representing Slovenia an unreformed, centrally planned economy, and Latvia I * one denoting the standards of a market economy Lithuania I * (figure 2.1). In 1998 the index was the highest for Croatia countries in Central Europe and the Baltics and Kyrgyz Republic 5 S the lowest for countries such as Belarus, Kazakhstan S I Turkmenistan, and Uzbekistan, which had yet to Bulgara I embark on a course of reform. Macedonia, FYR A substantial literature uses cross-country sta- Georgia * I tistical analysis of the transition from 1990 to 1999 Moldova - I ¢ to demonstrate that better policies are associated- Romania +.I¢. Armenia - I and significantly so-with higher annual growth Albania - I of GDP in Central and Eastern Europe and the Russia * I CIS, even when controlling for the effects of ini- Ukraine - tial conditions and external economic shocks (see Azerbaijan S + I * annex 2.1 for a summary of the main findings of Uzbekistan * this literature). The analysis allows for consider- Tajikistan * I * able variation in the nature of the relationship be- Belarus * . tween policies and growth. One hypothesis sug- Turkmenistan * - I 1 I l gests that a minimum critical mass of reforms needs 0 0.2 0.4 0.6 0.8 1.0 to be in place before economic reforms have the desired effects on performance. Below this thresh- Liberalization index old, it is possible that additional reforms could have a negative impact on output-that is, a clas- sic J-shaped response of output to policy reform. Source: de Melo, Denizer, and Gelb (1996); EBRD (2000). In this view, implementing a few limited re- forms could disrupt production in state enterprises 14 Explaining Variation in Output Performance without generating an attractive climate for re- entirely the same as those determining later structuring and new investment, resulting in lower economic performance. Some developments ac- output. One empirical test of this hypothesis (de companying the onset of transition, such as the Melo and Gelb 1996) finds that when countries breakdown of payment systems, are likely to have have made only limited progress in reform (de- had a stronger impact on the initial period of fined as a score of less than 0.4 on the liberaliza- the transition. Similarly, the effects of some un- tion index) additional reforms actually have a favorable initial conditions, such as repressed negative impact on growth. The response turns inflation, are likely to dissipate over time. positive once a minimum threshold of economic In our statistical analysis, the unbundling of reform has been reached. This suggests that there initial conditions into structure, distortions, and are important complementarities among elements institutions provides a more nuanced answer to the of a market-oriented reform program that have a question of the importance of initial conditions decisive impact on the relationship between re- versus policy reforms in explaining the recession forms and growth. and recovery periods of the transition experience Alternatively, it might take a long time for (annex 2.2). First, initial conditions are more imrl- policy reforms to exert their full impact on out- portant factors in explaining the differences across put, with the initially adverse consequences in countries during the initial period of output de- the short-run giving way to beneficial effects later cline (1990-94) than over the full 10 years of tran- on. The empirical results reported here imply that sition. The three (aggregate) indicators of initial policy reforms affect output growth over an ex- conditions defined earlier explain 51 percent of the tended period. Our analysis suggests that this variation in the average rate of growth across coun- "time lag" in the impact of reforms on perfor- tries during 1990-94, but only 41 percent of the mance can be estimated at approximately three variance in average growth during the decade. years (the hypothesis that the effect is entirely Second, our results suggest that different types contemporaneous can be strongly rejected; see of initial conditions were more significant in the Selowsky and Martin 1998). early and later stages of transition. Initial distor- There also are significant differences between tions in the economy-including factors such as the CIS and the other transition economies in the severe repressed inflation or high black market nature of the response of output to policy. A par- exchange rates and absence of pretransition policy ticular difference is on the immediate (contempo- reforms-are most closely associated with lower raneous) impact of reforms, which is negative in performance during the first years of transition. the CIS and positive in Central Europe. This means Initial institutions-including factors such as ab- that liberalization has an up-front "investment sence of "market memory," measured by the nunn- cost" in the CIS countries, consistent with greater ber of years under socialism, and general devel- economic distortions at the start of their transi- opment of national institutions, as determined by tions. A higher share of negative value activities the length of prior experience of nationhood-- required vast reallocation of resources after liber- are more strongly associated with variations in alization in the CIS. These countries also faced subsequent performance. greater obstacles to achieving that reallocation, Third, while initial conditions have a greater such as physical size, distance to external mar- impact on the initial collapse of output than on kets, large and inefficient company-town enter- the subsequent recovery, the impact of policies prises in isolated regions, and greater political and becomes stronger as the transition progresses-- constitutional turmoil inhibiting investment. although it is still significant in the early stages. Indeed, policy variables are statistically significant What ..ftial Conditions Matter and When Do in both periods, implying that market-oriented They Matter? policy reforms not only speed economic recovery and promote growth in the medium-term, but also T here are reasons to expect that the factors mitigate the effects of the transition recession in T explaining the initial output collapse are not the short term. 15 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Former Soviet Union What If Policies Themselves Are Endogenous? By contrast, advocates of slower reform point out that going ahead with reforms that can be P olicies, in general, are endogenous. The implemented quickly-"stroke of the pen" re- 1 choice might depend on initial conditions forms-without waiting for those that take more that might influence the possibility of policy con- time, such as the creation of institutions that sup- sensus behind reform (de Melo and others 1997). port markets, significantly reduces the benefits They might be influenced by the desire to reclaim of these reforms. The loss could be so severe as nationhood after the dismantling of the Soviet to generate output losses and also lead to the Union. Alternatively, what is feasible at one stage creation of interest groups opposing those re- might be partly determined by previous policy forms requiring more time. Some of these issues decisions. For example, private monopolies that are reviewed on part 3. emerged from privatization at an early stage of The amalgamation of different types of policy reform might attempt to block entry or resist a reforms into a single aggregate indicator prevents regulatory framework. Or new, small entrepre- statistical cross-country analysis from directly neurs may try to press for faster reforms in prop- shedding light on the desirability of progressing erty rights and in the court system. rapidly along all dimensions of reform. Some The literature reports some limited success indirect evidence is provided by the finding that in finding correlates to policy reform, such as output in each year is significantly associated the initial and contemporaneous level of politi- with the level of policy reforms achieved up to cal freedom (Dethier, Ghanem, and Zoli 1999; that year-that is, with cumulative policy reform. de Melo and others 1997). Part 3 of this report So the quicker a high level of liberalization is describes how choices about the structure of the reached and sustained the sooner the economy political system can shape the adoption of eco- can attain higher growth.5 nomic reforms. An important result from the These results are best seen as a broad-brush cross-country evidence is that policy reforms re- characterization of the main contours of transi- main strongly significant in explaining output tion. They do not provide the full story of the performance even when policies are taken as transition. Although policies and initial condi- endogenous (that is, when output growth and tions account for more than half the variability policies are jointly determined in a simultaneous of output growth across countries and years, they equation model; see de Melo and others 1997). still leave substantial room for other factors in- fluencing growth.6 A full explanation of output performance would have to include more coun- try-specific factors-as well as shocks and other T he relationship between the speed of reform omitted factors-and a detailed analysis of indi- T and economic growth has been the subject vidual countries or smaller groups of countries. of controversy. Some economists argued for ad- This is done in the rest of this report. vancing reforms in all areas as fast as possible; others criticized such a strategy as imposing un- Annex 2.1. Summary of Cross-Country necessarily high cost. The most interesting part Empirical Literature on Growth in Transition of the debate focuses on the sequencing of poli- cies-on the relative speed of different types of Economies reform. Advocates of moving fast in areas ame- A slund, Anders, Peter Boone, and Simon nable to rapid reform argue that the synergies AJohnson. 1996. "How to Stabilize: Lessons among different components-for example, from Post-Communist Countries." Brookings privatization together with liberalization of prices Papers on Economic Activity 26(1): 217-313. and trade-may generate enough gains and win- This paper finds differences in the determinants ners to maintain the reform momentum. The of output changes during 1989-95 (no significant need to take advantage of windows of opportu- association with policy reforms) and end-of-period nity is also cited as important in that decision. output level (liberalization and inflation significant). 16 Explaining Variation in Output Performance Berg, Andrew, Eduardo Borensztein, Ratna specifications that include 13-15 policy vari- Sahay, and Jeromin Zettelmeyer. 1999. "The ables, including lags, plus the controls and ini- Evolution of Output in Transition Economies: tial conditions (the precise number of para- Explaining the Differences." IMF Working Pa- meters estimated is not explicitly shown). per No. WP/99/73, International Monetary Some summary results: most of the variabil- Fund, Washington, D.C. ity in growth is associated with differences in An extensive exploration of the issues in model policies rather than initial conditions (depend[- specification with annual data for 26 countries ing on the time-declining impact of initial condi- from 1991 to 1996 (different period for a few tions); the difference between CIS and Central countries). A very general initial model includes: and Eastern Europe can largely be explained by * Macroeconomic variables (fiscal balance, in- differences in policies. flation, and exchange rate regime) Campos, Nauro F., and Fab.zio Concelli. 2000. * Structural reforms (Liberalization Index in its "Growth in Transition: What We Know, What three separate components, plus interactions We Don't, and What We Should." Centre for Eco- term defined by multiplying the Liberaliza- nomic Policy Research, London, United Kingdom. tion Index by the share of the private sector The authors use the contemporaneous Lib- in the economy) eralization Index and its components on their * Initial conditions (from de Melo and others 2001 inta.D e ait n rwh r regressions; only one component is statistically ba01inization, natura res e endgrowmth,u significant. Inflation is significant (with nega- banization, natural resource endowment, index of initial repressed inflation or actual tive effect on growth) and so is the presence of initia fiscal imbalance and inflation, share an International Monetary Fund program. In- intalrisclture imbale aependeinflt, a ashre stitutional variables are significant and positive of overicultrialization, and reforms before (rule of law and quality of bureaucracy). Initial collapse of central planning) conditions are measured as the principal com- * Other controls (average OECD growth, terms ponents of a set of reasonable-sounding indica- of trade, add s. tors (dependence on Council of Mutual Eco- of t , d snomic Assistance trade, repressed inflation, and In addition to the large number of variables, overindustrialization). they include first and second lag of macroeco- nomic variables and up to third lags of struc- de Melo, Martha, Cevdet Denizer, and Alan tural reform indexes. The initial conditions are Gelb. 1996. "From Plan to Market: Patterns of also parameterized by time to allow for their Transition." World Bank, Washington, D.C. impact to decline or dissipate after a period (but the precise specification and statistical tests are de Melo, Martha, Cevdet Denizer, Alan Gelb, not published with the paper). Both the level of and Stoyan Tenev. 1997. "Circumstance and GDP (in logs) and the annual rate of growth are Choice: The Role of Initial Conditions and Poli- used as endogenous variables. cies in Transition Economies." World Banlk, The initial specification contains too many Washington, D.C. variables to get significant results about effects These are two influential papers that intro- of individual variables, but it can be used to test duced the measures of policy reform and initial the hypothesis that certain categories of variables conditions most widely used in the literature. The are irrelevant. Thus the model rejects strongly first paper introduced the Liberalization Index, the hypothesis that none of the macroeconomic defined as a weighted average of policy reforms policy variables matters and/or that none of the in three areas: internal markets, external mar- structural variables matters for growth. kets, and privatization and private sector entry. The initial broad specification is progres- The paper's empirical analysis emphasizes "pat- sively simplified by eliminating variables with terns of transition" rather than models of policy low statistical significance to arrive at two final response and statistical test of hypotheses. 17 Transiion-The hrst Ten Years: Analysis and Lessons for Eastem Europe and the Former Soviet Union The second paper focuses on the role of initial and a "reform" pattern of growth, and coun- conditions, for which the authors calculated a set tries switch from the former to the latter when of 12 indicators widely used in subsequent work the CLI reaches 0.4 (that is, changes in CLI at (initial income, urbanization, natural resource lower or higher levels do not have any effect; endowment, location, pretransition reforms, ini- only those getting CLI over the threshold have tial repressed inflation, overindustrialization with an impact). On the other hand, the two "pat- respect to Chenery's "norms", shares of trade with terns" are estimated with high degree of gener- Council of Mutual Economic Assistance, black ality (with up to 13 parameters). The two re- market rate for foreign exchange, new versus old sulting patterns are plausible (for example, states, andyearsundercentralplanning).Thelarge "reform" has an initial drop in output but it is number of indicators is reduced to a more man- increasingly positive after the second year), and ageable set of two by the method of principal com- the hypothesis that they are equal (that is, that ponents. Thus the transformed variables are de- policies do not matter) can be strongly rejected. fined as linear combinations of the initial variables in a way that preserves the maximum of their Fischer, Stanley, Ratna Sahay, and Carlos A. variability. This is useful for statistical power, but Veight. 1998. "How Far is Eastern Europe from makes it more difficult to determine what initial Brussels?" IMF Working Paper No. WP/98/53, conditions really matter. International Monetary Fund, Washington, D.C. The authors estimate a model where growth The authors find that the Liberalization In- (in 1992-95) is explained by initial conditions, dex and stabilization variables (inflation and fis- policy reforms (measured by the aggregate Lib- cal deficit) are significantly associated with eralization Index or its cumulative value), and growth. They introduced "reform time" to ad- a war dummy variable. The authors also esti- just for different starting points of the transition mate a simultaneous equations model for process in different countries. growth and the Liberalization Index (with lib- eralization policies determined by initial con- Havrylysyhyn, Oleh, and Ron van Rooden. 1999. ditions plus a war dummy and an index of po- "Institutions Matter in Transition, but so Do Poli- litical freedom), but find that there is negligible cies." Paper presented at the Fifth Dubrovnik simultaneity bias on the growth equation. Both Conference on Transition, Dubrovnic, June 1999. initial conditions and the Liberalization Index The emphasis is on variables defining insti- are quite significant in the growth equation. The tutional development: "Index of Economic Free- paper also presents the "growth patterns" dom" from the Heritage Foundation (1994-97); model discussed below, augmented by the two indicators for democracy, rule of law, and oth- variables for the initial conditions. ers from the "Nations in Transit" reports from Freedom House; institutional environment from de Melo, Martha, and Alan Gelb. 1996. "A Com- the survey for the World Bank's World Develop- parative Analysis of Twenty Transition Econo- ment Report 1998/99: Knowledge for Develop- mies in Europe and Asia." Post-Soviet Geogra- ment (1998); and country risk ratings from phy and Economics 37(5): 265-85. "Euromoney." They follow de Melo and others The authors discuss "patterns of transition" (1997) on the treatment of initial conditions (the for the 26 European and Central Asian coun- same variables are aggregated into the same two tries plus China and Vietnam. The exogenous principal components). A similar aggregation ap- variables are the Cumulative Liberalization In- proach is used to summarize the eight institu- dex (CLI), plus dummy variables for regional tional variables into one or two principal com- tension and for Central and Eastern Europe. ponents. One innovation is to include a simple However, the paper postulates a model in which form of time dependence for initial conditions there is discontinuity when CLI = 0.4. This is (impact at t is bfc/t) and instrumental variables so because the authors define a "nonreform" (impact on t is t*bV). 18 Explaining Variation in Output Perlonnance Hernandez-Cata, Emesto, 1997. "Growth and openness from normal level + share of trade with Liberalization during the Transition from Plan FSU + share of trade with Central and Eastern to Market." IMF Staff Papers 44(4): 405-29. Europe/3), the initial level of GDP per capita, The paper models the reallocation of capital the decline in government revenues/GDP, and from the old to new sector (as in Berg and others shadow economy/GDP. 1999). The paper confirms the significance of lib- eralization and stabilization, with similar impacts Selowsky, Marcelo, and Ricardo Martin. 1998. on the CIS and Central and Eastern Europe. "Policy Performance and Output Growth in the Transition Economies." American Economic Heybey, Berta, and Peter Murrell. 1999. "The Review-Papers and Proceedings 87(2): 350-53. Relationship between Economic Growth and the The paper uses combined cross-section and Speed of Liberalization during Transition." Jour- time series data to test a model, allowing for a nal of Policy Reform 3(2): 121-37. delayed impact of reforms on growth, with crude The authors find that average growth over the control of initial conditions (Central and Eastern first four years of transition does not depend on Europe versus the CIS) and other factors (dummy the increase in the Liberalization Index over the variables for period of war or internal conflict). period, after accounting for possible endogeneity The Liberalization Index is found to be highly sig- of liberalization policies (the instruments used are nificant and to exhibit significant differences be- the initial level of liberalization, share of indus- tween its initial and long-term impact. There are try, and an index of political freedom). The au- also differences in the dynamic impact of reforn thors conclude that initial conditions are "much in Central and Eastern Europe and the CIS. more important than policy variables." Popov, Vladimir. 1998. "Will Russia Achieve Fast Annex 2.2. Additional Empirical Analysis Economic Growth?" Communist Economies and s mentioned in the text, the authors under Economic Transformation 10(4): 421-35. AL took additional empirical analysis in prepa- ration for this paper to elucidate the types of ini- Popov, Vladimir. 1999. "Shock Therapy versus tial conditions that made a difference for couI- Gradualism: The End of the Debate." Carleton try performance, as well as possible differences University, Ottawa. in the role of initial conditions and policies at different stages of the transition. Castanheira, Micael, and Vladimir Popov. 1999. The analysis of initial conditions was based "Framework Paper on the Political Economy of on the indicators developed by de Melo and otlh- Growth in Russia and Central and Eastern Eu- ers (1997), plus the terms of trade loss estimated ropean Countries." Global Development Net- by Tarr (1994). The 13 indicators were divided work, World Bank, Washington, D.C. Processed. in three groups and then aggregated into three The authors use only the cross-section vari- "synthetic" indices-initial structure, initial dis- ability in the data as they take average rate of tortions, and initial institutions. Table A2.1 growth (over the whole sample or separate shows that the three indices "explain" (as incli- subperiods) as a dependent variable. The Liber- cated by the R-squared of the regressions) 51 alization Index is found significant for growth percent of the variance on average growth across recovery (1994-98), but not for the overall pe- countries during the initial transition recession, riod (1989-98) or the subperiods (1989-96 and and 44 percent during 1995-99. 1996-98). Other control variables: war dummy Table A2.2 presents the results of the regres- and average inflation. Initial conditions are ac- sion of the annual rate of growth for all countries counted for by including an Index of Initial Dis- on the Liberalization Index and initial conditions, tortions (defense expenditures/GDP -0.03 + de- controlling for conflicts (war dummy), allowing viations in industrial structure and trade for a dynamic impact of policies-as in Selowsky 19 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Former Soviet Union TABLE A2. 1. and Martin (1998)-and for different effects dur- ing the early and late 1990s. The table shows that Regression of Average Growth on Initial Conditions, there are differences in the impact of policies and 1990-99 initial conditions in the two subperiods: liberal- ization policies have a stronger positive impact Average growth, Average growth, during 1995-99, while initial conditions have a Initial conditions 1990-9 4 1995-9 9. ' stronger impact on the earlier period. In both cases Structure -0.30 0.18 the difference is statistically significant, as the Distortions 1.03a 0.24 hypothesis of equal effect in both periods is re- Institutions 0.25 0.82a jected at a high level of confidence. R-squared 0.51 0.44 a. Significant at 95 percent. Notes Source: Martin (2000). 1. The three aggregate indicators (structure, distor- tions, and institutions) are a linear combination of the individual components, with weights equal to their TABLE A2.2. respective coefficients in a regression on average Regression of Annual Output Growth on Policies and growth. Initial Conditions Allowing for Differential Effects Early 2. The Council of Mutual Economic Assistance is in Transition, 1990-99 the free-trade area encompassing the Soviet Union and Eastern Europe. Endogenous variable: 3. Easterly (2000) shows that growth in most devel- Annual rate of growth oping countries slowed in the past two decades. Me- 1990-99 dian per capita income growth was 0 percent, com- pared with 2.5 percent in 1960-79. The slowdown is Exogenous variable Coefficient t-value attributed to the deterioration in the external envi- ronment, specifically lower growth in OECD coun- Liberalization (t) -11.01a -4.08 tries, particularly Europe and Japan, and higher real Liberalization (t- 1) 8.92a 2.60 interest rates. Liberalization (t - 2) 3.32 1.60 Liberalization (t), 1995-99b 8.26' 2.02 4. These indicators, developed by de Melo, Liberalization (t- 1), 1995_99b -4.03 -0.71 Denizer, and Gelb (1996), stop in 1997. More re- Liberalization (t - 2), 1995_99b -2.18 -0.60 cent work uses transition indicators of the Euro- War -11.52a ~~~~~~~~~~pean Bank for Reconstruction and Development, War -11.52a -5.45 which include 1998 and 1999. Initial conditions 0.09 0.38 Initial conditions, 1990-94' 1.82a 4.50 5. The usual caveats about attributing causality to R-squared (number of observations) 0.60 (200) regression coefficients apply. However, the coefficients Standard error of estimate 689 still provide the best inference available from the corn- bined experience of all transition economies. a. Significant at 95 percent. 6. For example, these two factors would allow b. The coefficient for thesc variables measured the additional ef- "predicting" the rate of GDP growth only with an fect of liberalization during 1995-99. uncertainty band roughly 11 points wide. More tech- c. The coefficient for this variable measured the additional effect nically, the estimated standard error of the regres- of initial conditions during 1990-94. sion with the percentage rate of growth as depen- Source: Martin (2000). dent variable is about 6.89, so that a two-way 90 percent confidence band would have a width of 11.37 points (1.65 times the standard error). This is an extraordinarily large number compared with variations in growth in the OECD countries. 20 Part 2 Policy and Institutional Challenges Ahead The Quest for Growth C ountries started the transition facing common challenges. The production system was de- signed for the exigencies of a command economy. Much industrialization was based on cheap energy and subsidized transport. In addition, the coordination and monitoring of central planning meant few links among enterprises, with large state enterprises forming production and delivery chains of vertically integrated, bilateral monopolies or monopsonies under the aegis of branch ministries. Opening to the world revealed productivity differences across sectors and enterprises, given the energy-intensive structure of production. In April 1992 the Russian Federation's domestic oil price was still only 3 percent of the world price. Shutting down any links in those extensive chains of production and delivery caused adverse output and employment effects to cascade through the economy. Together, changes in relative prices and dislocations of the production system meant that many enter- prises subtracted rather than added value at the new prices. All countries faced this problem at the beginning of transition. To realize the beneficial effects of liberalization, two further policy responses were required. The first was to impose market discipline on inherited enterprises-so that they would face the incentive to restructure and, in so doing, become more productive and able to compete at the new prices. Failure to do so would lead to closure. The second was to encourage the creation of new enterprises, which became possible after liberalization created the legal opportunity for private investment. It is reasonable to assume that investment in new enterprises would be undertaken only with an expected rate of return at least equal to what could have been realized by investments in an enterprise undergoing restructuring. Since restructured enterprises are, by definition, more productive than old enterprises, this yields a productivity ranking (figure 3.1). The yardstick used here and throughout the report is to rank enterprises by labor productivity, not total factor productivity. (The circumstances under which this is broadly permissible are explored in annex 4.2.) Empirical evidence collected fromn a broad cross-country survey of enterprises (box 3.1) throughout the region confirms that new enter- prises tend to outperform old enterprises along every dimension of performance (figure 3.2).' 23 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Former Soviet Union FIGURE 3.1. Productiviy Distribution of Old, Restructured, and New Enterprises Range of 0 old enterprises ' b Productivity of old enterprises Range of 0 restructured enterprises L-o. Productivity of restructured enterprises Range of 0 new enterprises I__ Productivity of new enterprises Note: The figure allows for outliers in both directions, as there is no reason why single old enterprises, everything else being equal, might not occasionally produce higher value added per employee than new enterprises, or why new enterprises might not occasionally have disappointing results. Source: World Bank data. Aggregate growth in the economy reflects Ruehl and Vinogradov 2001). The pattern of the interplay between old enterprises in need of growth is initially dominated by the negative state support that, by absorbing more resources contribution of old enterprises, which causes than they produce, reduce growth, and restruc- output to decline. Conditions at the beginning tured and new enterprises, which increase it (see of transition-such as repressed inflation, Box 3.1. The Business Environment and Enterprise Performance Survey The World Bank and the European Bank for Reconstruction and Development conducted a large survey of enterprises in 20 transiton economies in the early summer of 1999, adding five more transition economies later that year. The survey used face-to-face interviews with enterprise owners or senior managers and was conducted by the same international survey enterprise across all the countries (with the exception of Albania and Latvia, where local survey enterprises were used). The aim of the survey was to investigate how enterprise behavior and performance were related to and affected by the quality of the business environment and the relationship between enterprises and the state. The survey posed detailed questions about the enterprises' business and competitve environment and about the different restnrcturing actions they had taken in the recent past. (See Hellman and others 2000 and www.worldbank.org/wbi/governance/datasets.htm for a full description of the survey.) Sampling was random from the population of enterprises in each country, except that minimum quotas were imposed for state-owned and large enterprises. Box table A provides some basic information on the distribution by region, size, origin, sector, and location of the enterprise sample. Tne survey included some 125 enterprises from each of the 25 countries, with larger samples in Poland and Ukraine (almost 250 er'lerprises) and in Russia (more than 550 enterprises). The full sample comprises 3,954 enterprises, more than half from the Central and Eastem European region (including the Baltics) and the rest from the countries of the CIS. The sample is dominated by small and medium-size enterprises (SMEs); half of them employed fewer than 50 people, and only 8 percent employed more than 500. More than half the enterprises were newly established private enterprises, 27 percent were privatized, and the remaining 16 percent were state owned. The enterprises are divided fairly evenly between industry (52 percent) and services (48 percent), with 30 percent of enterprises from the manufacturing sector. Almost 50 percent of enterprises are located in large cities or national capitals, with the rest in small towns and rural areas. (box continues on following page) 24 The Quest for Growth Box 3.1 coNinNuED TABLE A. Characteristics of the Business Environment and Enterprise Perfomance Survey Sample, 1999 Number of enterprises Percent Region CIS 1,866 47.2 Central and Eastern Europe and the Baltics 2,088 52.8 Size SmalL (fewer than 50 workers) 1,944 49.2 Medium (50-500 workers) 1,690 42.8 Large (more than 500 workers) 318 8.0 Origin New 2,176 56.5 Privatized 1,050 27.2 State-owned 627 16.3 Sector Industry Farming 453 11.5 Mining 33 0.8 Manufacture 1,191 30.1 Construction 343 8.7 Power generation 16 0.4 Total 2,036 51.5 Services Trading 541 13.7 Retail 571 14.5 Transport 232 5.9 Finance 67 1.7 Personal services 214 5.4 Busiriess services 245 6.2 Communications 15 0.4 Other 30 0.8 Total 1,915 48.5 Location Capital city 1,220 30.9 Large city 704 17.8 Town 1,694 42.8 Rural 336 8.5 Source: Hellman and others (2000). Given the sample size and specific quotas, the survey cannot be used to measure the numberand share of new enterprises in each country. However, it can provide valuable evidence about how the performance and behavior of new enterprises differ from stateowned and privatized enterprises. It also provides an opportunity to test whether perceptions of the business environment differ systematically across different categories of enterprises. (For a full analysis of the data on small enterprises surveyed, see EBRD 1999). multiple exchange rates, and the terms of trade this depends on policy choices. Differences in losses associated with trade liberalization and conditions at the end of communism-and in the breakup of the Council of Mutual Economic exogenous shocks and policy choices in the Assistance trading area, which would in turn 1990s-put countries in vastly different circum- have been reflected in the inherited capital stances today. stock-were important determinants of this de- For many countries in the CIS-and for cline. With time, restructured and new enter- those in Southeastern Europe, such as Bulgaria, prises acquire the critical mass needed to over- FYR Macedonia, Romania, and the Federal Re- come the negative effects of old enterprises to public of Yugoslavia, which have seen steep de- generate economywide growth. The speed of clines in incomes since the onset of transition-- 25 Transition-The irst Ten Years: Analysis and Lessons for Eastern Europe and the Forner Soviet Union public goods and protecting the most vulner- FIGURE 3.2. able, is not enough for these countries. The .ene 1996-99 advanced reformers in Central Europe and the (percentagerm hance) ofOldandNewEnerrses,Baltics need to consolidate the gains of the first (percentage change) decade of transition and address what could be Sales called "second-generation" issues in the transi- 15 - tion toward an effectively functioning market economy. They have to secure control over 10__ quasi-fiscal and contingent liabilities. They have to undertake reform in labor and financial mar- Debt / > < \ Investment kets to allow the benefits of growth to be more widely shared across the population. They also -5 have to restructure social sector expenditures to make them fiscally more affordable without impairing the social safety net. Many of these reforms overlap with those required to join the European Union. Exports Employment A Tale of Two Approaches New enterprises Old enterprises T ooking at the assets of old, restructured, and L new enterprises reveals that countries restor- Source: Hellman, Jones, and Kaufmann (2000). ing sustained growth need to create a policy en- vironment that simultaneously disciplines old en- terprises (state enterprises, privatized but unrestructured enterprises, and agricultural col- restoring sustained growth and rebuilding the lectives) and encourages "new' enterprises (in- state are key priorities. cluding both greenfield investments and restruc- Without growth it will be impossible to gen- tured spinoffs from newly privatized enterprises). erate income-earning opportunities for house- Discipline entails hardening budget con- holds or to generate the resources to provide such straints (a concept owed to Kornai 1986), intro- basic public goods as a legal and judicial system, ducing competition to product markets, moni- secure property rights, and basic infrastructure. toring managerial behavior to generate incentives Nor will it be possible to maintain investments for efficient resource use and prevent such abuse in education and health that have been adversely as asset stripping and tunneling (box 3.2), and affected since the onset of transition, or to put providing viable exit mechanisms for inefficient in place a social safety net targeted toward the enterprises. It thus forces old enterprises to re- most vulnerable. Without a functioning state lease assets and labor, which then become avail- there will be no capacity to provide public goods able for more efficient reallocation to restruc- even if the resources were available. tured and new enterprises. Old enterprises also Growth is also important for the more re- divest themselves of social assets-such as hous- form-oriented countries in Central Europe and ing, health clinics, and kindergartens-which re- the Baltics. Per capita incomes in the three quires shifting the locus of social protection from wealthiest countries seeking accession to the enterprises to local governments. European Union as a fraction of the European Encouragement entails reducing excessive Union average are still only 68 percent for marginal tax rates, simplifying regulatory pro- Slovenia, 59 percent for the Czech Republic, cedures, establishing secure property rights, and and 49 percent for Hungary. However, an ex- providing basic infrastructure. That allows re- clusive focus on growth, while providing basic structured and new enterprises to absorb labor 26 The Quest for Growth Box 3.2. The Problem of Tunneling Tunneling is the legal expropriabton of income and assets belonging to minorty shareholders (Johnson, La Porta, Lopez-de-Silanes, and Shleiter 2000). Diverting cash flow and asset stripping are forms of tunneling. Managers usually do the tunneling and in purely private enterprises are usually acting at the behest of large or controlling shareholders. In partly state-owned enterprises, managers are usually acting by themselves or at the behest of private shareholders. Tunneling should be distinguished from theft, which is illegal. However, if expropriation by management occurs in a fully state-owned enterprise, this is almost always illegal. Tunneling should also be distinguished from rent seeking. Rentseeking refers to the efforts of enterprises to obtain advantages through privileges or subsidies granted by the government. Rents are usually extracted from a wide cross-section of society (for example, from all the consumers of a particular product). In contrast, the primary impact of tunneling is on minority shareholders (or the state, if it is a shareholder in the enterprise), although the evidence suggests there are also negative effects on the economy as a whole. Expropriation of shareholders has been a particular problem in some transition economies, particularly in the CIS. Some forms of tunneling were specific to the early transition, such as finding ways to legally expropriate state assets, though new forms have evolved over the decade. Nevertheless, tunneling is not a phenomenon exclusive to the transition. In the aftermath of the Asian financial crisis of 1997-98, for example, several companies are alleged to have engaged in some form of tunneling. Why Is Tunnelng Legal Tunneling is legal for two reasons. First, there may be important loopholes in the legal protection of investor rights. Laws goveming 'related party" transactions are often vague or inadequate. Consider the case where company A is controlled by the managers of company B and the two companies enterinto a transaction. If companyA supplies inputs at above-market prices (orsells assets at artificially low prices) to company B, someone is effectively tunneling the value out of B and into A through what is known as 'transfer pricing." Unless the law specifies clearly that related party transactions require full disclosure and supervision by independent parties (as in the United States), tunneling through transfer pricing is easy and legal. Second, the courts may be unwilling or unable to apply even precise statutes. There are cases in Westem Europe where courts have looked at instances of tunneling and pronounced it legal, For example, in civil law countries, if the courts cannot determine whether someone was harmed, they will often not punish an action that cleariy violated a statutory provision (Johnson, La Porta, Lopez-de-Silanes, and Shleifer 2000). Most transition economies have judicial systems based largely on civil law, and judges, in such instances, may be reluctant to interpret the law as preventing tunneling in complex situations. Tunneling can take a variety of forms. In countries with weak protection of shareholder rights, it is not uncommon for enterprises to consistently transfer small amounts of shareholdervalue out of companies. Such "seepage' of shareholder value is often considered an ordinary cost of doing business by investors in countries with a weak rule of law. But the evidence indicates that in environments with poor protection of shareholder rights corporate dividends are smaller, the valuation of companies is lower, and financial development is slower than in countries with stronger protection of shareholder rights (Johnson, La Porta, Lopez-de-Silanes, and Shleifer 2000). There is also some evidence that such seepage reduces the growth of and limits investment in capitalintensive sectors. In periods of economic crisis in a contextof weak investor protection, tunneling can increase substantially. If managers foresee theirenterprise's demise, they have strong incentves to tunnel enterprise assets. The Russian economic crisis of 1998 reportedly triggered widespread and extreme tunneling (Johnson, Boone, Breach, and Friedman 2000). The costs of such tunneling can be enormous. If investors expect that tunneling has become widespread, the entire capital allocation process in the economy can be disrupted, seriously delaying any economic recovery. Moreover, such expectations can lead to a precipitous drop in the economic value of enterprises due to strongly negative investor sentiment Tunneig and Capital Flight Theoretically, tunneling can remain purely domestic; that is, wealth can be transferred from minority shareholders to managers while remaining inside the country. Empirically, however, evidence suggests that episodes of large-scale tunneling tend to coincide with high levels of capital flight (Johnson, Boone, Breach, and Friedman 2000). There are several reasons for this correlation. The risk that existing laws permitting tunneling could be changed retrospectively creates strong incentives for holding tunneled assets abroad, out of reach of domestic authorities. Moreover, in an environment characterized by poor protection of investor rights, assets tunneled from enterprises are hardly well protected in (box continues on following page) 27 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Former Soviet Union Box 3.2 CONnNUED domestic banks that suffer from the same weaknesses of the broader environment. If managers can tunnel assets freely from their own enterprises, why should they expect domestic bankers to do otherwise? Despite the links between tunneling and capital flight, tunneling can also be used to prop up related enterprses during temporary periods of poor performance. In such cases tunneling becomes a form of cross-subsidizaton within a related group of companies. In environments where enterpnses might have limited access to short-term credit because, for example, the banking system is weak, 'propping' temporarily troubled enterprises through transfers from related companies could ease access to capital in the future. So, despite the negative consequences of tunneling, the function of propping as a backstop in insufficiently developed financial markets could explain why investors are still willing to invest in such an environment (Shleifer and Wolfenson 2000). Increasing Vulnerability to Economic Crisis Institutions protecting entrepreneurs against government expropration are crucial to transition and economic development generally. A lack of protecton for the property rights of entrepreneurs hampers growth. However, to grow in an environment where investors are not fully protected against expropriation by other entrepreneurs-that is, where tunnel- ing is common-is not impossible. The main effects of tunneling are to limit financial development and divert resources toward sectors that are not capital intensive. Nevertheless, weak investor protecton and widespread tunneling do increase a country's vulnerability to economic crisis. Note: Tunneling was first identified analytically by Jensen and Meckling (1976), who focused on the United States, where tunneling is limited. Shleifer and Vishny (1997) survey the literature through 1996. and assets made inexpensive by the downsizing. Discipline and Encouragement By creating a stable and predictable business en- vironment, the policies of encouragement gener- As the case of Poland suggests, an important in- ate incentives for enterprises to invest. As more gredient of market discipline is the hard budget and more new enterprises enter the market, an constraint on state enterprises, together with suf- increasingly competitive environment develops. ficient standards of corporate governance to pre- At the same time the policy environment must vent large-scale asset stripping before privatization. restrain predatory behavior by new enterprises It was understood that no subsidies would go to that seek to extract special preferences from the state enterprises after the beginning of 1990, but state and, in so doing, erect barriers to competi- it took 18-24 months for the government's com- tion and further entry. mitment to hard budget constraints to be seen as The more advanced reformers in the CSB, fully credible by enterprise managers. Poland's now facing second generation issues, are much economic growth-which resumed in 1992, the farther along in implementing an environment earliest among the transition economies-was first of discipline and encouragement. Discipline has due largely to better use of existing assets by en- been established, but it needs (as in Poland) to terprises spun off from state enterprises. Not until be maintained, because loss-making state en- 1995 was there a big boom in domestic invest- terprises in coal mining, steel, and railways still ment-and not until 1996, with Poland in its fifth impose a costly burden on the budget. In these year of growth, was there a take off in foreign countries new and restructured enterprises still direct investment. lead growth, but unemployment remains stub- At the same time successive governments un- bornly high. In addition to maintaining the dis- dertook structural reforms to generate an invest- cipline established during the first decade of re- ment climate favorable for the entry of new en- form, the focus of encouragement for job terprises, in particular small and medium-size creation in new enterprises needs to be remov- enterprises (SMEs). Like many other countries ing bottlenecks in infrastructure and reforming with a socialist past, Poland had a high indus- labor and financial markets and the social pro- trial concentration, with the leading enterprise tection system. having more than 30 percent in more than 60 28 Tlhe Quest for Growth percent of the markets (at a three-digit level). The result was to prevent or postpone closure The new enterprises signaled that product mar- of the least productive old enterprises and the re- ket competition would press state enterprises to structuring of enterprises with good prospects. In become efficient. During 1990-98 the number addition, the entry of new enterprises that would of individual- and family-owned enterprises rose be viable without state support was restrained. from 1.2 million to nearly 2.8 million. Similar The protect-and-discourage strategy thus creates figures held for Hungary. In both countries the an environment where resource transfers flow in share of employment and value added of new a direction opposite to that in a discipline-and- enterprises, the engines of growth, rose to 50 encourage environment. percent or more. The Associated Fiscal Adjustment... Protection and Discouragement he two contrasting approaches of discipline The disposition of assets among old, restructured, T and encouragement and protection and and new enterprises also provides a useful per- discouragement are also broadly mirrored by the spective on such countries as Romania and Rus- associated fiscal adjustment that has taken place. sia. They have protected rather than disciplined Enterprises used to be a captive source of revenue old enterprises through subsidies-granted in transition economies. The state's loss of con- through the budget, energy consumption, and trol over them also meant the loss of fiscal con- the banking sector. In addition, their institutions trol and the need for political acceptance of a re- of public and corporate governance are not duced public sector. Government revenues as a strong enough to prevent asset stripping. They share of GDP fell from around 38 percent in 1992 discourage or at best only selectively encourage to 31 percent in 1998 in the CIS and from 44 the entry of new enterprises because the state's percent to 39 percent in the CSB. Stabilizing in- capacity to provide key public goods is weak and flation, in practice often the first order of busii- the investment climate poor. Tax rates are high. ness, required that expenditures be reduced too. Licensing and registration procedures are open They fell precipitously from 57 percent of a rap- to abuse. Furthermore, the legal and judicial sys- idly declining GDP in 1992 to 37 percent in 1998 tem is unable to enforce property rights. in the CIS-and, less sharply, from 45 percent in Adding to this discouragement, subnational 1990 of a more modestly reduced GDP to 41 per- governments engage in anticompetitive practices cent in 1998 in the CSB. This situation gave rise to protect established enterprises at the expense to two kinds of adjustment. of new enterprises in their jurisdictions. Examples First, a substantial amount of spending was of protection abound. Transfers to selected enter- moved out of the budgetary arena. Explicit bucl- prises and conglomerates in Russia getary subsidies to enterprises generally fell across in 1992 through credit at highly subsidized rates most of the region, but enterprises were still sup- are estimated at 33 percent of GDP, financed ported in a variety of ways. Implicit subsidies through a massive inflation tax on households and were channeled largely through the energy sec- enterprises without political connections. Subsi- tor, which would then pass the costs back in the dies implicit in soft budgets amounted to another form of arrears to the budget. In addition, bank 5 percent of GDP in 1996 and 3.5 percent in 1997. loans were rolled over and the enforcement of Romania reverted in 1994 to directed credit, tax and other rules was lax. In Russia, for ex- price controls, and budgetary and extrabudgetary ample, explicit budget subsidies to the enterprise transfers-and reopened enterprises earlier de- sector declined from 10.2 percent of GDP in 1994 clared closed. These attempts to use the economy's to 5.9 percent in 1998. But total budget subsi- old capital stock precipitated a macroeconomic dies to the enterprise sector, which also include crisis in 1996. Again, such transfers were financed the net increase in tax arrears as well as inflated through implicit or explicit taxes on households, prices of goods procured by the government and new enterprises, and enterprises that restructured paid for by offsetting tax arrears, rose as a share enough to survive the market test. from 10.9 percent of GDP to 16.3 percent. 29 Transition-The First Ten Years: Analysis and Lessons for Eastem Europe and the Former Soviet Union Moreover, implicit subsidies extended through decline in employment significantly smaller than bank loans and public utilities were siphoned off the massive collapse of output and labor demand. along the way, coexisting with wage payment ar- The adjustment took the form of lower real rears by enterprises, thus not helping workers who wages as well as the emergence of arrears and would have benefited if soft budget constraints nonpayment of wages. Without vibrant new en- had indeed been an instrument of social protec- terprises, labor moved to low-productivity ser- tion. Faced with a shrinking tax base and more vices and subsistence agriculture. Together with and more claimants being assisted outside the labor hoarding by enterprises, these sectors budget, governments attempted to collect more served as shock absorbers in view of the lack of revenue by raising taxes. This discouraged poten- a functioning social policy. tial entrants and, with abuses of discretion by gov- The second pattern, broadly prevalent in the errment officials, drove them underground.2 In ad- CSB, saw employment decline with output. Job dition, this pattern of fiscal adjustment, by destruction was concentrated in existing enter- signaling business as usual, helped protect enter- prises, while job creation was to be found almost prise managers from restructuring or closure. exclusively in new enterprises. Here too there is A second mode of fiscal adjustment reallocated cross-country variation. For Poland, which exem- public expenditures to the social sectors to cush- plifies the discipline-and-encourage approach to ion the impact of transition on the vulnerable. transition, enterprise restructuring in key sectors That made it politically possible to discipline old of the economy and concomitant increases in la- enterprises into shedding labor. In Poland, for bor productivity meant that output, growing rap- example, pensions were critical in preventing the idly since 1992, has outpaced job creation to a elderly from falling into poverty. Indeed, high so- point where unemployment stood at 17 percent cial spending was more affordable in the CSB in August 2001. Budget constraints were softer because government revenue was significantly and enterprise restructuring more limited in the higher than in the CIS countries. That also helped Slovak Republic, which grew at an annual aver- to create a constituency for the discipline neces- age rate of 4.7 percent between 1994 and 2000. sary to ensure a return to growth. The stop-and-go nature of the reform effort re- However, this alone would have been un- sulted in an insufficient creation of jobs in new sustainable without the rapid growth of new firms. Employment remained largely unchanged enterprises. Their high labor productivity gave and unemployment reached nearly 19 percent in them the potential to offer displaced workers a the second quarter of 2000-the highest rate of viable outside option rather than a return to unemployment in Central Europe. subsistence activities, and that created a con- In both countries, the aggressive use of social stituency for encouragement. The CSB still re- sector expenditures in the form of generous so- sorted to off-budget activity and created con- cial assistance and unemployment insurance pro- tingent liabilities. But these countries show the grams cushioned the risk of poverty in the face of overarching role of fiscal policy and the insti- high unemployment. But as in many countries in tutions of budget management in supporting a Central Europe and the Baltics, these programs growth-oriented adjustment by redistributing diminished the incentives for workers to look for part of the "reform dividend" to those who jobs. This, together with high payroll taxes, has would otherwise bear its costs. constrained the growth of employment. ... And the Role of Labor Markets Notes A s with fiscal policy, two broad patterns of 1. "Old" enterprises are defined as enterprises es- Ax lab or market adjustment can be identified. tablished before 1989 in Central Europe and the The first, largely associated with the CIS and the Baltics and before 1991 in the CIS. Increments in countries of Southeastern Europe, involved a growth rates are 5 percent, with -5 percent as the 30 The Quest for Growth origin and zero the first increment, to capture the 2. Estimates of the unofficial economy as a share of negative export growth rate reported by the old en- GDP in 1995 are 42 percent in Russia, 49 percent in terprises in the sample for 1996-98, a time when Ukraine, and more than 60 percent in Azerbaijan and overall growth across the region had picked up. Georgia (Johnson, Kaufmann, and Shleifer 1997). 31 Discipline and Encouragement W hat policies and institutions are required to bring about a discipline-and-encouragement environment? While no one policy can be assigned to a single outcome in an interrelated system, it is convenient to think of policies in two groups: those primarily directed at disciplining the old sector and those primarily directed at encouraging the new sector. Discipline requires imposing hard budget constraints on enterprises and banks. This entails elimi- nating a wide range of explicit and implicit mechanisms to channel public resources to enterprises and banks, including tax exemptions, fiscal and financial subsidies, budget and tax offsets, directed cred- its, and contingent liabilities. However, discipline also refers to measures to prevent the misuse or theft of assets in both privaite and state-owned enterprises through asset stripping, tunneling, and expropriation of minority share- holders. To prevent such abuses, incentives for managerial behavior need to be aligned with the goal of enhancing efficiency through such reforms as privatization, strengthening the legal framework (particularly the enforcement of property rights), bankruptcy regulation, accounting reform and dis- closure, and creditors' and shareholders' rights. Some countries, such as Belarus, Turkmenistan, and Uzbekistan, have managed to maintain discipline over managers in state enterprises without liberal- ization or hard budget constraints, but this has occurred largely through administrative methods held over from the Soviet-style command economy. Encouragement starts with liberalizing prices and trade to enable the entry of new enterprises. But it goes much further to encompass policy and structural reforms that promote an attractive invest- ment climate. There are an enormous number of factors that affect the decisions of economic actors to invest. These can be roughly divided into two categories: the quality of public goods and the extent of nonmarket obstacles to competition. The key public goods that most directly affect the quality of the investment climate are a legal and judicial system capable of enforcing contracts and protecting property rights, a social system that pro- motes the development and maintenance of human capital, a macroeconomy that ensures stability over time, a banking system that provides effective financial intermediation, and a network of basic infrastruc- ture. Public expenditure must be prioritized to provide the goods that promote investment and growth. 33 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Former Soviet Union Tearing down the obstacles that discourage early in the transition. However, the capture investment and competition is just as important. of the state by a narrow set of vested inter- The inheritance of the command economy in- ests-old enterprises and well-connected early cluded stifling administrative barriers to entry, entrants-discouraged further entry at later over-regulated labor markets, and tremendous stages of transition and created a poor invest- discretion for bureaucrats. New barriers have ment climate. This resulted in a pattern of pro- emerged in the transition. The most prominent tection and selective encouragement. of these has been an overly complex and * Belarus, Turkmenistan, and Uzbekistan, distortionary tax regime that pushes new entre- which neither liberalized nor hardened bud- preneurs into the informal economy. These ob- get constraints, strongly discouraged new stacles create an environment in which corrup- entry. Access to foreign exchange and credit tion and uncertainty undermine investment. on special terms softened the budget for state Juxtaposing discipline and encouragement enterprises. Reliance on mechanisms inher- with protection and discouragement highlights ited from the command cconomy continucd. two contrasting modes of adjustment. In reality, The survival of a highly centralized structure country outcomes span a range of intermediate of political power did limit the extent of as- possibilities depending on whether liberalization, set stripping and other forms of theft that hard budget constraints, and an enabling busi- proved so damaging to growth in the previ- ness environment were pursued-and in what ous groups of countries. Yet this element of order and how vigorously. Though it is not pos- discipline came at the cost of a highly pro- sible to categorize transition economies into a tective stance that discouraged entry of new spectrum of different modes of adjustment, coun- enterprises and SMEs. try examples can be used to exemplify alterna- * For an example of partial liberalization and tive transition paths: weak discipline of state enterprises (though with strong restrictions on asset stripping) * The policies of discipline and encouragement coupled with an environment strongly support- were pursued most consistently in Estonia, ive of new entry, one must look outside East- Hungary, and Poland. ern Europe and the CIS to China (box 4.1). * Within the broad category of discipline and encouragement, softer budget constraints- Why particular countries or groups of coun- and hence less discipline-prevailed for a long tries get on paths of either discipline and encour- time in the Czech Republic, Lithuania, and agement or protection and discouragement can the Slovak Republic. Indeed, the resulting be traced to economic and political conditions lack of industrial restructuring is widely held they faced at the onset of transition; the struc- to have precipitated the Czech crisis in 1996- ture of political institutions that determined the 98. Harder budget constraints and faster re- relative power of winners and losers from re- structuring can help reorient these economies forms; and initial choices regarding the pace, toward the path followed by the first group. comprehensiveness, and sequencing of reforms. * Bulgaria, the Kyrgyz Republic, Moldova, Ro- Understanding the forces that shape the environ- mania, the Russian Federation, and Ukraine ment helps define the key challenge in the politi- liberalized their economies, but for a long time cal economy of discipline and encouragement; failed to maintain discipline through hard fostering new coalitions of winners and losers is budget constraints-and they could not con- crucial for shifting the incentives of old enter- tain tunneling and theft through either law or prises and promoting the development of new administrative control. The competition for enterprises and thus for improving the probabili- resources between old and new enterprises ties of reform. makes it difficult to provide encouragement if Meeting the challenges of discipline and en- the discipline for old enterprises is relaxed. couragement is important if growth is to be re- Russia and Ukraine encouraged new entry stored, its quality enhanced, and its benefits 34 Discipline and Encouragement Box 4.1. Can the CIS Leam from China's Refonn Experence? From 1978 to 1995 GDP per capita in China grew at 8 percent a year and lifted 200 million people out of absolute poverty. Does this experience offer lessons for the countries of the CIS-particularly for such countries as Uzbekistan and Belarus, which have yet to embark substantially on economic reform? Or are conditions so different between China and these countries that China's experience does not offer lessons for these countries? Since the onsetof reforms in 1978, China has been in the midst of two historic transitions: first, from a rural agricultural society to an urban, industrial one and, second, from a command economy to a market-based one. The first transition was driven by reforms in agriculture, a sector that employed 71 percent of the labor force and started a virtuous cycle, not only in agricultural development but also in rural industry. The substantial increase in agricultural productivity fed back into the economy in a number of ways. First, the increases in productivity freed surplus labor that had been hidden in the commune system to move into rural industry. Second, higher incomes from higher agricultural prices and production provided a market for goods and services produced by rural industry, with product quality upgraded as incomes rose over time. Third, the high savings rate-more than 30 percent of GDP-together with an implicit public guarantee of savings in the banking system, led Chinese households to hold deposits with the banking system. The ratio of M2 to GDP rose from 25 percent In 1978 to 89 percent in 1994. This allowed the banking system to channel netflows of 5 percent of GDP to borrowers. Some of the flows covered state enterprise losses and helped finance high-productivity investment in new and nonstate enterprises. Seigniorage accruing to the government in the 1990s allowed the budget deficit and substantial needs of loss-making enterprises to befinanced through money creation with low inflation, leaving financial workouts and loan recovery to a later phase of reform. Fourth, township and village enterprises developed rapidly and achieved high productivity growth and, despite unclear property rights, functioned as private enterprises in almost every way. In coastal areas growth depended less on high domestic savings, as the opening of markets was to lead to a massive inflow of foreign direct investment. The explosive growth of enterprises operating under the banner of townships or villages in rural areas was a source of intemal compettion, while extemal sector liberalization in coastal areas was a source of external competition for state enterprises. In general, no other sector in the CIS countries could provide a comparable boost to the economy. In Russia, for example, only 13 percent of the labor force was engaged in agriculture in 1990, compared with 71 percent in China. In principle, the energy sector in Russia, which gained from partial price liberalization on the order of 11 percent of GOP, could have been used to compensate the losers from reform. However, the state's loss of control rights over the sector, which had already occurred during the years of reform socialism, implied that much of the gains went abroad through capital flight, reflecting, among other things, a poor investment climate in the country. This is reflected in the difference in the ratio of fixed investment to GDP, which was 22 percent in Russia compared with around 34 percent in China, a large part of which-perhaps 10 percent of GDP-could be attributed to capital flight and foreign direct investment. Nor were surpluses available from the household sector. Increasing shortages, caused by the erosion of political control before the breakup of the former Soviet Union and before price liberalization, had led to substantial involuntary savings, resulting in a 'monetary overhang"-mostly household claims on the state banking system, estimated at roughly a third of household wealth for the former Soviet Union in 1990. Attempts to sterilize this overhang were unsuccessful, and the savings were extinguished by the burst of inflation following price liberalization at the onset of transition. Not surprising, financial savings were slow to recoverfrom this episode that, together with currency confiscation, led to substitution away from the domestic currency to commodities and foreign exchange. In marked contrast to China, the ratio of M2 to GDP in Russia fell from 68 percent to 17 percent in 1992. To place the second transition-that from a command to a market economy-in comparative perspective, note that only 19 percent of the Chinese laborforce worked in the state sector (and hence were enttled to a range of social benefits), compared with 90 percen: in, for example, Russia in 1990. In general, sectors enjoying soft budget constraints were much more important in Russia. In 1985,93 percent of the labor force was employed in state and municipal enterprises and organizations, including state farms, and a further 6 percent worked in collective farms and consumer cooperatives, leaving only 1 percent in individual and private enterprises. Because of the collapse of demand for industries producing capital goods and defense-related outputs and the difficulty of switching away from trade with the Council of Mutual Economic Assistance to Westem markets, a much largershare of the Russian economy was not viable following price and trade liberalization. The costs of propping up ratherthan restructuring the state sector would have been prohibitive. Thus, a strategy of protecting the state sector through fiscal and financial transfers from the rest of the economy, unlike in China, would not have allowed adequate economic space for new enterprises to emerge as sources of growth. (box contnues on following page) 35 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Former Soviet Union Box 4.1 cownNuED Finally, the govemment's ability to manage the process was critical. China contained the abuses from partial reform to some extent by exercising tight political control over asset stripping, arbitraging between controlled and market prices for private gain, and corruption. In contrast, the exit from communism, particularly in the countries of the CIS, led to a collapse of state institutions and, in the absence of a framework of property rights, set the stage for widespread tunneling and theft of state assets. In summary, some aspects of China's experience are relevant and parallel the experience of the most successful transition economies of Central Europe and the Baltics. These include, for example, the strong role played by new and nonstate enterpfises in generating growth and the positive role of increasing trade with market economies and foreign airect investment. Yet the large differences in initial conditions between China and the CIS point to the difficulties CIS countries face in trying to follow a similar path. In China, the first transition brought large gains (liberalizing repressed sectors such as agricuiture and nonstate industry). Part of these gains was available for transfer to the losers from the second transition from a command to a market economy. Equally important, those losers-those from the unviable sectors-did not account for a big part of the economy. The state's capacity to manage public assets enabled a slower move to market conditions for loss-making state enterprises, before they were subjected to full market discipline in a growing economy. If a country is able to follow such a phased strategy, there is less reason for it to experience a contractionary transition. However, these conditions were largely absent in the CIS countries where, following price liberalization, loss-making state enterprises accounted for a higher share of the economy and needed to be subjected to hard budget constraints for resources to be liberated and used by the winners (new private enterprises). This led first to a cut in output and activity, then a recovery. In addition to these differences in political and economic conditions, central planning was far more entrenched in Russia than in China. In the 1970s central government agencies in the former Soviet Union physically allocated about 60,000 different commodities throughout the plan. In China the number was about 600, unchanged from 1965. Nor, unlike the former Soviet Union, did China have the problems posed by 'giantism' and enormous regional specialization underpinned by high transport intensity and fuel prices far below world levels. Moreover, the costs of China's approach to transition, arising from soft budget constraints, remain to be fully addressed. The banking system, which was used to support loss-making state enterprises, is saddled with nonperforming loans of 30-40 percent of annual GDP, Restoring tne nealin of the state banking system on account of those loans could lead the stock of domestic government debt to rise from about 20 percent to 75 percent of GDP. Servicing this debt will pose a major fiscal challenge for the government. This mirrors the expefience of transition economies such as Bulgaria and Romania, where support for state enterprises resulted in a high proportion of nonperforming loans in the banking sector. Moreover, as in several transition economies, surveys report that asset stripping and excessive wage competition in China's state enterprises are now widespread. So, despite the different conditions in the CIS and China, what are the lessons for such countries as Belarus and Uzbekistan? In 2000, GDP as a proportion of its 1990 level was 85 percent in Belarus and 94 percent in Uzbekistan, significantly higher than the corresponding numbers in other CIS countries. The initial fall in GDP in Uzbekistan during the transitional recession was more like the shallow dip in Central and Eastern Europe, which may be attributed to low rates of industrialization and urbanization that, together with self-sufficiency in energy, made these countries less vulnerable to the disruption of payments and market links attending the breakup of the former Soviet Union. The experience of Belarus and Uzbekistan echoes that of China in two ways. First, governments of Belarus and Uzbekistan exercised enough political control over state enterprises to limit the risk of spontaneous privatization and excessive asset stripping. Such control has not always been effective; thus, Uzbekistan has seen sizable capital flight amounting to nearly 20 percent of merchandise exports. Second, governments were able to channel an infusion of resources to priority state enterprises and sectors, such as large-scale chemicals and automobiles in Uzbekistan and agriculture and housing in Belarus, to maintain production. In Uzbekistan this came from redirecting cotton and gold exports, which accounted for 60 percent of export revenues, to other markets at significantly better prices, as well as from newly opened oil deposits that turned the country into a net oil exporter in 1996. The effect of policies that discriminated against agriculture was a transfer of an estimated 4-5 percent of GDP out of the sector during 1996-98. In addition, state banks, used to transfer resources to favored sectors and enterprises, face the prospect of large losses. In Belarus the infusion of resources came from goods-for-energy barter deals with Russia on terms that were highly favorable, both to Belarusian exports and imports (see box 4.3). But China differs from Belarus and Uzbekistan in one critical way. Little has been done in Belarus and Uzbekistan to create an investment climate conducive to entry of new enterprises. In Belarus, for example, the share of employment accounted for by small enterprises (employing 50 or fewer people) is less than 20 percent, well below the 40 percent threshold needed for a return to sustained growth. It is in this respect, rather than in the infusion of resources to state enterprises, that countries such as Belarus and Uzbekistan would be wise to learn from China's experience and strongly encourage the growth of new enterprises as a basis for wealth creation and economic growth. 36 Discipline and Encouragement widely shared. The challenges are also interre- seeking and widespread theft of erstwhile state lated. Entrepreneurs have no incentive to bring assets dominate economic activity. The state can- in new management, develop new products, or not afford the social safety net required for re- seek new markets needed for growth-oriented structuring or investments in human capital if, restructuring if explicit subsidies or implicit sup- together with enterprises, it is caught in a perva- port extended through banks and public utili- sive web of arrears and unpaid taxes. ties and protection from competitive pressures The reform agenda required to enable growth keep them afloat. Evidence from business sur- to resume calls for effective public and private veys confirms that enterprises facing hard bud- institutions. Political institutions are critical in get constraints and a degree of competition are mediating the interactions between winners and more likely to undergo managerial turnover and losers from various reforms. The ability of the state develop new products (Carlin and others 2001). to secure property rights depends on the effec- Industrial enterprises also face major ob- tiveness of the legal system. Mechanisms of cor- stacles to downsizing if arrangements to help porate governance dictate what kind of owner- them divest social assets, such as housing, sana- ship structure is most likely to bring about toriums, and kindergartens, are unavailable. This efficiency-enhancing enterprise restructuring. Im- problem is further compounded by the lack of proved budget management is necessary to ean- public resources to finance an adequate system sure that the state maintains a responsible fiscal of social protection to replace these divested as- policy. Effective administration of programs of sets for well-targeted groups. Startups and social assistance and insurance are required so that spinoffs that could fuel innovation and growth the "reform dividend" generated by growth is used do not enter the fray if the tax and regulatory to assist those adversely affected by transition. system does not level the playing field between But policy choices are critically important as them and incumbent state-owned and privatized well; the example of East Germany demonstrates but unrestructured enterprises. If the state can- how poor policy choices at the start of transi- not administer the rule of law, particularly con- tion can undermine even the most favorable ein- tract enforcement and secure property rights, rent vironment (box 4.2). Box 4.2. The German Experience East Germany opened its border in 1989 and was united with West Germany in the fall of 1990. In many ways the experience of the transition in East Germany has been unique amongthe former socialist countries. It appeared to start from a privileged position in at leastAthree respects. First, all economic, political, and social institutions and arrangements of West Germany, as well as its entire legal framework, were adopted without change and without delay. East Germany did not face the difficult challenge of building a new institutional framework frcm scratch. Given that it is now conventional wisdom to argue that the key deficiency of the transition in many countries has been insufficient attention to building an institutional framework to support a market economy, East Germany appears to have had considerable advantages in its transition. Second, though critics faulted Western governments early in the transition for not offering enough aid or assistance to the transition economies, East Germany received huge transfer payments financed partially by a new "solidaritytax" leveled on West German incomes. Annual transfers from West to East Germany have been massive, averaging 40-60 percent of East Germany's GDP, totaling about DM 1.2 trillion (US$571 billion) forthe 1991-97 period. Currenttransfersstill average DM 140 billion, or more than 4percentof WestGermany's GDP, a year. In comparison, total lending and grants of the World Bank to all transition economies in Europe and Central Asia amounted to DM 43 billion (US$19 billion) in 1991-2000. The transfers to East Germany have financed public expenditure programs, supported private sector investment, and secured transfer payments for social protection that have arisen from the sudden eligibility of East German citizens, including massive retraining and public works programs. (box continues on following page) 37 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Former Soviet Union Box 4.2 counNUED Third, unification gave East Germany automatic membership in the European Union (and the World Trade Organization). While the CIS suffered from European antidumping rules and Central Europe's Central Europe Free Trade Agreement members complained of being blocked from the Common Market (while being subjected to cheap imports, especially in the agricultural sector), East Germany benefited from speedy and complete incorporation into a large external market In light of these considerable advantages, few would have predicted the difficulties East Germany encountered over the past decade. East Germany's growth performance has lagged behind other transition economies with far fewer advantages (see box figure A). Its initial economic decline was deeper than in its transition economy neighbors. Though East Germany's growth rates were higher than those in other countries in the region for a few years, its subsequent growth has been among the slowest in Europe. The hope that East Germany would quickly catch up with West Germany has not materialized. FiruRE A. Companison of Real GDP, 1989-2000 Index of real GDP (1988 = 100) 150 - 125 - - Average for Central and Southeastem Europe and the Baltics 100 ' Germany ....... East Germany 75 ...... -Poland 50 - , , , , t , 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Note: The average for the CSB is weighted by population. Source: World Bank country office data. How can such poor performance be explained despite the enormous advantages that East Germany enjoyed in comparison with other transition economies? A simple answer is that even a good institutional framework backed by massive subsidies cannot overcome the negative consequences of inappropriate policies. Two early policy choices had particularly damaging consequences. First, the conversion rate of West to East German marks reflected political pressures and not economic realities, vastly overvaluing the East German currency. Second, the attempt to bring East German wages in line with West German wages, while average labor productivity in the East remained at approximately a third of the West, severely damaged East Germany's competitiveness. As a result, a larger segment of the inherited capital stock was scrapped as unproductive and less employment was preserved than would have been warranted with more realistic macroeconomic policy choices in effect. Erasing the existing capital stock, together with extending West Germany's generous eligibility criteria, increased the need for social transfer payments. Such payments continue to constitute between 55 percent and 65 percent of gross national product Clearly, the current level of social payments in East Germany would not be sustainable in any economically independent geographic area. The high degree of subsidization in East Germany, coupled with the relatively high unit labor costs, have coincided with a marked slowdown in the rate of new enterprise growth, as shown in box figure B. The net creation of new enterprises, which had outperformed West Germany in 1991, has persistently declined each year through 1999, falling to a seventh of West Germany's level. Now East Germany seems to suffer from exactly the same phenomenon as the least successful transition economies in the CIS-anemic growth of new enterprises that have been the main drivers of growth elsewhere in the transition. Indeed, investment data suggests that even German investors are 'jumping over" East Germany to invest in the Czech Republic, Hungary, and Poland, where unit wage costs are considerably lower. (box continues on following page) 38 Discipline and Encouragement Box 4.2 CONTINUED Though the slow rate of new enterprise growth is mostly attributed to an insufficient institutional framework in the CIS, the East German case demonstrates how poor policy choices at the start of transition can undermine even the most favorable environment. FIGURE S. Net New Enterprise Registration in East and West Germany In 1000 300 250 - 150 ;- , West Germany 100 -- - East Germany 50 _ 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Source: Ruehl and Vinogradov (2001). New Enterprises Drive the Transition T he success of a discipline-and-encourage 1998 was around 55-65 percent of GDP in the strategy is predicated on the ability of new Czech Republic, Hungary, and Lithuania, corn- and restructured enterprises to emerge as engines pared with 10-20 percent in Belarus, Kazakhstan, of economic growth. Moving assets from the Russia, and Ukraine.' Data on small enterprises public to the private sector is thus an important as providers of employment divide countries into element of transition. In 1999 the share of the two groups: leading reformers and countries fur- private sector in CIS GDP was 20 percent in ther behind. Small enterprises' share of employ- Belarus; 55 percent in Kazakhstan and Ukraine; ment in 1998 was about 50 percent for leading 60 percent in Armenia, Georgia, and the Kyrgyz reformers such as the Czech Republic, Hungary, Republic; and 70 percent in the Russian Federa- Latvia, Lithuania, and Poland, roughly the same tion (figure 4.1). These figures compare favor- as the European Union. For countries less far along ably with Central and Eastern Europe: 55 per- the path to a market economy, such as Belarus, cent in Macedonia and Slovenia, 60 percent in Kazakhstan, Russia, and Ukraine, the share was Bulgaria and Romania, 65 percent in Latvia and between 10 and 20 percent. Poland, and 80 percent in the Czech Republic The share of small enterprises in employmeint and Hungary. and value added differs widely across the region The picture is quite different, however, for new (see figures 4.2 and 4.3). In many parts of the enterprises, which typically need encouragement CIS and Southeastern Europe-where growth in the form of a favorable business environment is low and income per capita is substantially be- (figures 4.2 and 4.3). Using small enterprises em- low pretransition levels-the share of employ- ploying fewer than 50 workers as a proxy for new ment in small enterprises hovers around 20 per- enterprises, their contribution to value added in cent. In contrast, the share of small enterprises 39 Transition-The First Ten Years: Analysis and Lessons for Eastem Europe and the Former Soviet Union consistently account for a larger fraction of to- FIGURE 4.1. tal value added than of total employment. La- Private Secor Share in GDP, 1999 bor productivity is indeed higher in small enter- prises compared with large enterprises. Belarus Moreover, this is true independent of progress Turkmenistan toward a market economy. New companies are Tajikistan more productive than inherited enterprises in Azerbaijan countries as diverse as Hungary and Ukraine. Moldova Second, is aggregate labor productivity, Uzbekistan measured as aggregate value added per em- Kazakstan ployee, higher in an economy where small en- Macedonia, FYR terprises account for a higher proportion of Slovenia value added and employment than in another Ukraine where those shares are lower? Small enterprises Armenia have had high and growing shares of employ- Bulgaria ment and value added in the leading reformers Croatia in the CSB (figures 4.4 and 4.5). However, there Georgia appears to be a threshold-of around 40 per- Kyrgyz Republic cent for the shares of small enterprises in em- Romania ployment and value added-below which Latvia _ economies do not take off in terms of growth. Poland This echoes the finding that there must be a Lithuaniia Russian Federation _ minimum critical mass of reforms, below which Albania the economy does not respond to policies. The Estonia shares of new enterprises in employment and Slovak Republic value added are high and above the threshold Czech Republic in the leading reformers in the CSB. But both Hungary shares remain low and well below the thresh- a 20 40 60 80 100 old in the slow-growing economies of the CIS. The notion of a threshold is important in the Percent interaction between the old and new sectors of the economy. Source: EBRD (1999). Third, can differences in labor productivity between small and large enterprises constitute a source of growth? The higher productivity of in total employment and value added is more small enterprises needs to be complemented by than 50 percent in the CSB, characterized by an incentive for labor and capital to move to that high and sustained economic growth and either sector-for small enterprises to increase aggre- approaching or surpassing pretransition per gate growth. Higher labor productivity in small capita incomes. enterprises implies lower labor intensity per unit The rapid growth of small enterprises in help- of output. This observation and the assumption ing bring about a higher contribution to value that small enterprises are less capital-intensive added and employment among the leading re- than large enterprises imply that labor and capi- formers shows that economic policy has directed tal have a higher marginal product in small en- the process of factor reallocation. That prompts terprises (annex 4.2). Movements from large to three questions (table 4.1). small enterprises are adding value and thus are a First, is labor productivity in small enter- source of growth. The large gap in labor pro- prises, measured as value added per employee, ductivity in countries such as Kazakhstan and higher than in large enterprises in transition Ukraine shows an unrealized potential for economies? In the aggregate, small enterprises growth in the new sector. 40 Discipline and Encouragement FIGURE 4.2. Share of Employment in Small Enterprises, 1989-98 Percent Hungary 60 - Czech Republic - - H *Poland 40 -P ...... ..Lithuania ...... Latvia 20 - Russia _ = -- Ukraine O - Belarus 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 Kazakhstan Source: World Bank database on SMEs. FIGURE 4.3. Share of Value Added in Small Enterprises, 1989-98 Percent Hungary 80 - Czech Republic 60 - Poland . ......... Lithuania 40 .Latvia Russia 20 ~ ~ ~ ~ ~~~~~. - - ~~~~~~~~~~~~Ukraine 0 I l l l l l l I _ Georgia 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 Kazakhstan Source: World Bank database on SMEs. The difference between the two sets of enter- A Small Number of High-Productivity Small prises-old and new-begins to erode over time Enterprises Is Not Enough for two reasons. Old enterprises either close or restructure, raising labor productivity, and em- s the transition proceeds, labor shed by ployment growth in new enterprises at some point A downsizing old enterprises either finds its reduces labor productivity in those enterprises. way into new and more productive employment There is a marked productivity difference between or migrates to unemployment or subsistence ac- the two sets of enterprises for Russia and Ukraine, tivities. Labor hoarding in the old sector may but a noticeable narrowing of that difference for persist for a long time, with new enterprises act- the Czech Republic, Hungary, and Lithuania far- ing merely as passive receptacles for such trans- ther along the path to a market economy. fers. Alternatively, when the investment climate 41 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Former Soviet Union TABLE 4.1. SMEs Have Higher Labor Productivity, 1998 (percent) Share of total Value added Value added Countries SMEs employment (total) (per employee) Belarus 37.7 15.9 - - Czech Republic 97.0 48.7 53.5 109.9 Georgia 88.6 39.6 39.3 99.2 Hungary 96.1 54.9 63.6 115.8 Kazakhstan 88.6 15.6 22.4 143.6 Latvia 91.2 45.5 50.4 110.9 Lithuania 97.4 55.1 55.3 100.4 Poland 92.1 45.7 54.4 118.9 Russia 56.3 18.6 23.0 123.7 Ukraine 69.2 16.9 30.0 177.5 - Not available. Note: Average for enterprises of all sizes = 100. Small enterprises are defined as having 50 or fewer employees. Source: World Bank database on SMEs. the old sector and contribute to sustainable FIGURE 4.4. growth. Simply having a small number of highly Value Added per Employee in Small Enterprises, productive small enterprises is not enough. Un- 1998 less it is combined with rapid growth in the share of employment, the small sector will not develop Czech Republic the critical mass to lead aggregate economic Georgia growth (see figures 4.4 and 4.5). Hungary Russia illustrates the inability of the new sec- Kazakstan tor to contribute to sustained growth. The share Russia of value added for small enterprises more than Ukraine doubled, from 10 percent to 23 percent between European Union 1991 and 1998, in part reflecting a steadily de- 0 10 20 30 40 clining GDP. But their share in employment has remained low and largely stagnant. They do not Thousand euros seem to have incentives to increase their employ- Source: Eurostat (1998). ment or to multiply-so they continue to fall well short of the threshold for growth to take off. A large chunk of the labor force remains mired in unrestructured state and private enterprises. is conducive to entry, new enterprises compete In countries where aggregate employment with the old sector, rapidly increasing their share picked up, it did so after the recovery of aggre- in employment and typically attracting the most gate output (figure 4.6). This empirical investi- qualified individuals. gation of new enterprises and their interaction As noted, a threshold of about 40 percent with old enterprises suggests the following: for the shares of small enterprises in employment and value added needs to be crossed for new * A sharp and early decline in aggregate em- enterprises to absorb the resources released by ployment precedes the rapid growth of new 42 Discipline and Encouragement Figure 4.5. Index of GDP and Shares of Value Added and Employment Accounted for by Small Enterprises, 1989-98 Czech Republic Russia Index Percent Index Percent 120 - , 60 120 - _ 60 100 - > 50 100 _ - 50 80 - 40 80 - 40 60 - - 30 60 - 30 40 - 20 40 - 2 20 - 10 20 .- 10 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 Lithuania Ukraine Index Percent Index Percent 120 - 60 120 - 60 100 _ 50 100 _ 50 80 - 40 80 - - 40 60 - 30 60 30 40 - 20 40 - 20 20 - 10 20 - 10 0 I I I I I 0 0 L I I LfI~ 0 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 Hungary Kazakhstan Index Percent Index Percent 120 - 70 120 _ _ 60 100 _ 60 100 _ - 50 80 - 50 80 - - 40 40 60 - 60 - - 30 30 40 - 20 40 - 20 o 210 2 10 0 I I I I I0 0 I I I oL 0 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 - GDP index Percentage of employees in small firms - Value added by small enterprise as percentage of total value added Source: World Bank database on SMEs. 43 Transtion-The Frst Ten Years: Analysis and Lessons for Eastern Europe and the Former Soviet Union FIGURE 4.6. Employment and GDP, 1990-98 (index 1990 = 100) Hungary Czech Republic Index Index 120 - 120 100 100 80 80 60 - 60 40 40 - 1990 1991 1992 1993 1994 1995 1996 1997 1998 1990 1991 1992 1993 1994 1995 1996 1997 1998 Poland Ukraine Index Index 140 - 120 120 - 100 100 - ___ _ 80 80 - 60 60 40 40 20 1990 1991 1992 1993 1994 1995 1996 1997 1998 1990 1991 1992 1993 1994 1995 1996 1997 1998 Russian Federation Kazakhstan Index Index 120 120 100 100 80 80 60 -60 40 40 l l l 1990 1991 1992 1993 1994 1995 1996 1997 1998 1990 1991 1992 1993 1994 1995 1996 1997 1998 Bulgaria Romania Index Index 120 120 100 100 80 80 60 - 60 - 40 40- 1990 1991 1992 1993 1994 1995 1996 1997 1998 1990 1991 1992 1993 1994 1995 1996 1997 1998 - GDP index Employment Source: World Bank database on SMEs. 44 Discipline and Encouragement enterprises. With the former as an indica- resource transfers for nonviable state-owned en- tor of the fast dismantling of old industries, terprises, other things being equal, either require the rapid demise of the old sector seems nec- the consolidated government to loosen its fiscal essary but not sufficient for the growth of stance, raise taxes elsewhere in the economy, or new enterprises. A plausible reason would engage in off-budget activity. be that the rapid dismantling of the old sec- tor lowers the price of its assets. Cheap re- * A looser fiscal stance can weaken the cred- sources become easily available to new en- ibility of the government's stabilization pro- terprises, useful when financing is not gram and, by increasing market perception available and investment not forthcoming. of risk, raise domestic interest rates, thus That is why discipline is a crucial element crowding out the new private sector.2 of growth. But it is not enough, for encour- * Intensifying taxation of the potentially more agement is also required. efficient emerging private sector can start: a * Countries that reached the trough of the tran- vicious cycle that pushes enterprises into the sition recession sooner had faster growth in informal sector, thus lowering tax revenue the new sector. Where sustainable growth and further increasing tax rates. returned, the share of small enterprises had * Off-budget activities include issuing loan reached a critical mass. Where it did not, guarantees, establishing insurance schemes, people remained "unemployed on the job," and providing explicit or implicit cover gen- as in the CIS and countries in Southeastern erally for politically motivated activities tlhat Europe. Aggregate employment started to fall tend to benefit incumbents. only late in the process. These observations suggest a sequence where hard budget con- These outcomes end up protecting state en- straints are imposed and the old sector de- terprises and collective farms and discouraging clines before the new sector can grow. new enterprises. However, the relationship does * While disciplining the old and encouraging not run simply from the costs of propping up the new appear to be complementary, the old unviable enterprises (protection) to discourag- sector has generally, though not invariably, ing the new private sector. The lack of a vibrant proved unable to survive even where budget private sector limits the outside options that constraints have been soft and the new sec- might attract workers and potential entrepre- tor has not emerged. In such cases, mainly in neurs from old enterprises. It also limits prod- the CIS, agriculture and low productivity ser- uct competition, an essential element of a disci- vices have served as "shock absorbers" for plining environment. Both weaken the incentives those forced to leave old industries. for closure and restructuring of nonviable en- terprises. These considerations explain why im- Can We Have It Both Ways, That Is, pediments to the efficient exit of enterprises dis- courage the growth of new enterprises, and why Encouraging New Enterp rises? the emergence of a new private sector would accelerate the closure and restructuring of state- T he analysis makes clear that policies encour- owned enterprises. T aging the new private sector while subject- Further evidence on how protection of the ing them to market discipline must get the old sector discourages the new sector conies highest priority. But must discipline and encour- from the following examples. The first is pro- agement go together? Or is it possible to encour- tection through the banking sector. Small en- age the new sector while protecting the state en- terprises have grown less in such CSB countries terprises and privatized enterprises that continue as Bulgaria and Romania, where the banking to behave as if they were public? The economic sector was a major conduit for loans to the old argument seeing discipline and encouragement state-owned enterprises and farm collectives. In as complementary goes as follows. Continued those countries nonperforming loans increased 45 Transition-The irst Ten Years: Analysis and Lessons for Eastern Europe and the Fonmer Soviet Union as a share of total banking sector loans during through favorable foreign exchange regimes, much of the 1990s. Contrast that with the vi- directed credit, and high trade protection. brant growth of new enterprises in the Baltics, Specific large foreign investments have also been Hungary, and Poland, where these loans were favored. As a result new smaller enterprises face sharply reduced over time. In 1998 nonper- the residual of these allocations in the market. forming loans still represented 34 percent of What remains in credit and foreign exchange must total loans in Romania, while they were 4 per- be purchased at costs several times higher than cent in Estonia, 6 percent in Hungary, 11 per- would have been paid in unified markets. In cent in Poland, and 13 percent in Lithuania. Uzbekistan small enterprises have to pay three These loans to old enterprises prevented the ex- times more for foreign exchange than do large pansion of bank credit to new, small, and less state enterprises to finance their imports. As a re- politically connected enterprises. In addition, sult new enterprises have been squelched. In they became a major factor in triggering a bank- Belarus, the share of small enterprises in the total ing and macroeconomic crisis that called for a number of enterprises is a mere 26 percent, sub- sharp stabilization and credit tightening-all stantially lower than other CIS countries. While reducing the growth of new enterprises. these special regimes may have helped avoid a The second example is protection through spe- sharp output decline in the old industrial sector cial allocation of inputs. In Belarus and Uzbekistan and aggregate output, they cannot provide the the old industrial sector has remained protected basis for sustained future growth (box 4.3). Box 4.3. Belarus A decade after the beginning of transition, Belarus has emerged as one of the least reformed, but seemingly most resilient, economies in the CIS. GDP in 1999 reached 89 percent of its 1990 level, well above the CIS average of 62 percent (box table A). The country's relatively favorable growth performance stands in stark contrast to its extremely poor record on macroeconomic slabilizatior, and structural reform. Apart from a brief spell of relative stability in 1996 and 1997, inflation ran at triple digits throughout the 1990s, reflecting money-based financing of public sector deficits. The European Bank for Reconstruction and Development ranks Belarus 25th out of 26 transition economies in overall progress in economic reform. Although Belarus has largely abolished the Soviet command system, a valiety of controls over the economy remain. The state has continued to support priority sectors, such as agriculture and housing, with both direct subsidies and soft loans channeled through the banking system. Price controls remain widespread, partly aimed at regulating the prices of certain commodities and partly designed to fight inflation through administrative means. A multiple exchange rate system remained in place until September 2000, operating as a heavy tax on exporters and a subsidy for certain importers. While intemational trade has been liberalized, some restrictions remain in conjunction with domestic price controls. Privatization of large enterprises has hardly begun, and even many small enterprises remain in state hands. Given the poor record on macroeconomic stabilization and structural reform, what explains Belarus's relative success in having contained output decline? There are four important considerations. First, Belarus is successful only in relation to the other CIS countries. Compared with transition economies outside the CIS, such as neighbors Lithuania and Poland, Belarus is no more than an average performer. Second, Belarus has had the benefit of continued cheap energy imports from Russia. In addition to this direct subsidy from Russia, Belarus also tended to accumulate substantial energy arrears to Russia, which were later settled through barter transactions that artificially maintained demand for Belarusian exports-another important form of indirect subsidy from Russia. Third, by providing subsidies and credit to large enterprises, the state kept industrial output from collapsing in the early stages of transition, as in many other CIS countries. Fourth, by not privatizing large enterprises, the state has retained greater control over productive assets and prevented extreme cases of asset stripping and tunneling. These considerations suggest that continued state ownership and control of large enterprises, as well as preferential treatment by Russia, have played a key role in maintaining industrial production and bolstering economic growth. Yet the resources that have been used to maintain industrial production have been drained from other sectors of the economy. In particular, the Belarusian system has stifled the growth of new (box continues on following page) 46 Discipline and Encouragement Box 4.3 comuNUED enterprises. The multiple exchange rate system and pervasive state controls have shifted private initiative into arbitrage activites and the shadow economy, while investment has remained largely in the state sector. There has been little evidence of restructuring in the state sector, suggesting that the pain of such restructuring has merely been postponed. These factors do not bode well for long-term growth in Belarus. TABLE A. How Belarus Compares with Other Transfiton Economies Real GDP in 2000 Average inflation,' EBRD transition Government revenue County (199= 100) 1991-2000 indicator rating, 1999b to GDP, 1999' CSB average 106 41 3.1 39 Czech Republic 99 13 3,5 41 Hungary 109 20 3.7 42 Lithuania 68 88 3.2 32 Poland 147 26 3.5 43 Romania 83 102 2.8 32 CIS average 62 18S 2.3 24 Belarus 89 344 1.6 42 Georgia 30 257 2.9 16 Kazakhstan 65 163 2.7 19 Russian Federation 66 163 2.5 34 Ukraine 43 244 2.5 35 Uzbekistan 97 182 2.0 32 a. Consumer Price Index. b. Unweighted average of transition indicator ratings across reform categories, which range from 1 (least progress) to 4+ (most progress). c. General government revenues divided by nominal GDP. Source: EBRD (2000). Nevertheless, Belarus's experience suggests important lessons about transition. When compared with the partial and poorly implemented reforms in many CIS countries, Belarus's inaction enabled it to avoid some key mistakes in the early stages of transition. First, while continued state ownership did little to promote more efficient operational or strategic decisionmaking at the enterprise level, it did deter the large-scale asset stripping, tunneling, and tax evasion that has damaged growth in the early stages of transition in some other CIS countries. Nevertheless, the rapid decentralization and even fragmentation of power in other CIS countries at the start of transition might have precluded such a policy option. Second, the capacity of the Belarusian state to maintain high levels of tax collection (see box table A) highlights the importance of these taxes in smoothing the initial output decline. The strength of fiscal revenues has kept the scope for supporting declining sectors and socially oriented expenditures. Belarus will need to maintain its ability to collect taxes if it chooses to embark on economic reform. Though it may be tempting to explain this fiscal performance as a result of the authoritarian state, such a political regime has not guaranteed similar outcomes in Turkmenistan or Uzbekistan. The third lesson is that current account revenues matter, especially in transition. Other CIS countries saw their exports to Russia collapse while their energy imports became far more expensive, leading to a massive terms-of-trade shock. While Central European transition economies were redirecting their exports to the European Union, CIS countries had nowhere to go but Russia. Belarus continued to earn export and service revenues by maintaining close economic ties with Russia, thus softening the initial trade shocks. In summary, although Belarus does not have a viable strategy for sustainable growth, it has-perhaps inadvertently-avoided some mis- takes. The ability of the state to prevent the collapse of its capacity and control early in transition appears to have saved the Belarusian economy from the rapid decline felt elsewhere in the region. Yet such an outcome was hardly guaranteed by the lack of economic refcrn, but rather by dependence on Russia, whose willingness to subsidize, directiy and indirectly, Belarus's inaction was a function of the geostrategic importance of the country. Such an option was not necessarily available to other CIS countries. A key question for the future is whether Belarus will enjoy certain 'advantages of backwardness' if it embarks on serious structural reforms. The long delay in privatization could allow Belarus to learn from earlier mistakes of other transition economies. Whether the capacity of state institutions can be simply shifted from maintaining the status quo to implementing forward-looking economic policies in an environment (box continues on following page) 47 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Former Soviet Union Box 4.3 coNnwum characterized by the rule of law is uncertain. Surely, if Belarus were to change course now and embark on economic reform, its initial conditions would be far more favorable than those of the other CIS countries in the early 1990s, when demand was collapsing across the old Council of Mutual Economic Assistance. Most of Belarus's trading partners are now growing and this should soften the negative shock of any reforrn effort. While these considerations suggest that Belarus may still have the opportunity to develop into a strong economic reformer-once the political will is there-an important question remains. Will economic reforms lead to the same erosion of public institutions observed in other transition economies? Once opportunities forprivate wealth creation in the private sector arise, skillful civil servants may have greater incentives to extract rents from the private sector or may even switch careers. The key challenge for Belarus, once it starts reforming, will ultimately be the same as for all other transition economies: to establish good govemance and strong market-supporting institutions. The third example is protection through tax Tax exemptions for large enterprises and ag- and utility arrears. Nonpayments by enterprises riculture collectives-popular until recently in to the main utilities in the energy sector remain a Ukraine, where these initiatives could be unilat- major source of subsidization, particularly in Rus- erally introduced by the legislature-generate a sia and Ukraine, but also in Georgia, the Kyrgyz highly discretionary tax regime and are an im- Republic, and Moldova. This has delayed restruc- portant source of bribes. The same is true for the turing of energy-intensive industries and shedding use of negotiated offsets as a way to pay taxes in of assets that smaller new enterprises could use. Russia, and tax avoidance by large enterprises in The discretion of negotiating and settling Georgia. New small enterprises are usually less nonpayments complements the natural behavior powerful in this environment and end up paying of these utilities to act as discriminating monopo- higher bribes as a fraction of total profits. lies. New, more energy-efficient enterprises are In sum: the softer the budget constraint and charged more to compensate for the revenue losses thus the stronger the barriers to exit, the lower from the old, less energy-efficient enterprises. the contribution of small enterprises to employ- ment (figure 4.7). FIGURE 4.7. Annex 4.1. Assumptions for Small and New Soft Budget Constraints and Employment in Enterpnses Small Enterprises, 2000 his discussion is based on the hypothesis that l small enterprises can be taken as approxi- Share of employment in small ente-rpses (percent) mating new enterprises. For this hypothesis, the Hungary Cech Republic margin of error for Central and Southeastern Eu- rope can be deduced from table A4. 1. Although it Estonia Poland seems safe to assume that most new enterprises 40 - \ * Georgia are small enterprises, the opposite need not be true.3 Table A4. 1 lists the share of new and small enter- 20 - prises over total active enterprises in Central Eu- Romania n U ropean countries for 1995. For example, of all Kazakhstan Ukraine active enterprises in Albania in 1995, 68 percent 0 l l (from column 2) were created as greenfield. For 2 4 6 the same year our assumption would estimate the proportion of small enterprises to be 98.7 percent Soft budget constraints index (from column 3). The difference of 30.3 percent Source: EBRD (2000); World Bank database on SMEs. (column 4) is the share of small enterprises that we consider to be new ones but actually are old. 48 Discipline and Encouragement TABLE A4.1. Differences between New Enterprises and Small Enterprises, 1995 and 1998 (percent) Enterprises active January 1995 Projections for 1998 Small old Small old enterprises as enterprises as New Small Small old percentage of percentage of Country enterprises enterprises enterprises small enterprises small enterprises Albania 68.4 98.7 30.3 30.7 18.8 Bulgaria 96.0 98.1 2.1 2.1 1.1 Czech Republic 86.1 98.6 12.5 12.7 7.6 Estonia 81.9 96.4 14.5 15.0 7.8 Hungary 84.5 99.0 14.5 14.6 9.6 Latvia 83.2 95.2 12.0 12.6 6.2 Lithuania 75.4 95.6 20.2 21.1 10.9 Poland 88.6 98.7 10.1 10.2 4.9 Romania 95.8 99.1 3.3 3.3 2.0 Slovak Republic 92.6 98.4 5.8 5.9 3.4 Slovenia 75.7 97.6 21.9 22.4 14.5 Central European countries 88.3 98.6 10.3 10.4 5.7 Source: Eurostat (1998); authors' calculations. Column 5 shows the ratio of small old enter- The error tends to decrease over time. Be- prises over small enterprises, that is, the probabil- cause the stock of companies in the hands of the ity of picking up a wrong element when sampling state at the beginning of the transition is either from the set of small enterprises. It is therefore a liquidated or privatized, the source of the error measure of the error implied in assuming that small is extinguished. The last column of table A4.1 enterprises equal new enterprises. The size of the shows a projection for 1998 once a mortality error for Albania is large: 30.3 percent means that, rate of 20 percent is assumed for both new and within the set of enterprises we assume to be new old enterprises. (Most of the data in this report ones, one out of three is not. But for Bulgaria the refer to 1998.) The probability of error is esti- error is 2 percent. Although Albania is extreme, mated to have shrunk considerably for all thie many countries show high errors. countries this paper discusses. Figures for the Czech Republic, Hungary, and Further evidence can be gathered through thie Lithuania are a source of concern because this answers to the Business Enterprise Environment study relies heavily on evidence from the three and Performance Survey (see box 3.1). The sur- countries. However, this crude estimate is an vey results show that 52 percent of new enter- upper bound for the actual error. In fact, neither prises are controlled by an individual owner (ver- small-scale spinoffs nor new enterprises created sus 24 percent of old enterprises), and that in as private cooperatives are counted in column 2, two-thirds of the new enterprises the majority although both can be considered new enterprises. ownership is held by no more than three share- The bias is particularly relevant for the coun- holders (true for less than half of old enterprises). tries in the Federal Republic of Yugoslavia, such Moreover, new enterprises are the ones that rely as Slovenia, where many new enterprises were most on internal funds and family financing and created as cooperatives. that have more trouble getting bank credit. The 49 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Former Soviet Union higher concentration in ownership and the ex- employee means that primary factors of produc- clusion from conventional credit channels is im- tion earn more in that sector. So, by moving into portant indirect evidence that new enterprises the sector, workers can get better wages, or en- are small. trepreneurs can expect a better return for their capital, or both. That is, primary factors in SMEs have more income to distribute among them. Productivi m a of SMEs Figure A4.1, based on the aggregate num- bers for Russia, illustrates the argument. The I n many transition economies SMEs have two lines show the combination of wages (or higher average value added per employee than marginal product of labor, MPL) and profit the rest of the economy. As a result: rates (or marginal product of capital, MPK) compatible with the observed average product * Factors of production are likely to have an of labor in each sector (normalized so that av- incentive to move to the SME sector, thus, erage value added per worker in the economy increasing its size as a fraction of the and total employment are both equal to 1). The economy. slope of each line reflects the average capital * An increase in the size of the SME sector is intensity of the sector.4 The intersection with likely to increase the growth rate of the the vertical axis is the average value added per economy. employee in each sector (as a ratio to the economywide value added per employee). Because SMEs represent the main source of Assume that the "large enterprise" sector is new enterprises, we refer to the SME sector as characterized by point A, where the wage rate is "new." The basic argument is that since SMEs about 60 percent of the average value added per are likely to be less capital intensive than larger worker in the economy. The production possi- or old enterprises, a larger value added per bilities indicated by the SME line would mean that the new sector can pay wages 50 percent higher at the same profit rate (indicated by a point directly above A), or a much larger profit FIGURE A4.1. rate if paying the same wages (points directly at Factor Price Frontier: SMEs and the Rest of the right of A, on the new line). So there would the Economy be strong incentives for both labor and capital to move to the SME sector. Wage rate (MPL) Labor mobility in Russia has probably not 1.5 been strong enough to equalize wages across sectors. So, on average, MPL for the SME sec- tor is likely to exceed MPL for the large enter- 1.0 prise sector at the current position-particularly B considering the effective wage resulting from A old enterprises not paying their workers on time 0.5 or paying in kind, rather than the official wage rate. Because the existing capital stock is likely 0.0 s \ E \ E to be more sector-specific than labor, its returns 0 10 20 30 could be even further from being equalized. Thus, a point like B appears as a reasonable Rate of profits (MPK, percent) comparison point on the SME sector for a "typi- cal" bundle of labor and capital. - SMEs - Others What are the gains to the economy when fac- Source: Authors. tors move, say, from A to B? A critical element is the extent to which factors, particularly capital, 50 Discipline and Encouragement can be productively reallocated to new uses. For that, we define a "coefficient of shiftability to the FIGURE A4.2. SME sector" equal to the ratio of productivity of Efficiency Gain from 10 Percent Factor capital from the old sector to the productivity of Reallocation to the SME Sector new capital.5 Some types of capital, such as spe- cialized machinery and tools, may not be usable Increase in GDP (percent) at all (a zero shiftability coefficient). Others, such 4 as liquid assets used as working capital, office equipment, and some real estate, are likely to be 3 fully usable in the emerging sector. In many old enterprises the main obstacle to productive real- 2 location of capital may be its technological ob- solescence-though this also means that it is of 1 low productivity even for the old enterprises. The figure shows the gain in GDP from real- l locating to the SME sector 10 percent of the total 0 0.5 1.0 labor and capital of the economy, as a function of the "shiftability coefficient" defined above. The Shiftability coefficient (O = null, 1 = comp ete) polar case of perfect capital shiftability-when - Only labor moves capital can move without any loss in productiv- O labor moves ity-is also of interest because it represents the Labor and capital move case of new investment. Figure A4.2 shows the Source: Authors. large efficiency gain in allocating those new re- sources to the emerging sector instead of using them on old enterprises; if 10 percent of labor and capital move to the SME sector, aggregate ter because they reflect restructuring, and that the valued added (GDP) would increase by almost 4 evolution of labor productivity in the manufactur- percent (assuming constant returns to scale, as ing sector can be taken to approximate the extent of standard on this type of model). The extra value restructuring (see annex 4.1). The corresponding added is the ex post validation of the ex ante po- number for Georgia-40 percent-reflects growth of new enterprises plus the destruction of old produc- tential earning difference, which provides the en- tive capacity due to armed conflict. couragement for the creation of new enterprises. The large productivity gap in such countries as 2. In Russia subsidies implicit in soft budgets ac- Kazakhstan and Ukraine may thus be taken as an counted for two-thirds of net borrowing by the gov- indicator of how much unrealized growth poten- ernment. As a share of GDP this was 7.5 percent in 1996 and 5.6 percent in 1997. With the government's tial exists in the new sector in these countries, disinflation policy, this kept real interest rates in the The other polar case is when only labor can triple digits in 1995 and 1996, coming down to single move to the new sector.6 In this case (showed by digits only in 1997. the horizontal line in the figure) the gain is sim- ply the difference in marginal productive of la- 3. Caves (1998) finds that most newly born enter- bors across the sectors-about 1.2 percent in the prises in market economies belong to the first tvwo deciles of the size distribution. A similar result can be "base" case marked by points A and B on the expected in transition economies, where starting large factor price frontier. enterprises from scratch is more difficult. 4. Thus the SME line's lower slope reflects the as- Notes sumption that SMEs are on average less capital inten- sive than large enterprises. For the average capital in- 1. The analysis is based on the hypotheses that small tensity of the economy we use a capital-output ratio enterprises can be taken as a proxy for new enter- of 6 (EBRD 2000). The SME line is constructed as- prises, that spinoffs from the state sector do not mat- suming that they used about 8 percent less capital per 51 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Former Soviet Union worker than the economy average (a capital-output (O for fixed factors; 1 for perfect mobility and sub- ratio of 5, versus 6.23 for the rest of the economy). If stitutability across sectors), the percentage gain in both sectors had lower overall capital intensity, the aggregate value added from reallocating 1 percent two lines would rotate out around the w-axis (for each of labor and capital to the new sector is dV/V = wage they would show a higher profit rate). W, - wo + k, ( a r, - rO). Of course, if the shiftability coefficient, a, is too low-specifically lower than 5. Thus (1-shiftability coefficient) is the share of a = r1/ro-it is not worthwhile to reallocate capital, productivity "lost" when moving from the old to the as the loss of output in the old sector would exceed new sector. The values on the figure are calculated productivity in the new. as follows. Value added per worker in each sector, v, can be written as v = r, *k + w where kj = capital- 6. This is different from the case of a = 0 and capital labor ratio in sector i, wj= wage (or marginal prod- and labor reallocation. Here it is assumed that old uct of labor) in sector i, r, = profit rate (MPK) in capital is left in use in the old sector instead of being sector i, and i = 0 (old/large enterprises) or i = 1 (new reallocated, despite its low (or null) productivity in enterprises/SMEs). If a is the "shiftability coefficient" the new sector. 52 Imposing Discipline E nterprises that subtract value at the prices prevailing after liberalization contribute negatively to growth. They must be subjected to market discipline through such instruments as hard budget constraints and competition in the markets for their products. Managers' behavior should be monitored using workers' councils or banks while the institutions of corporate governance are strengthened. At the same time, governments need to finance a targeted social safety net for those hurt by the imposition of discipline on inherited enterprises. Both governments and enterprises need to meet their expenditure obligations on time and in cash to prevent payments arrears. Soft Budget Constraints Can Create Macroeconomic Crises The loss of fiscal control accompanying the transition has led to a mushrooming of implicit and 1contingent liabilities. Soft budget constraints are generally more prevalent in the CIS than in the more advanced market reformers of Central and Eastern Europe. An index of soft budget constraints-- measured by the proportion of enterprises in the Business Enterprise Environment and Performance Survey (see box 3.1) reporting arrears to the tax authorities and to state-owned energy producers, which are two of the significant sources of implicit subsidy-is highest in the Caucasus and Moldova, followed by the Czech and Slovak Republics and Croatia, followed by the Russian Federation, Ukraine, Roma- nia, the Kyrgyz Republic, and Kazakhstan. By repeatedly allowing enterprises and banks to shift theiir losses to the budget, governments have injected an enormous moral hazard into the economic environ- ment, with severe consequences for the economy and the credibility of fiscal policy. Some examples: * The state budget deficit was about 1.3 percent of GDP in the Czech Republic in 1997-98, but the "hidden" deficit out of budget "transformation agencies" and guarantees was almost three times as large. The main reason was the softness of the budget constraint of the enterprise sector-not only for remaining large state-owned enterprises but also for newly privatized enterprises. * In Croatia and Lithuania part of the very high current account deficit stemmed from the practice of public utilities to borrow abroad with government guarantees. In Bulgaria and Romania an 53 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Fonner Soviet Union excessive use of off-balance-sheet items notes, and tax offsets, where government spend- erupted into full-blown macroeconomic cri- ing arrears and overdue tax payments are mutu- ses during 1996-98 when contingent liabili- ally cancelled). In turn, the public utility compa- ties arising from the growth of bad debts in nies have passed on the costs of the implicit the banking sector were converted into a fis- transfers to the general fiscal accounts by run- cal burden requiring explicit financing. ning up huge tax arrears and unpaid dues to * As a result, all these countries needed "sec- extrabudgetary funds, so that the burden of sub- ond generation" stabilization efforts in the sidies has eventually led to the accumulation of second half of the 1990s. public debt. * In Russia the failure to impose hard budget In Ukraine it is estimated that regional en- constraints through widespread use of arrears ergy companies alone have provided annual fi- and noncash payments led to government nancing to nonpayers on the order of 4 to 5 per- borrowing on an unsustainable scale, particu- cent of GDP. Moreover, governments faced with larly in light of the global financial crisis in tax and expenditure arrears have complicated 1998. That eventually led to default on a large the nonpayments problem by encouraging bar- part of sovereign debt, an event that had se- ter and other nonmonetary instruments to con- rious repercussions throughout the CIS. duct mutually offsetting operations.' Much of this has continued regardless of government reso- In the poorer CIS countries-Georgia, the lutions and legal acts forbidding the absorption Kyrgyz Republic, Moldova, and Tajikistan- of further energy arrears by the budget. Enter- none of which entered the transition with exter- prises, their domestic and foreign suppliers (such nal debt, soft budget constraints on public en- as Gazprom of Russia), and creditors believed, terprises, the difficulties associated with raising correctly, that despite the statements, some ex- energy prices and disconnecting nonpayers to plicit or implicit government support would be reduce consumption, and low tax collections forthcoming. The absorption of gas arrears of have contributed to a rapid buildup of external public institutions by the Moldovan budget in debt. External shocks, such as the Russia crisis 2000 was the fifth cycle of contingent liability of 1998, have further exacerbated the problem. creation and moral hazard since the country's All four have a long-term solvency constraint and independence. Payments arrears eventually had face tight liquidity in the next few years. Pre- to be absorbed by the budget, often through tax liminary analysis suggests that new financing on offsets, or, for energy-dependent CIS countries, highly concessional terms and, in some cases, through intergovernmental agreements with Rus- generous debt relief may be needed to restore sia or Turkmenistan. them to sustainable debt. The case for interna- Nonpayments problems have been com- tional action will be strengthened if the debtor pounded in countries whose subnational govern- countries implement strong up-front adjustment ments have considerable autonomy, such as Rus- measures, eliminating tax exemptions and con- sia and Ukraine. Clearance of tax arrears through fronting powerful special interests (box 5.1). tax offsets gives subnational governments oppor- tunities to increase their retention of shared taxes Non payments Weaken the Incentives for at the expense of the center. In Russia subnational governments are more actively engaged than the federal government in clearing tax arrears M /[ any governments have chosen to keep through offsets and individual tax exemptions lVl unviable enterprises afloat by extending and deferrals for inefficient enterprises. implicit subsidies through the energy companies. The consequences of the nonpayment syn- In Russia implicit energy subsidies to manufac- drome go beyond the budget. The failure to turing enterprises averaged 4 percent of GDP in harden budget constraints and the opacity of 1993-97. They take the form of arrears and non- noncash payments have weakened incentives cash settlements (including barter, promissory to use existing assets efficiently and to 54 Imposing Discipline Box 5.1. External Debt and Fiscal Sustainability in the Low-Income CIS Countries Armenia, Georgia, the Kyrgyz Republic, Moldova, and Tajikistan are among the poorest countries in the word-with per capita incomes in 2000 ranging from US$170 in Tajikistan to US$610 in Georgia. All five have small economies, narrow export bases, and depend heavily on energy imports. Four of the five are landlocked, and several face natural or conflict-related constraints on intemational trade. All began economic transition a decade ago with almost no debt, but by the end of 2000 their total nominal external debt exceeded US$7.1 billion, or 84 percent of their combined GDR The net present value of their debt outstanding at the end of 2000 averaged 158 percent of their eKports and 358 percent of their central government revenues. Many factors have contributed to the accumulation of debt. Sharp increases in energy prices and the loss of transfers irom the central government of the former Soviet Union after its breakup were massive shocks to these economies. Regional and intemal conflicts hindered economic recovery. Policy failures, corruption, and weak governance have played an important role, too. These countries have not been able to attract much in grant aid from bilateral donors and have relied heavily on the International Monetary Fund and World Bank for financial assistance. In the mid-1990s, the intemational financial institutions and other international creditors overes- timated the implementation capacity of their governments to conduct the complex and socially painful reforms associated with the transition. The terms of some of the financing received were not always suitably concessional. Finally, the 1998 financial crisis in Russia severely disrupted their trade and financial sectors. Four of the five countries (Armenia is the exception) were forced to deeply devalue their currencies, leading to sharp increases in the domestic currency cost of extemal debt service. Although the severity of the external debt burden varies among the five countries, the weakness of their external financing and fiscal positions is likely to be a serious constraint to growth and poverty reduction, even in the medium term. Action is needed on two broad fronts if they are to make better progress toward sustained economic growth and poverty reduction with strengthened financial viabilitt. First, the low-income CIS countries need to improve their policy environments. Fiscai reforms are required to ensure improved efficiency and effectiveness of public expenditures within the tight overall resource constraint. The pace of reform needs to be accelerated in key sectors, such as energy, which have strong links to the fiscal and external debt situation. Major reforms are needed to improve the investment climate, notably the environment for entry by new enterprises, which has proven to be a key success factor in Central and Eastern Europe. Institutional changes to strengthen public and private sector govemance will be important in this regard. Second, they need to work closely with their extemal partners to secure a volume and blend of financial assistance consistent with their absorptive capacity and policy efforts, as well as their fiscal and debt-carrying capacities. International Monetary Fund and World Bank policy- based credit operations are either ongoing or planned in each of the countries. Other donors need to consider increasing their assistance, notably through grants and other highly concessional funds. The scope for and costs and benefits of external debt rescheduling for these countries also needs to be examined. Donors should provide debt relief promptly and on highly concessional terms where this is demonstrated, as warranted by both financial need and policy efforts. restructure enterprises. Nonpayments reinforce for numerous formal and informal financial- existing interenterprise relationships. They also industrial groups, thus impeding competition andA weaken competition by segmenting product new entry. Furthermore, subnational governments markets and dealing with a few key suppliers have protected incumbent enterprises with which and customers. Furthermore, by reducing trans- they have built close relationships through parency in accounting and transactions, they nonpayments at the expense of new entrants, in- complicate monitoring of enterprise managers. hibiting investments and postponing the resump- Even healthier enterprises have benefited by tion of growth. colluding with unviable enterprises to divert the The experience of the 1990s clearly demon- implicit subsidies available in the system rather strates that sustainable growth and low inflation than look for ways to restructure and increase require enterprise restructuring and exit, new en- efficiency. try, and fiscal adjustment. But none of these can The widespread use of arrears and noncash be addressed without tackling the nonpayments payments led to the proliferation of vertically in- problem that affects them all. For this, the chal- tegrated conglomerates and provided an impetus lenge facing Russia and other CIS countries 55 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Former Soviet Union differs somewhat from the hardening budget con- compliance could lead to federal transfers being straints that the reforming countries of Central cut. The result is a chain reaction: the federal and Eastern Europe faced at the onset of transi- budget does not transfer the funds needed for tion. The following actions comprise the main residential electricity and heat, so the local au- elements of a solution to the nonpayments crisis thorities insist that the municipal utility compa- (Pinto, Drebentsov, and Morozov 2000): nies foot the bill. The latter in turn insist that the households pay. For enterprises, the main * The government pays its bills in cash and on time and refrains from engagig in m l pressure comes from taxes and cash payments time and refrains from engaging in mutual fo enry.nadto,te etutrn fps offsets to cancel tax and budgetary arrears. for energy. In addition, the restructuring of past • Enrgy ompaiespay heirtaxs incashand arrears is taking place by mutual consent and on * Energy companies pay their taxes in cash and condition that current obligations are met in full. On time. * Companies enforce a policy of disconnect- ing those delinquent on payments, thus im- Exit Mechanisms-Implement Now, posing a hard budget constraint on those Revise Later enterprises. urning off the spigot of fiscal and quasi- * Social assets traditionally provided by enter- u Tfofafiscal and quasi- prises (such as housing, health clinics, and kin- Ifiscal support and keeping government prises (such as housing, health clinics, and kin- agencies current on their obligations is necessary. dergartens) are divested to local governments. gg dergtenets) are tagested to theloal -irncments. However, this alone is liable to lead to more accu- * Benefits are targeted to the low-income mulation of arrears and, without complementary households most affected by an increase in policies, will not be enough to lead to restructur- energy prices to international levels. ing or closure. Two such policies are important. * Some types of one-company towns in distant First, restructuring and exit will require a transfer regions of the CIS, such as distressed settle- of responsibility for housing and other social as- ments in isolated areas and settlements with sets from enterprises to local governments, an area a p o n re o twhere the roles and responsibilities of local gov- cial arrangements. ernment suffer from lack of clarity and absence of The substantial real devaluation of 1998 has financing. Second, measures of administrative liq- given Russian enterprises some (temporary) uidation need to be strengthened. breathing space, making it easier to eliminate the Almost all countries in the region have bank- implicit subsidies transmitted through tax and ruptcy laws. In a number of countries there were energy payments. Noncash settlements continue significant reforms of these laws in 2000. But to abate as a result of improved enterprise li- these formal exit mechanisms have not been par- quidity from the devaluation. Although tax rates ticularly effective. The European Bank for Re- continue to be negotiated, almost 100 percent construction and Development's legal indicator of current taxes are paid in cash. survey indicates that there is a gap between the Budget constraints are hardening though the adoption or amendment of bankruptcy legisla- very channels that once transmitted hidden sub- tion and its effective implementation (EBRD sidies: government budget management, taxes, 2000). Experience suggests that it is preferable and energy payments. Cash collections from en- to implement existing insolvency laws and revise terprises by regional energy companies have in- them once they have been put into practice. creased significantly, a development attributed to unremitting pressure from such infrastructure Competiion Is Linked to Innovation monopolies as Gazprom and RAO Unified En- and Growth ergy Systems. Indeed, these monopolies have become potent instruments for hardening bud- ncouraging competition in product markets get constraints and insisting on cash payments. E is another important ingredient of discipline. Tax offsets are forbidden for the value-added Because socialist economies were highly con- tax, profits tax, and income tax, and lack of centrated, exposing enterprises to internal and 56 Imposing Discipline external competition is important. Results from innovate. Where monitoring of managerial per- the Business Enterprise Environment and Per- formance by debtors and external shareholders formance Survey (see box 3.1) show that nearly is weak-as is generally the case in the CIS and 30 percent of state-owned enterprises face no Southeastern Europe-competitive product competitors, compared with 5-9 percent for markets can help discipline enterprise manag- private enterprises. While more than half of ers. The link between competition, innovation, state-owned enterprises have more than three and growth emerges as particularly important. competitors in their main product market, 80 Governments should thus vest the competition percent or more of privatized and new private policy agency with the authority to enforce comt- enterprises find themselves in that situation petition laws strictly. (EBRD 1999). The survey also shows that enterprises fac- ing an intermediate degree of competition (one Note to three competitors) develop new products, re- to three competitors) develop new products, re- 1. As an illustration of failing to harden budget con- place managers, or change their organizational straints by disconnecting nonpayers, the Russian gov- structure if they are subject to a hard budget ernment threatened to block oil producers' access to constraint, with their moderate degree of mar- export pipelines if producers stopped supplying oil to ket power providing the reward necessary to nonpaying domestic refineries. 57 Extending Encouragement: W hat is needed to encourage the formation, and then growth, of new enterprises? This chapter identifies administrative barriers to entry as one of the most important obstacles. lt also looks at aspects of legal and judicial reform that impinge on the security of prop- erty rights. This chapter describes what needs to be done to develop a financial sector capable of nurturing new enterprises and monitoring enterprise reform. It also reviews reforms in the tax system and in intergovernmental fiscal relations that support the discipline-and-encourage strategy. Corruption and Anticompetitive Practices Mar the Investment Climate The Business Environment and Enterprise Performance Survey (see box 3.1) unbundled factors influencing the investment climate into microeconomic variables (including taxes and regula- tions), macroeconomic variables (including policy instability, inflation, and exchange rates), and lavv and order (including functioning of the judiciary, corruption, street crime, disorder, organized crime, and mafia). According to the respondents, taxes and regulations are consistently among the most important impediments to expansion by new enterprises. Within this category, the granting and an- nual renewal of business licensing-and the opportunities for corruption that this can provide-are seen as serious obstacles. The other obstacles, in order of importance, are inflation, lack of access to finance, corruption and anticompetitive practices, and lack of access to infrastructure services (Hellman, Jones, and Kaufmann 2000). The situation is reported to be consistently worse in the CIS and Southeastern Europe than in Central Europe and the Baltics. The survey also shows that small enterprises across the region pay, on average, 5.4 percent of their annual revenues in bribes, as opposed to 2.8 percent for large enterprises (World Bank 2000c). The regressiveness of the bribe tax is highest in Armenia, Moldova, Ukraine, and Uzbekistan. The frequency of bribe payments is about double for small enterprises: 37 percent report paying bribes frequently, compared with 16 percent for large enterprises. New enterprises see corruption and anticompetitive practices as two of the most difficult obstacles across all parts of the region. Business licensing, which falls in the category of taxes and regulations, is 59 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Former Soviet Union seen as particularly troublesome by new enter- In Ukraine an International Finance Corpora- prises, given the opportunity for arbitrary be- tion study showed that small enterprises endure havior in interactions between officials and en- an average of 78 inspections a year, requiring 68 terprises in granting and renewing licenses. Also written responses. Dealing with inspections and troublesome are regulations for customs, foreign audits consumes two days a week of the average trade, foreign currency, foreign exchange, and manager's time and requires cash outlays of taxes. Examples of such obstacles abound. about US$2,000. A survey of Russian enterprises by the World The administrative problems that new en- Bank in 1996 shows that the average new busi- trants in Armenia (box 6.1) face in registering, ness applicant had to deal with about 25 differ- locating, and operating their companies are per- ent agencies and complete about 70 registration vasive throughout the region. The solutions lie forms. There were 30 kinds of licenses for a busi- in simplifying processes to make them transpar- ness startup. About a third of surveyed enter- ent and consistent across all government agen- prises indicated that they were forced to obtain cies. Similarly, customs procedures need to be re- a license that in their opinion was not required. engineered to make them more transparent and Box 6.1. Reducing the Cost of Entry and Doing Business in Anmenia New enterprises face many problems in Armenia that could be reduced with appropriate measures. Registaton Company registration is cumbersome, involving half a dozen agencies and includes such antiquated practices as obtaining a company seal. Procedures are described in general terms in the laws, leaving substantial discretion to civil servants. Registration normally takes three to six months. Policy recommendations are to: * Develop a centralized registration process with one oversight agency * Create transparent registration requirements and procedures and make them easily accessible to the public * Eliminate the outmoded requirement for a company seal. Registering ownership rights is a convoluted bureaucratic process, designed more as a data collection system than as a process to register legal rights. Many steps appear irrelevant. Armenia's land cadastre system records ownership based on street addresses or passport numbers, rather than a unique cadastral code established through surveys. Site development is equally difficult, with numerous municipal and state agencies involved in issuing the approvals and clearances required for construction and occupation permits. There is no comprehensive description of the required procedural steps, and the authority of various agencies often overlaps, resulting in frequent delays and contradictory decisions. Policy recommendations are to: * Redesign ownership registration by eliminating or streamlining all steps not directly related to registering legal titles * Assign a unique cadastral code to all real property units and create a consistent register * Develop clear and transparent rules for site development that are consistent across all government agencies * Harmonize all requirements of the process throughout the country * Eliminate duplication of effort by varous agencies * Introduce application forms for all agencies and attach a reliable and clear guide * Set timetables for each phase. Source: World Bank data. 60 Extending Encouragement to reduce the scope for arbitrary decisionmaking and abuses of power. FIGURE 6.1. Insecurity of Property Rights in Transition Enterprises Lack Confidence in Legal and Economies, 1999 Judicial Institutions Estonia_ T he state can create a good climate for in- Uzbekistan T vestment by protecting the security of prop- Poland erty and contract rights. Variation in the Slovenia security of property rights in the region is high Hungary (figure 6. 1). Fewer than 30 percent of enterprises Azerbaijan in such countries as Croatia, Estonia, and Po- Croatia land lack confidence in the security of property Slovak Republic rights, compared with more than 70 percent in Georgia Russia and Ukraine. Bulgaria Institutions contributing to the security of Armenia property rights are diverse: property registries; a Roma Alania stable and appropriate body of laws, regulations, Albaniac and administrative procedures; effective systems Czech Republic to adjudicate administrative and civil disputes; larus and legal expertise (lawyers, notaries, judges, Kazakhstan prosecutors, police, and bailiffs) to ensure that Lithuania the legal framework is enforced. Data are avail- Kyrgyz Republic able on the quality of legal drafting (figure 6.2), RussiaC which is eventually linked to the stability and Ukraine appropriateness of the legal framework and to Moldova the quality of the judicial system (figure 6.3). 0 20 40 60 80 On the basis of the Business Environment and Enterprise Performance Survey, more than Percent 90 percent of enterprises in 13 of the sample countries report that they are not adequately Source: EBRD (2000). consulted in the drafting process for new laws or policies. More than 80 percent of enterprises in 13 of the sample countries report that they complain about low enforcement rates, while are not adequately informed of changes in rules in Albania almost 90 percent of enterprises do. that affect them before these rules are adopted. Almost two-thirds of the variation in the se- The quality of the judiciary is unbundled into curity of property rights across countries can be fairness (bias, honesty, consistency), cost (finan- attributed to variation in their systems of legal cial and time), and likelihood of enforcement drafting and the effectiveness of their judiciaries. (see figure 6.3). There is wide variation in per- Thus, attention to the process of drafting and ceptions of the quality of the judiciary. In Hun- enforcing laws may help ensure that laws are gary, Estonia, and Slovenia only a third of en- suited to achieve economic outcomes-in this terprises report that court decisions are unfair, case, secure property and contracts. while in Armenia and Lithuania almost 90 per- A study based on a survey of manufacturing cent of enterprises do. In Uzbekistan fewer than enterprises in Poland, Romania, Russia, the Slo- half of the enterprises complain about the high vak Republic, and Ukraine found that enterprises cost of litigation, while in Poland more than 90 with the least secure property rights invested percent of enterprises do. In Bulgaria and nearly 40 percent less than those with the most Slovenia only about a quarter of enterprises secure property rights.' Moreover reinvestment 61 Transition-The hrst Ten Years: Analysis and Lessons for Eastern Europe and the Former Soviet Union FIGURE 6.2. Quality of Legal Drafting in Transition Economies, 1999 (percentage of enterprises that complain they are seldom or never consulted about new rules) Azerbaijan Slovenia Czech Republic Estonia Uzbekistan Georgia Bulgaria Croatia Poland Slovak Republic Belarus * Voice Armenia * Publicity Albania Latvia _ Hungary Moldova Kazakhstan Ukraine _ Russia Kyrgyz Republic Lithuania Romania 50 60 70 80 90 100 Percent Note: This figure shows partial results from two questions on the Business Environment and Enterprise Performance Survey (see box 3.1). The first question measured enterprises' voice in legal drafting: "In case of important changes in laws or policies affecting my business operation, the government takes into account concerns voiced either by me or by my business association." The second question measured the extent to which the state publicizes new rules (publicity) before their implementation: "The process of developing new rules, regulations, and policies is usually such that businesses are informed in advance of changes that will affect them." For both questions the possible responses were "always, mostly, frequently, sometimes, seldom, or never." Source: World Bank data. of profits was a bigger source of investment capi- strengthen property rights, a well-functioning tal than either bank funds or trade credit in all financial sector is needed to provide credit to five countries. In such countries as Russia and enterprises making new investment and requir- Ukraine, which have made less progress in tran- ing financing beyond their retained earnings. sition than Poland and the Slovak Republic, re- forms to secure property rights are more impor- Financial Deepening Is Slow but Progressing tant for promoting the investment climate than reforms in the banking sector. Then, as the tran- Ithough securing property rights appears to sition proceeds and legal and judicial reforms A be more important than reforming the 62 Extending Encouragement FIGURE 6.3. Quality of Judiciary in Transition Economies (percentage of enterprises that complain that courts sometimes, seldom, or never exhibit positive qualities when resolving business disputes) Slovenia Bulgaria Uzbekista n Slovak Republic Estonia Romania Belarus Hungary Armenia Azerbaijan * Unfair Lithuania * High cost Poland E:i Weak enforcement Georgia Czech Republic Croatia Ukraine Moldova Kazakhstan Kyrgyz Republic Russia Latvia 'Albania_ 0 20 40 60 80 100 Percent Note: This figure shows partial results from the following question on the Business Environment and Enterprise Performance Survey (see box 3.1): "Now, thinking about our country's legal system, how often do you associate the following descriptions with the court system in resolving business disputes?" The descriptions were: fair and impartial; honest/uncorrupted; quick; affordable; consistent/reliable; able to enforce its decisions. The possible responses were "always, usually, frequently, sometimes, seldom, or never." The "unfair" bar comprises the descriptions of fair and impartial, honest/uncorrupted, or quick. The "high cost" bar comprises the descriptions of quick or affordable. The "wealk enforcement" bar comprises the descriptions consistent/reliable and able to enforce decisions. Source: World Bank data. banking system in the early stages of transition, way in developing banks and capital markets, financial deepening and development is key to their financial sectors remain underdeveloped by sustaining the growth of the new private sector. international standards. Banking systems in most While transition economies have come a long transition economies remain noticeably small iin 63 Transition-The First Ten Years: Analysis and Lessons for Eastem Europe and the Former Soviet Union international perspective, even more so once years (figure 6.6). In 1997 all Central European bank credit to the public sector, often reflecting countries except Slovenia had operating cost to subsidization of unprofitable activities, is re- total asset ratios that were lower than those in moved and attention is confined to the private countries with comparable incomes elsewhere. sector (figure 6.4). While the Baltic countries had not quite matched Similarly, stock market capitalization, a mea- their peers in the world, their average ratio of sure of the capacity of securities to provide fi- operating costs to total assets fell from 7.8 per- nance to the real economy, shows that financial cent in 1996 to 5.6 percent in 1997. The remain- development lags behind other countries of the ing transition economies with data have ratios world with similar per capita incomes (figure 6.5). less favorable than their peers, but they too (ex- Relative to their peers in the region, Croatia, the cept Romania) saw efficiency improve. An alter- Czech Republic, and Hungary boasted the high- native measure of efficiency-the difference be- est stock market capitalization in 1998.2 tween the average interest rate on loans to the private sector and the resident deposit rate-also shows Central Europe and the Baltics in a favor- able situation, but not Southeastern Europe and Although the financial sector appears underde- the CIS (figure 6.7). veloped from an international perspective, this Although domestic credit to the private sec- is not necessarily true for financial intermedia- tor (as a share of GDP) in the transition econo- tion in the banking sector. The ratio of operat- mies is generally low in comparison with coun- ing costs to total assets reveals that several Cen- tries of similar per capita incomes (figure 6.4), tral European countries have progressed in recent in many of the most advanced reformers an FIGURE 6.4. Domestic Credit by Deposit Money Banks to the Private Sector, 1998 Percentage of GDP 100 _ 80 - World fitted line * Central Europe and 60 -the Baltics 60 vak~~~~~~~~~~~~~~~~~~Republic Czech RepublicthBaic 40 - Slovenia * Commonwealth of *Estonia * Independent States 20 Moldova *iR LatviaBelarus * Poland o Southeastern Europe Armenia . a Ukraine o $ $Russian Federation O - AermbenjaOanb *# c d Romania Lithuania 0 5,000 10,000 15,000 Per capita income (US$) a. Kyrgyz Republic b. Albania c. Georgia d. Kazakhstan Note: The world fitted line represents the average of all countries in the world in that per capita income range. Source: Eichengreen and Ruehl (2000). 64 Exteniding Encouragement FIGURE 6.5. Stock Market Capitalization and Per Capita Income, 1998 Percentage of GDP 70 - 60- 50 World fitted line 40 - Central Europe and the Baltics 30 -CehRpbi CrC Hungary zechRepublic * Commonwealth of 20 a tia Poland Slovenia Independent States Armneniga Ukraine n Fedriai * Slovak Republic O Southeastem Europe O ..............c* * Russia Moldova Azerbaijan Kazakhstan Romania - 10III 0 5,000 10,000 15,000 Per capita income (US$) Note: The world fitted line represents the average of all countries in the world in that per capita incomie range. Source: Eichengreen and Ruehl (2000). FIGURE 6.6. Operating Costs to Total Assets in the Banking Sector, 1997 Percentage of total assets 14 Kazakhstan o 12 World fitted line 1 0 Ukraine Romania * Russian Federation * Central Europe and 8 - Bulgaria the Baltics 6 °Lavi *a Lithuania * Commionwealth of 6 _ Latvia Estonia Slovenia Independent States 4 Poland Hungary + Southeastern Europe 2 - Slovak Republic Czech Republic 0 0 5,000 10,000 15,000 Per capita income (US$) Note: The world fitted line represents the average of all countries in the world in that per capita income range. Source: Eichengreen and Ruehl (2000). 65 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Former Soviet Union FIGURE 6.7. Interest Rate Spreads, 1998 Percentage points 50 40 _ Kyrgyz Republic Wodd fitted line Ukraine . Central Europe and 30 - Arrnenia * Georgia Russian Federation the Baltics 20 Romania* Commonwealth of BIgaa ° Belarus Crti Independent States 10 u gr * CrtEstonia Slovenia O Southeastem Europe Moldova Macedonia. FYR Latvia * * 0 Uthuania Poland Slovak Republic | Hungary Czech Republic 0 5,000 10,000 15,000 Per capita income (US$) Note: The world fitted line represents the average of all countries in the world in that per capita income range. Source: Eichengreen and Ruehl (2000). important part of that credit is now flowing Hungary. Thus comprehensive financial restruc- to new enterprises. In Estonia, Hungary, and turing and curtailment of loans to the banks' tra- Latvia between two-thirds and three-quarters ditional clients have been associated with faster of total private credit is going to those enter- growth of financial intermediation to new private prises. clients. The experience of Hungary suggests that The Czech Republic and Hungary provide an restructuring the financial sector and developing interesting contrast. The Czech Republic has the financial intermediation can go hand in hand. highest share of credit to the private sector, about 60 percent of GDP, (see figure 6.4), but only one- tenth of it flows to new enterprises. In Hungary, Quickn the Dicie the credit provided to the private sector is much Quicken the Pace? lower (about 25 percent of GDP), but about three- Hard budget constraints should apply to banks quarters of it goes to new enterprises. Why? Be- and enterprises alike. A credible threat of exit is cause in the Czech Republic most banks were still necessary for a competitive banking system. But state-owned and largely unreformed, and they a bank differs from a nonfinancial enterprise, so continued to lend to their traditional clients: large a decision on liquidation or restructuring, espe- semi-privatized enterprises. This is consistent with cially for large banks, must take into account the a situation of banks not having been positioned systemic risk of closure. The efficiency of banks to do their job of screening, together with the improves if they are privatized to strategic inves- limited industrial restructuring and lack of growth tors. If such investors are not forthcoming, in the Czech Republic in recent years. privatization to concentrated owners should be Hungary, by contrast, sold off almost all its considered, so long as there is a clear separation commercial banks to foreign investors early on in between investors and borrowers from the bank. the transition. These new private banks applied Facilitating the entry of foreign financial in- to their borrowers the hard budget constraints termediaries can accelerate the development of imposed on them by their headquarters and strict the banking sector and in turn help the growth supervision by the supervisory agencies in of start-ups not yet ready to tap capital markets. 66 Extending Encouragement Entry by foreign banks and the acquisition of response to improved policies and improved domestic banks by foreign banks are quick ways overall business environment. Some studies have of importing managerial and supervisory exper- shown a clear statistical relationship between tise, the latter because supervisory responsibility the growth of foreign direct investment and the resides mainly with the home country regulatory indices of economic liberalization discussed in authority. While encouraging competition is criti- chapter 2 (Selowsky and Martin 1998). The cal for the long-term health of commercial banks, ability to attract foreign direct investment is thus it should be done without relaxing prudential a good proxy for the attractiveness of a norms. These norms should cover minimum capi- country's overall investment climate and is also tal requirements, regulatory capacity, legal re- crucial to promote the entry and growth of course, and the independence of supervisory au- SMEs. Actually both are well associated as thorities from political interference. shown in figure 6.8. Recent work on the relationship between fi- In addition to the positive spillovers of ira- nance and growth suggests that banks and secu- proved technology, better management skills, arid rities exchanges provide different services, both access to international production networks, for- required in a mature market economy (Levine eign direct investment has proved resilient in vola- and Zervos 1998). The financial system in tran- tile international capital markets. Although the sition economies will most probably remain Russia crisis of 1998 led to sizable outflows of heavily bank-based for a while. New enterprises bond and portfolio equity capital, this was offset need to be nurtured and this phase calls for a by an increase in foreign direct investment in the bank-based financial sector capable of provid- advanced reformers in Central Europe and the ing risk assessment and venture capital services. Baltics. Russia and Ukraine, reluctant to attract A later phase will call also for an active and li- foreign participation in their privatization pro- quid securities market that is more suited to fund- grams, had attracted only modest amounts of for- ing risky projects and offers fuller risk diversifi- eign direct investment but substantial portfolio cation. This later phase will require even more flows. Much of the latter found its way into fi- demanding institutional developments such as a nancing public sector deficits arising from an over- regulatory framework requiring information dis- all climate of soft budget constraints. When the closure and preventing insider trading and other government's solvency was questioned in the 1998 forms of market manipulation. crisis, portfolio flows to Russia quickly reversed. Privatization Attracts Foreign Direct Tax Reform: Broadening the Base and Investment, and Positive Spillovers Follow Lowering Rates T here are significant differences in the levels T he agenda of tax reform in the transition T of foreign direct investment flowing to the T economies is large. Its guiding principle, transition economies. In the 1991-2000 decade, based on the lessons of theory and experience, is the CIS received cumulative foreign direct invest- that a tax system raising revenue for the govern- ment per capita of about US$115 compared with ment at least cost to the economy should be stable US$800 for the CSB. The Czech Republic and and broad-based and have low statutory rates. Hungary had the highest amounts, almost The focus in this section is narrow, restricted to US$2000, followed by Estonia, Latvia, and aspects of taxation that bear directly on disci- Croatia in the US$300 to US$1000 range. pline and encouragement. Cash privatization of large enterprises has The key objective of tax reform is to broaden driven much of these foreign direct investment the base of taxation and reduce statutory rates. flows; cumulative foreign direct investment is This would eliminate tax exemptions enjoyed by highly correlated with cumulative privatization favored enterprises, thus hardening budget con- revenues (EBRD 2000). However, foreign di- straints. It would also reduce the burden of taxa- rect investment has also flown to countries as a tion on viable enterprises and encourage the 67 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Former Soviet Union FIGURE 6.8. Cumulative Foreign Direct Investment Per Capita and Employment in Small Enterprises, 1998 Cumulative foreign direct investment (US$) 1,800 - 1,600 Hungary* 1,200 1,000 - Estonia Czech Republic 800 Latvia 600 - 400 Kazakhstan Lithuania 200 Bulgaria Romania G 0 F a J *1 I 0 10 20 30 40 50 60 Contribution of small enterprises to employment (percent) a. Ukraine. b. Belarus. c. Russian Federation. Source: EBRD (2000); World Bank database on SMEs. informal sector to come within the tax system. using simplified accounting. Such an approach The informal sector in 1995 was estimated at 60 would relieve the administrative and reporting percent of GDP in Azerbaijan and Georgia, be- burden on the taxpayer and reduce the interac- tween 40 and 50 percent in Russia and Ukraine, tion between the taxpayer and the tax authority. and between 30 and 40 percent in Bulgaria, For agriculture, a property tax needs to be Kazakhstan, Latvia, and Moldova. In the Kyrgyz developed as land privatization continues. This Republic tax exemptions, mostly to foreign joint could be the primary method of agricultural ventures, have in the past cost the budget the taxation in the foreseeable future. Paralleling equivalent of 5-7 percent of GDP, while revenue the earlier discussion, a distinction needs to be collections remain below 16 percent of GDP. In made between small agricultural units and Russia the potentially large tax base was eroded larger agribusinesses. Smaller agricultural en- by exemptions granted by subnational authori- terprises should be subject to the small busi- ties on federal taxes, estimated at 0.8 percent of ness tax (or personal income tax, as appropri- GDP (World Bank staff estimates). ate). Larger enterprises should become more Taxes on small businesses are widely reported fully integrated into the modern income tax and to be a disincentive to creating small enterprises value-added tax systems. in the region. The turnover threshold for becom- ing a value-added taxpayer should thus be set high enough to exclude small enterprises, which should Sueroertngent Fal RElations be subject to a small-enterprise tax regime, per- haps as simple as a fixed annual fee in lieu of taxes Jt is important for the incentives facing federal other than wage and social taxes. Larger enter- and local governments to be aligned in sup- prises might be subject to either a turnover tax at port of the discipline-and-encourage strategy. a moderate rate or to a simplified cash flow tax Subnational governments should have incentives 68 Extending Encouragement to identify with small business. To save jobs and the service at a higher level of government. In to protect existing political structures these gov- yet other cases it means a complete redesign of ernments often protect large (often bankrupt) the provision of social services and benefits. enterprises in their localities, even though job Many countries in the region have shifted so- growth generally increases with the development cial responsibilities to local governments, but of small businesses. However, the latter takes time. have not ensured that local governments have To help subnational governments identify enough resources to manage these responsibili- with new emerging enterprises, small business ties, either through bigger transfers or greater taxes and property taxes could be allocated to tax autonomy. In addition, the divestiture of so- that level of government. Subnational govern- cial services is frequently done in a way that leads ment officials will then see that the success of to confusion over which level of government is small business means more local revenues. Com- responsible for what. Roles and responsibilities pliance with the property tax will be enhanced if of subnational governments have to be clarified. the tax is tied to improvements in public services, In addition, enough resources must be allocated such as police protection demanded by local to the additional responsibilities of absorbing businesses. That in turn increases property val- social assets from enterprises. That reduces the ues, reinforcing mutual interests. incentive for localities to favor old enterprises It is also necessary to ensure that local gov- over new. ernments do not face an incentive to treat en- terprises differently because they continue to provide social services, such as hospitals, kin- Notes dergartens, and housing recreation facilities. In 1. The property rights variables used in the analysis many cases municipalities do not have the re- combine an index of property rights security compris- sources to provide these services. This creates ing extra legal payments for licenses, extra legal pay- incentives for subnational governments to pro- ments for services, and payments for protection with vide special treatment to unrestructured and a measure of the effectiveness of courts. See Johnson, McMillan, and Woodruff (2000). sometimes nonviable old enterprises, with the attendant deleterious consequences for the 2. Several transition economies, notably in Central growth of new enterprises. Europe, appear to perform better when market liquid- It is important that an effective strategy be ity is measured by the ratio of turnover (a measure of developed and implemented for divesting enter- the liquidity of the securities) to market capitaliza- tion. These high turnover ratios, rather than reflect- prises of these social services. In some cases this ing gcnuine liquidity of rhc markcts, arc driven by means placing the responsibility of service pro- high privatization sales captured (the numerator) in vision on the municipality. In others it means conjunction with low levels of capitalization (the de- reassigning responsibilities, or at least financing nominator). 69 Privatization: Lessons and Agenda for the Future P rivatization has been key in the transition from plan to market. Rapid privatization early in the transition aimed to get the state out of enterprise management, to create a broad constituency for reforms, and to arrest "spontaneous privatization" from pretransition reform attempts in countries as diverse as Hungary, Poland, the Russian Federation, and Ukraine. Privatization was a way of imposing hard budget constraints and promoting restructuring. It was also a way of creating demand for stronger property rights and institutions of corporate governance, thus contributing to private sector development. What is the evidence on privatization's effect on restructuring in medium-size and large enterprises over the past decade in transition economies? Djankov and Murrell's (2000) recent review of the literature, covering about 30 studies, concludes that: * Privatization to concentrated outsider owners (investment funds, foreigners, and blockholders) has benefited restructuring * Privatization to diffuse owners and to enterprise workers and managers (so-called insiders) has not been beneficial; indeed, privatization to workers in the CIS has been worse than state ownership for restructuring. A reasonable hypothesis for the ineffectiveness of diffuse or insider ownership is the lack of an adequate institutional framework. That framework would include strong mechanisms of corporate governance, including rules to protect minority shareholders; rules against insider deals and conflicts of interest; and adequate accounting, auditing, and disclosure standards. It would also include take- over, insolvency, and collateral legislation, as well as strong creditor surveillance by well-run private banks. Without such institutions those in control of enterprise assets face few restrictions to prevent diversions for private gain. Other factors may also have contributed to the poor performance of enterprises privatized to insiders or diffuse owners. These owners probably had neither the ability to provide the capital to modernize the enterprise, nor the market experience to cope in a competitive environment, nor the drive and vision to guide the radical restructuring often required to improve enterprise performance. 71 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Forner Soviet Union Furthermore, workers in the CIS, traditionally a private ownership politically acceptable. It was weak force in the enterprise, had almost no own- also thought that any inefficiency of diffuse own- ership or management experience. Also, owner- ership would be only temporary, as secondary ship by workers may have done little for restruc- markets would consolidate share in the hands of turing if employment had to be reduced sharply. effective owners. In any event, there were few Concentrated ownership tends to offer those concentrated domestic owners to buy state as- controlling enterprises the incentive to maximize sets. Selling to foreigners was politically difficult, long-run enterprise profits. Owners can align and such investors were sometimes less available their interests with minority shareholders and to the CIS countries, given the overall uncertain- creditors, reducing the likelihood of asset strip- ties about the investment climate and lack of ping for private gain. But privatization in transi- track record. Other factors included country size, tion economies shows that concentrated owner- geography, political uncertainty, inadequate le- ship alone is not enough for effective governance. gal framework, and weak institutional capacity. Concentration of ownership must be deep Might it not have been preferable to keep the enough to reduce the divergence between the assets in state hands, waiting to identify and then interests of controlling shareholders and other sell the enterprises to reliable strategic investors? stakeholders and to lead controlling sharehold- Yes, on two conditions. First, the privatization ers to pursue long-run profit maximization. An agency needs the autonomy to discharge its func- ownership control of only 20-30 percent may tions with transparency and without political in- not be high enough to align interests. terference. Second, there has to be enough insti- The type of concentrated ownership also mat- tutional capacity to prevent asset stripping by ters. Enterprises controlled by strategic investors state managers in the interim. In many countries have performed much better than those controlled these conditions were not met, resulting in "spon- by investment funds, other stakeholders (hold- taneous" privatizations by managers when the ing companies), or other financial institutions. enterprises were still owned by the state. Take Enterprises controlled by foreign strategic inves- Russia at the end of the Gorbachev era, when tors have generally been the best performers. This the state had lost control over enterprises after is not entirely unexpected. Such investors, able the collapse of institutions during the dissolu- to provide more resources and skills for restruc- tion of the Soviet Union. turing, had an advantage over domestic strategic The challenge in economies with a large un- investors early in the transition. finished agenda of privatization is for the state The selection of strategic investors matters, to exercise its control rights to avoid tunneling too. Enterprises sold through transparent ten- and theft by enterprise managers during the years ders or auctions have generally attracted better it could take to complete privatization and while owners, outperforming enterprises sold directly enterprises remain in state hands. Yes, divesti- to politically connected parties, frequently at ture to reputable core investors who put their highly subsidized prices. capital at risk is best practice. However, that is So, the ideal strategy is to transfer assets as difficult to achieve on a large scale in a short rapidly as possible to individual investors or period. A major question then is whether inter- concentrated groups of strategic investors mediate modes of privatization-such as mass through open, fair, and transparent methods. privatization through vouchers, or management- Some countries-such as East Germany, Esto- employee buyouts-might eventually lead to the nia, and Hungary-adopted this strategy, with same result, even if they temporarily worsen en- satisfactory outcomes. Why did other countries terprise performance. not follow suit? Navigating between continued state owner- Privatization to diffuse owners and worker- ship with eroding control rights and a transfer owners was appealing on equity grounds, and to ineffective new private owners with an inad- in several countries this was the only way to make equate institutional framework is possibly one 72 Privatizafion: Lessons and Agenda for the Future of the most difficult challenges confronting the performance of privatized enterprises by policymakers in charge of privatization. There assisting a transparent consolidation of widely is a substantial risk of asset stripping and losses held shares and new private enterprises. It of economic value in both cases. would also facilitate the development of bank How much can tunneling and theft in the in- and nonbank intermediation. termediate stages be prevented? How fast can The success of the two approaches requires reputable concentrated investors become the ul- openness to foreign investors. The importance timate owners? Does the outcome depend on the of foreign capital in the first approach is obvi- path followed? The present value of the benefits ous. It makes little sense to exclude foreign in- of a specific privatization strategy depends on the vestors from a privatization program designed answers. For example, do rapid intermediate to transfer enterprises directly to effective own- modalities-such as privatization to diffused ers, as foreign investors are often in the best po- owners or insiders-accelerate or retard the even- sition to provide the resources and skills for re- tual takeover of the enterprise by the "right" kind structuring state enterprises. However, foreign of investor? What are the relative magnitudes of investors are equally important in the second the gains or losses of these intermediate stages of approach. They can usually buy large blocks of ownership in relation to state ownership? The shares in the secondary market relatively quickly, answers depend on initial conditions and the ca- while also complying with all capital market pacity to reduce tunneling (see box 3.2) and theft regulations, thus facilitating the concentration and thus enhance the benefits of the strategy. of ownership through fair and transparent sec- ondary trading. Excluding foreign investors from mass Privatization? privatization programs using vouchers could force policymakers into difficult choices. For example, Jf countries choose traditional methods of the authorities might have to tolerate unfair prac- privatization to effective owners, establishing tices to allow cash-strapped domestic investors the necessary legal and institutional framework to rapidly gain control. The concentration of and giving the privatization agency autonomy ownership in many voucher programs, as in tihe and resources are essential. However, if coun- Czech Republic, owed much to poor capital mar- tries choose methods of rapid privatization that ket regulation and weak rule enforcement. But if lead to diffuse or insider ownership, strengthen- the authorities had been willing and able to en- ing and enforcing the regulatory and supervisory force an adequate regulatory framework, they framework are crucial to enhance the account- probably would have been forced to accept a ability of corporate boards and managers, to longer period of diffuse ownership.' protect the rights of minority shareholders, to promote disclosure, and to ensure that second- ary trading is conducted at fair and transparent prices. Building these institutions requires time A t the end of the 1980s the best-known and resources in both cases. l vprivatizations were negotiated sales to stra- Where court enforcement of contracts is tegic investors-or privatization by issuin-g weak, these provisions should be supplemented shares-along the lines applied in Great Brit- by stock market regulation for financial inter- ain and a few other OECD countries. The cir- mediaries, such as investment funds and bro- cumstances in transition economies demanded kers, which can monitor compliance by other different approaches. For example, all transi- participants in financial markets. This would tion economies contained thousands of state- help set the stage for medium-size and large owned small business and service units, of a type privatizations in countries that still have a sig- unknown in the West. Most countries began nificant reform agenda. It would also improve their privatization by divesting these small 73 Transition-The First Ten Years: Analysis and Lessons for Easten Europe and the Former Soviet Union enterprises, usually by simple methods based minority of shares to insiders, but it was neces- on auctions. Small-enterprise privatization is ev- sary to settle for 51 percent preferential subscrip- erywhere regarded as a success. But as the size tion of share ownership by managers and work- of the enterprise increased, so did the complex- ers. When coupled with their vouchers, insiders ity of privatization solutions. ended up owning on average 66 percent of the At the beginning of the 1990s there was a shares in privatized enterprises. general perception of the advantages of a fairly Even now, enterprises subject to mass rapid ownership transfer. If speed was important, privatization in Russia have done little restruc- what could be done? Quick cash auctions, it was turing. A fair number of them were probably feared, would put most assets in the hands of unviable under any form of ownership. Many the old nomenklatura or the frontmen for for- were producing goods that people did not need eigners-with the possibility that the public or want, and some were using resources not would turn away from privatization and reform. readily shifted to the production of other items. Speedy privatization was also seen as a response However, it is also true that heavy insider control to pretransition attempts at enterprise reform, by managers in an environment of weak or inef- such as in Russia, that had empowered manag- fective shareholder oversight caused significant ers and allowed leasing, cooperatives, and other asset stripping. Delays in implementing a legal and arrangements often characterized as "spontane- governance framework, aggravated by opposition ous privatization." Rapid transfer of ownership from powerful insiders with a stake in weak cor- was seen as a way to stop unfair leakage of as- porate governance, led to poorly regulated capi- sets to managers. tal markets and weak enforcement of regulations Vouchers were seen as the answer, for they protecting investors' rights. These factors made would give purchasing power to the population the phase of insider control and diffuse owner- in a transparent and fair way. Secondary trad- ship very costly, raising questions about what ing, facilitated by investment funds, would al- might have happened with slower privatization low transparent consolidation of shares. Russia and a longer period of state control (box 7.1). and then Czechoslovakia applied voucher schemes universally, and Armenia, Azerbaijan, The Czech and Slovak Republics: Diverging Paths Georgia, the Kyrgyz Republic, Lithuania, and Moldova among others used them as the princi- The Czech and Slovak republics are an interest- pal method of divestiture. Vouchers were a sec- ing contrast to Hungary and Poland, neighbor- ondary privatization method in several other ing countries where vouchers were used on a countries (see table 7.1). much smaller scale (World Bank 1999a). The most advanced industrial economy in the former Russia: From Spontaneous Privatization to Diffuse Communist bloc, the former Czechoslovakia re- Ownership mained firmly in the grip of the Communist Party right up to the Velvet Revolution. There was little Russia was unique because of the large number price liberalization, little tolerance of small pri- of enterprises to be privatized (about 25,000 vate enterprise, and little experimentation with medium-size to large companies) and the strength workers' councils in enterprises. The fear of a of spontaneous privatization by enterprise man- reversion to communism was perhaps stronger agers after the dissolution of most branch minis- than in any neighboring country. tries under Secretary Gorbachev. Auctions had Economic reformers believed that nothing to be conducted across 89 regions over territory could be expected from a slow or evolutionary covering six time zones. A decentralized and sys- approach. The government apparatus could not tematic approach had to be developed at the cen- be turned into a force for positive change. A ter, with autonomous regional implementation. fast and massive transfer was needed to create The intent of the reformers was to award only a new owners who would support further 74 Privatization: Lessons and Agenda for the Future TABLE 7.1. Methods of Privatization of Medium-Sized and Large Enterprises Country Direct sales Vouchers Management-employee buyout CSB Albania n.a. Secondary Primary Bosnia and Herzegovina Secondary Primary n.a. Bulgaria Primary Secondary n.a. Croatia n.a. Secondary Primary Czech Republic Secondary Primary n.a. Estonia Primary Secondary n.a. Macedonia, FYR Secondary n.a. Primary Hungary Primary n.a. Secondary Latvia Primary Secondary n.a. Lithuania Secondary Primary n.a. Poland Primary n.a. Secondary Romania Secondary n.a. Primary Slovak Republic Primary Secondary n,a. Slovenia n.a. Secondary Prim,ary CIS Armenia n.a. Primary Secondary Azerbaijan Secondary Primary n.a. Belarus n.a. Secondary Prim,ary Georgia Secondary Primary n.a. Kazakhstan Primary Secondary n.a. Kyrgyz Republic n.a. Primary Secondary Moldova Secondary Primary n,a. Russia Secondary Primary n.a. Tajikistan Primary Secondary n.a. Turkmenistan Secondary n.a. Primary Ukraine Secondary n.a. Primary Uzbekistan Secondary n.a. PrimLary n.a. Not applicable. Source: European Bank for Reconstruction and Development data. market reforms. Vouchers thus became the cor- enterprise restructuring through active gover- nerstone of the reformers' approach. Investment nance (venture funds and limited partnerships). funds were created to consolidate the diffuse The Czech Republic implemented its mass ownership of vouchers and to diversify risk. voucher in two stages or "waves," the first un- Voucher holders could either bid directly for der the former Czechoslovak Federation and the enterprise shares or buy shares in investment second after the split. The government that ruled funds, which would then use the vouchers to the Slovak Republic shortly thereafter decided bid for enterprise shares. The funds were ex- to abandon the second wave of mass voucher pected both to manage a diversified portfolio privatization in the belief that the voucher pro- of shares (equity mutual funds) and to drive gram would not be conducive to sound 75 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Former Soviet Union Box 7.1. Historical Counterfactuals: Mass Privatization in Russia Russia's rapid and massive privatization, which began in the early 1990s, has been the center of significant criticism and debate. The major criticism is that privatization was implemented too quickly and withoutfirst putting in place the insttutional and legal framework that would have provided the checks and balances needed for proper govemance incentives and competition. The results, it is argued, have been significant stripping of assets, concentration of economic power that has prevented further reforms, loss of fiscal revenues, and worsening income inequality. Whatwould have happened under a slower privatization process-one that waited for strategic investors to emerge, at least for the medium- size and large enterprises, and for key legislative and enforcement capabilities to be in place, particularly in the area of corporate governance? What would have been the costs and benefits of having proceeded more slowly? At the outset, it is important to distinguish between two different privatization waves that were implemented in Russia since 1992: the mass voucher privatization program and the later loans-for-shares scheme for privatizing a small but highly productive number of very large enter- prises. This discussion refers to the voucher program. The second scheme is universally regarded as a poor choice of a privatization strategy. Russia's mass voucher privatization program was implemented to give managers and workers-'insiders7-the incentive to acquire majority ownership of their enterprises' shares in order to obtain their support for privatization, with managers eventually becoming the most important shareholders. However, because of a weak legal framework for corporate govemance plus major political uncertainties in Russia, managers may have had an incentive to maximize their short-term capital gains by selling assets for personal gain (rather than keep the enterprse as a going concern and maximize future profits) and thus decapitalizing the enterprise at the expense of other (smaller) shareholders. This may have had two negative effects: one concerning efficiency, the other, equity. First, there was a bias against keeping enterprises operating as a going concern. Second, regarding equity, wealth was redistributed from the rest of the population to the managers. Asset stripping may have in the end also increased the incentives to send those gains abroad-where they were less likely to be discovered and prompt legal action. Would a slower approach-keeping enterprises under the control of state and line ministries while legislation and enforcement were im- proved and strategic investors identified-have led to less asset-stripping and more satisfactory outcomes? Would there have been adequate incentives to develop legislation for decentralized private ownership in the absence of a minimum degree of experience with private property? In fact, the state did not have full control over its enterprises at the beginning of the 1990s. By 1988 'spontaneous' privatization acceler- ated as a result of legislation passed in 1987, allowing labor collectives and directors to become independent from the state and in practice receive the rights of owners. The Law on Cooperative Activities of 1988 allowed the formation of cooperatives (headed by the directors) within the enterprises. These cooperatives engaged in the most lucrative activities of the enterprises, while the liabilities remained with the state. As a matter of fact, this situation left enterprises in legal limbo and subject to soft budget constraints. The experience of a wide variety of countries, from Bulgaria to Romania to Ukraine, shows that keeping enterprises in that situation strongly encourages asset stripping as managers and workers remain, in the interim, uncertain about how much they will benefit from the eventual privatization of their enterprise in the future. That uncertainty is a strong incentive for quick asset stripping. This was aggravated by the collapse of state institutions potentially capable of exercising oversight as a result of the nature of the exit from communism in Russia. The alternative of giving insiders (workers and managers) a smaller share of ownership and thus encouraging the entry of outsiders was considered at that time, but faced strong political resistance from managers and line ministries. But even if it had been feasible, it may not have yielded better results in the particular circumstances of Russia. With weak minority shareholders' rights, the incentives of managers to exploit other shareholders is even higher when the latter hold a largershare of the enterprise, unless they are concentrated and can organize to prevent it. While too much share distribution to insiders may prevent strategic outsiders from coming in, too little may accelerate the incentives for asset stripping by these insiders. In the absence of encouragement for foreign strategic investors and with the collapse of state institutions, the choice between the mass privatization program and a longer period of state ownership, while developing institutions of corporate governance, is an open question. In summary, these and other historical questions can only be answered overthe long run in the context of judging Russia's overall transition experience, both economic and political. In contrast, most policy issues in this area today are more straightforward because of changed initial conditions and do not require a resolution of these questions. governance. It implemented the second wave The performance of Czech and Slovak enter- through direct sales. But the sales were not con- prises privatized through mass voucher methods ducted through open and transparent methods, was disappointing, for three reasons. First, few and they favored politically connected parties. investment funds had the resources and the skills 76 Privatization: Lessons and Agenda for the Future to drive enterprise restructuring. Second, little was compromise with the civil opposition, the ma- done to introduce an adequate regulatory frame- jority of which had never been underground. work governing enterprises, investment funds, and This more pragmatic Communist regime had capital market activities, particularly to protect largely abandoned central planning, allowing a minority shareholders' rights. Third, surveillance growing share of private ventures and speeding by creditors was weak because of delays in priva- reforms in taxation, banking, foreign trade, and tizing the largest banks and weak insolvency and corporate governance during the five years be- collateral legislation. Indeed, largely unreformed fore the political changes. commercial banks, operating in a climate of weak The new ruling elite in Hungary did not see a regulation, controlled many of the largest invest- return to communism as a threat. The problem ment funds. The funds also owned bank shares. lay with the managerial group that governed thou- This cross-ownership weakened the governance sands of more independent, but still formally state- structure and softened budget constraints further. owned, enterprises-and stripped their assets. All these problems resulted in significant "tun- Transferring ownership titles through mass neling"-extraction of assets by enterprise and privatization to the population while leaving gcv- fund managers (see box 3.2). ernance to these old managers would not be a so- In sum, the lack of appropriate accompanying lution. In addition, Hungary was the most hightly institutional policies and lagging banking sector indebted country in the region on a per capita ba- reform made mass privatization unnecessarily sis, and the government was obliged to think about costly in equity, transparency, and microeconomic generating cash from privatization. These factors efficiency. It eventually contributed to a large led Hungary to base its privatization on sales to buildup of contingent fiscal liabilities in the insuf- strategic investors and to open the process entirely ficiently reformed, state-dominated commercial to foreign investors. The full opening of the banks. By contrast, the best performers tended to privatization program to foreign capital was seen be enterprises and sectors sold to strategic inves- as a bold step that no other government (except tors through transparent methods, including one Estonia) felt able to take at that time. Foreign capi- of the earliest cases of a sale to foreign investors: tal into Hungarian enterprises and banks brought the Skoda-Volkswagen transaction. much needed investment, know-how, and compe- In the Slovak Republic governance of enter- tition-the main reasons for Hungary's good prises privatized by direct sales (during the second growth in the second half of the 1990s. wave) has not been significantly better than that of By the mid 1990s Hungary had made sub- those privatized by voucher methods (during the stantial progress in selling banks to reputable first wave), despite the more concentrated owner- foreign strategic investors and in adopting a strict ship, as judged by the overall poor performance of banking regulation law and a strict bankruptcy Slovak enterprises in the second half of the 1990s. law. Hungary had also worked out much of the The Slovak experience with direct sales provides a large stock of bad loans in the banking systern. sobering lesson and reinforces the conclusion that By contrast, this painful task had not even started concentrated ownership is a necessary, but not suf- in the Czech and Slovak republics and would ficient condition for effective ownership. A pro- gain momentum only in the late 1990s. gram of direct sales to concentrated owners that Hungary's economic performance in the second follows open and transparent methods is much half of the 1990s shows that creditor discipline more likely to produce positive outcomes than a matters as much as good ownership and gover- program of direct sales that favors politically con- nance structures in driving economic restructur- nected parties and has the potential for corruption. ing and microeconomic efficiency. Hungary: Bold Moves, Big Gains Poland: Diversify and Discipline In Hungary the rapidly reforming Communist As early as 1990 Poland planned to privatize a Party led democratization by making a historic mass of enterprises through vouchers. But politics 77 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Former Soviet Union delayed the scheme, not any calculated assessment investment funds as vehicles to achieve a con- by reformers that Poland's promising economic solidation of shares, requiring that all vouchers and institutional situation would allow them the be placed with funds. Moreover, before latitude to proceed through other privatization privatization, the government established sector options. From the beginning Poland pursued a rich holding companies that allowed the state to re- menu of privatization options, including the first tain a 39 percent residual share in all enterprises five flotations on the Warsaw Stock Exchange and being privatized. A legal framework governing the sale of assets and selected liabilities of some these funds was put together, but was then poorly 1,000 medium-sized enterprises through install- enforced. The upshot was a dilution of the value ment sales, a process described as "privatization of shares held by the initial holders of vouchers. through liquidation." Other variants included state and private enter- Poland's success in privatization and its good prise funds in Romania and divestiture by leas- growth performance in the second half of the ing to worker cooperatives in Ukraine. 1990s was also a result of other important comple- mentary reforms. Like Hungary, Poland intro- Were Vouchers a Mistake and Other duced hard budget constraints in the early stages "What Ifs" of the transition, the result of early efforts to re- structure and privatize the banks and address the C ach country pursued its own strategy; there bad loan problem. Poland reinforces the notion E was no homogenous approach driven by that creditor discipline matters as much for enter- blueprints. This diversity and considerable dis- prise restructuring and performance as effective appointment raise many questions about the ap- ownership and corporate governance. propriateness of strategies. The most fundamen- tal question has two parts: was mass voucher privatization a mistake? Could that have been Slovenia: "Internally Privatized" Enterprises Have Contributed Lttle to Gmwthforeseen? Would countries that went through mass voucher schemes, with disappointing re- Slovenia inherited the old Yugoslav concept of sults, have been better off keeping their enter- social capital and social ownership, with the state prises in state hands while trying to accelerate holding title to few industrial enterprises and economic reform and creating an institutional banks. Slovenia's privatization (and that of all and legal framework to attract reputable con- the parts of former Yugoslavia) reflects this start- centrated investors? ing point. The Slovene program allocated the This approach has been successful in China, majority of shares to state-owned institutional which has enjoyed sustained high growth rates investors (the pension fund and a development without, until recently, allowing much in the way fund) and to employees through several subsi- of formal transfer of state ownership. There is dized schemes. Few of the former socially owned considerable debate over how China achieved this. enterprises ended up in the hands of strategic Box 4.1 argues that initial conditions in China investors. The evidence suggests that the "inter- were different enough to allow growth to be fu- nally privatized" enterprises have not flourished, eled by a massive entry of new enterprises while though they seem to be doing slightly better than the state and subnational governments could the socially owned enterprises that were not monitor state enterprises to prevent egregious as- privatized. Most of the growth in industry and set stripping by managers or their allies. But most services is explained by new entrants, the recipi- of the CIS started the transition with a major po- ents of most foreign direct investment. litical transformation in parallel with an economic one. This weakened or eliminated many of the policy and disciplinary mechanisms required to replicate this approach, and the recentralization Several other CIS countries relied heavily on mass of political power needed to achieve it would not privatization through vouchers. Kazakhstan used have been feasible given the political trends. 78 Privatization: Lessons and Agenda for the Future Answering these broad questions requires a Summarizing Lessons historical counterfactual against which to com- .his experience suggests the following pare the mass privatization option (see box 7.1). 1agenda: If political centralization or recentralization was not an option, what mechanisms were available * Privatization should be part of an overall to improve governance of state-owned enter- strategy of discipline and encouragement. prises? Would hardening budget constraints- * Small enterprises under state ownership (gen- a key accompanying policy that influences en- erally enterprises with fewer than 50 employ- terprise performance-have been easier under ees) should be sold quickly and directly to state ownership? Could market and legal insti- new owners through an open and competi- tutions supporting property rights evolve in an tive auction, without restrictions on who may environment of state ownership? The cases of bid for the shares. Belarus (see box 4.3) and Uzbekistan do not Medium-size and large enterprises should support this notion. target sales to strategic outside investors. Only time will permit a fuller assessment With a concentrated, controlling interest, they of these fundamental questions. But many use- will have a clear stake to best use the enter- ful and practical lessons still emerge from more prises' assets. Although several transaction modest questions. For example, to what extent methods may be used, including negotiated can the outcomes of privatization be influenced sales, this can be brought about most effec- by the complementary policies generally under tively through competitive "case-by-case" the control of policymakers? In the former methods, more deliberative than voucher Czechoslovakia, mass voucher privatization schemes or rapid, small auctions. They use would have had a better chance of producing independent financial advisors who both pre- more restructuring and less corruption if the pare the enterprise for sale and act as sales legal framework governing companies, invest- agents on behalf of the state. ments funds, and capital market activities had * Investor protection should be enshrined in been sharply enhanced and enforced from the the legal system and enforced, covering rules very beginning. Earlier efforts at privatizing to protect minority shareholders; rules banks and strengthening creditor rights through against insider dealings and conflicts of in- improved insolvency and collateral legislation terest; creditor surveillance accounting, ala- might have aided overall restructuring as well. diting, and disclosure standards; and take- These reforms would have required intense- over, insolvency, and collateral legislation. but manageable-effort from the Czech and When court enforcement of contracts is Slovak authorities. Moreover, the efforts prob- weak, these provisions should be supple- ably would have succeeded in attracting larger mented by a stock market regulation for fi- flows of foreign capital, facilitating secondary nancial intermediaries, such as investment trading, and consolidating ownership by stra- funds and brokers, who then have an incen- tegic investors. tive to ensure compliance by other partici- In Kazakhstan the inability to vigorously en- pants in financial markets. This will set the force the legal framework governing investment stage for privatization of future medium-size funds was a major issue. Russia suffered similar and large enterprises. It will also improve weaknesses in the way it implemented the performance of existing privatized en- privatization. Managers and local officials try- terprises by assisting transparent consolida- ing to prevent competition dominated many of tion of shares where ownership is diffuse. It the auctions to voucher holders. A weak legal will also facilitate bank- and nonbank-based framework to protect minority shareholders in- financial intermediation. hibited outsider participation in secondary trad- * Privatization should be accompanied by in- ing, guaranteeing that the managers would con- creasing competition in the market for the tinue to control the enterprises. products sold by the enterprise in question 79 Transition-The First Ten Years: Analysis and Lessons for Eastem Europe and the Former Soviet Union and vigorously enforced by the competition to improve the legal framework, assist a trans- policy authority. This can discipline manag- parent consolidation of shares, and develop in- ers in an environment where corporate gov- stitutions to help monitor managerial behavior. ernance is weak. The remaining privatization agenda must rely on * The cash flow and property rights of the transparent and competitive case-by-case state should be clarified for enterprises in privatization, including the transfer of residual which the state continues to hold an owner- state ownership. ship stake. In countries such as Belarus, Turkmenistan, * Divesting enterprises in sectors characterized and Uzbekistan, where most medium-size and by natural monopoly or oligopoly (where av- large enterprises remain in state hands, a ma- erage production costs decline continuously as jor challenge remains. In Belarus and scale increases and the market and society are Uzbekistan, as in China, strong central con- best served by one or a few enterprises) must trols have prevented egregious asset stripping proceed with great caution, if at all. Advances in state-owned enterprises. However, there has in technology have made such sectors increas- been little restructuring because these enter- ingly rare. But where they exist-as in local prises have been protected through the trade distribution of natural gas-an efficient regu- regime and the special allocation of foreign ex- latory regime that protects the public interest change and credit at subsidized prices. Foreign is a prerequisite, lest divestiture transform an direct investment and entry of new enterprises inefficient public monopoly into a poorly regu- remain heavily controlled. The challenges in lated or nonregulated private monopoly. these countries, when they decide to reform, will be to liberalize, encourage new entry, and In Central and Eastern Europe and the Baltics assemble a legal framework (for private prop- (except the Federal Republic of Yugoslavia), SMEs erty and contract enforcement) and market in- have been substantially privatized. Voucher stitutions to monitor managers. Small enter- privatization or competitive cash auctions are not prises should be quickly privatized through relevant options. Bulgaria, Croatia, and Roma- competitive auctions, and medium-size enter- nia-with large commercial enterprises still to be prises through case-by-case methods, with a privatized-should implement (and, indeed, are majority share sold to outsiders. Very large en- implementing) transparent and competitive case- terprises should be privatized only when a clear by-case methods to mobilize strategic investors. strategic investor is identified. Progress in the European Union accession will make this more likely to succeed. The Federal Note Republic of Yugoslavia, now privatizing medium- size enterprises through competitive auctions, will t. Capital market regulations in most OECD coun- start the privatization of large enterprises through tries require that a major shareholder or group of con- case,by-case methods. A major positive feature nected shareholders acquiring 30 percent of the shares case-by-case methods. A major positive feature in a company offer to buy out minority shareholders of these methods is to restrict the distribution of at the assessed market value. Full implementation of subsidized shares to insiders to a maximum of 30 this regulation alone would have considerably slowed percent of shares, thus making the majority of consolidation of ownership in the Czech Republic in shares available for strategic investors, the 1990s. Other regulations restricting insider deals and transactions with connected parties, enhancing In such CIS countries as Armenia, Georgia, disclosure, and ensuring an integrated pricing mecha- the Kyrgyz Republic, Russia, and Ukraine, and nism would also have slowed the consolidation of where privatization has been important during ownership. For a review of progress with privatization, the past decade, the most critical steps now are see EBRD (2000). 80 Supportive Social Policies H ow can social policies support a strategy of discipline and encouragement? By targeting social safety nets to the most vulnerable, such as those affected by the increase in utility prices and by the labor shedding resulting from hard budget constraints on enterprises. By helping local governments take over divested social services previously provided by enterprises, such as housing, kindergartens, and clinics, to permit enterprise restructuring to go ahead. And by reforming expenditures on education and health to allow workers to acquire skills more adapted to new market realities and, more generally, to ensure that the benefits of growth, once it resumes, are widely shared. The loss of fiscal control accompanying the transition and the need to reduce the fiscal deficit to stabilize inflation reduced government expenditures as a share of GDP everywhere. In Central Eu- rope and the Baltics, which are farthest along in the transition, an important policy priority now is to restructure social sector expenditures to make them fiscally more affordable. In the low-income CIS countries, where the resumption of sustained growth has proved elusive, the priority is to pro- vide a social safety net for the most vulnerable. Meeting those objectives is a challenge. For the advanced reformers of Central Europe and the Baltics the ratio of consolidated public expenditures to GDP is around 45 percent, comparable to those in the high-income countries of Western Europe (World Bank staff estimates). To reduce the tax burden on the private sector while confronting the new costs of complying with European Union directives, these countries will need to improve their efficiency in social service provision and modify pension systems to reduce their fiscal costs. In the Caucasus and Central Asia public expenditure ranges between 20 and 25 percent of GDP, approaching 30 percent in the Kyrgyz Republic. In these countries public sector revenues were lower than in the rest of the CIS. Several of them, facing unsustainable public debt, need an international workout to reduce debt service and domestic efforts to increase tax revenue to maintain social expenditures. Refonning Pension Systems Spending on social insurance programs, which cover pensions and unemployment insurance, ac- counts for about 10 percent of GDP in the CSB. In Croatia, Poland, the Slovak Republic, and 81 Transition-The First Ten Years: Analysis and Lessons for Eastem Europe and the Former Soviet Union Slovenia public pension expenditures have climbed country. In Central and Eastern Europe it ranges to more than 13 percent of GDP-twice the from 120 to 250 percent of GDP.) There are three pretransition level and not sustainable. The fiscal possible sources of financing. The first is to limit burden of pension systems increased in the tran- first pillar expenditures to create savings in the sition partly because benefits eased the social costs state pension schemes and reduce future liabili- in its early years. Indeed, in most CSB countries ties. The second is to increase payroll taxes and the incidence of poverty for households headed contributions. The third is to use fiscal resources by pensioners is much lower than for other socio- from debt instruments or privatization proceeds. economic groups. However, the aging of the popu- Given the magnitude of the financing needs and lation and the rising payroll taxes needed to fi- the fact that governments may have largely ex- nance more generous benefits undermine the hausted the first two sources, the resources for financial viability of pension systems and threaten paying obligations to current pensioners will have employment creation. The challenge is to make to come from the budget. pension systems more sustainable without under- The up-front fiscal costs of moving to a mining the socially desirable goal of providing multipillar system are large. In reforming countries adequate income for retired workers. a second pillar financed by a contribution rate of 6 Large pay-as-you-go pension systems, with percent of gross wages, as in Hungary, would re- their attendant large unfunded liabilities, burden quire resources equal to around 1.9 percent of GDP public finances and savings in the long term. annually in the first years. Other countries prepar- Higher contribution rates by themselves do not ing to establish such a scheme-Bulgaria, Croatia, solve the problem because they raise labor costs, Estonia, and FYR Macedonia-will also need to shift incentives toward informal market employ- plan and budget for these costs in ways consistent ment, and undermine job creation, particularly with preserving macroeconomic stability. The size by new enterprises. In addition, the opportunities of the fiscal costs depends on the design of the sec- for using contractual savings for capital market ond pillar-its size and choice of cohorts-and on development are lost. Most countries have begun the pace that governments move along this path. to reform their pension systems by tightening the However, the institutional requirements of intro- link between contributions and benefits, shifting ducing multipillar systems are demanding. Finan- to notionally defined contribution schemes, rais- cial markets have to be adequately developed, and ing retirement ages, reducing replacement rates, governments have to regulate and supervise funds. and changing pension-indexation formulas. In The CIS typically spends about half what the addition, several countries in Central and East- CSB spends on pensions. But in many cases even ern Europe are taking steps to implement that spending is fiscally unsustainable, given poor multipillar pension schemes (table 8.1). tax compliance and high ratios of dependents to A multipillar pension system allows people to contributors. Pension systems in the CIS have gen- diversify risk across countries, regions, and assets. erally done a much poorer job of keeping the eld- It provides for a portion of the mandatory contri- erly out of poverty. They need to undertake the butions to the public pension system to fund the basic reforms to put their pay-as-you-go systems individual accounts of each worker. These contri- on sounder financial footing. Some low-income butions are managed and invested by private insti- countries-such as Georgia, where resources are tutions, and pensions are paid according to a de- constrained and tax compliance especially poor- fined contribution. The eventual pension system may need to move to a flat benefit structure to comprises a downsized pay-as-you-go scheme (the protect the poorest elderly until fiscal conditions first pillar), a benefit from a fully funded scheme permit a move to a more differentiated structure. (the second), and personal savings (the third). Implementation of these reforms will face po- A key element of these reforms is making ex- litical opposition and administrative weakness. plicit at least part of the implicit debt of pension For example, flat benefits to the most needy may systems, which then has to be financed. (The size not enjoy broad-based support, as Georgia's re- of the full implicit debt varies from country to cent experience showed. Coverage needs to be 82 Supportive Social Policies TABLE 8.1. Refonn Options for Social Protection Programs Social insurance Income level Social assistance Unemployment Pensions Fiscal implications Higher income * Means-tested cash Insurance Multipillar system Reduced fiscal burden (Central Europe and benefit assistance with minimum of pay-as-you-go the Baltics) program, possibly poverty-based benefit system, but with supplemented by up-front fiscal cost indicator targeting Deinstitutionalization Middle to low income * Categorical cash Flat benefit or Reformed pay-as-you- Reduced fiscal burden (Bulgaria, Romania, the benefit (universal or severance go system, with of pay-as-you-go Russian Federation, targeted by category; minimum poverty- reallocated toward and Ukraine) means tested only based benefit targeted social where local assistance institutions are strong) * Lifeline utility tariffs * Deinstitutionalization Very low income (Caucuses and low- . Limited cash benefit, Flat benefit and Flat benefit Increase fiscal income Central probably based on severance allocations for all Asian countries) geographic targeting, community targeting, or indicator targeting * Lifeline utility tariffs * Self-targeting (workfare schemes) * Deinstitutionalization Source: Authors. broad enough to ensure that schemes are con- diminished considerably. In addition, the real in- tinued and adequately funded. Making the de- comes of households fell. Although the Central sign and implementation of reforms more diffi- European countries have spent the bulk of their cult are weak actuarial capacity to forecast social protection budgets on pensions, these coun- expenditures and revenues, a lack of auditing of tries allocated somewhat more resources to so- pension funds, and problems in collection. As cial assistance spending to address rising poverty. wages begin to rise, tax collection improves, and The CIS generally spent less on social assistance, institutional capacity develops, countries can while indirectly protecting employment by not im- move toward a multipillar pension system. posing hard budget constraints on old enterprises and by maintaining subsidies on housing and utili- Social Assistance Should Protect Children ties. The subsidies on utilities have most often been and the Most Destitute, Adding More as met by forcing the utilities to defer maintenance, Budgets Allow or by expecting enterprises to finance them even where housing has been divested. Despite the ex- W lith the transition to market, the guaran- pansion of social assistance, most programs, in- VV teed employment, retirement security, and cluding the subsidies on utilities, have not reached consumer subsidies of the socialist system great numbers of the poor. 83 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Former Soviet Union The key objectives for social assistance pro- less prone to error in these economies, to the grams-and probably all that the poorest coun- extent that the tests are more formalized. Un- tries can afford-are to help the most destitute employment insurance may also be more fea- and to ensure that children's mental and physical sible in these countries, with structural unem- development is not impaired. In addition to en- ployment falling and capacity to implement a suring that young children receive essential pre- true unemployment insurance program grow- ventive health care and can afford to attend school, ing. But more generous provision of benefits many countries also face the challenge of offering carries the risk of creating a culture of depen- alternatives to institutions for elderly people un- dency and reducing incentives of workers to find able to live on their own; adults with physical and employment-added, of course, to the ever- mental disabilities; and children disadvantaged by present problem of becoming a major fiscal bur- poverty, ethnicity, or disability. Some countries are den and a drag on growth. introducing community-based services such as home and day care and special educational pro- Severe Cuts Have Compromised the Quality of grams for children with disabilities, adapting ap- proaches that have been successful in Western Education Europe and the United States. ome of the fiscal adjustment in the CIS has In addition, to facilitate the enterprise restruc- 3 been at the cost of severe cuts in education, turing critical to sustained growth, countries may inevitable given the plunge in GDP for all the wish to offer some protection to those hurt by CIS countries (World Bank 2000b). Public restructuring. The best way of doing so is to re- spending on education ranges from less than 2 move barriers to entry of new enterprises. Assis- percent of GDP for Armenia and Georgia to tance should be targeted to workers whose skills almost 8 percent of GDP for Uzbekistan. The and experience make them the least likely to be average for OECD countries-with 10 times the employed by new enterprises. GDP per capita-is about 5 percent of GDP. For low-income countries with limited re- Several countries have maintained reasonable sources and pervasive poverty, targeting cash ben- spending on education relative to their GDP. efits-with a possible increase in the resource en- The challenge is to ensure that resources are velope in some countries in the Caucasus and allocated to best use-across levels, between Central Asia, where spending is low-may have wage and nonwage expenditures, and across to be improved through geographical targeting, regions within each country. But in the poorest community-based identification, or other indica- countries of the Caucasus and Central Asia tors. Given the pervasiveness of the informal spending cuts have compromised the ability of economy and self-employment, targeting will con- the education system to prepare students for the tinue to be imperfect. There may be some scope emerging market economy (figure 8.1). for self-targeting through some form of public The cuts have affected opportunities, access, works scheme. There also is considerable scope and coverage (World Bank 2000b). Falling edu- for improving the targeting of utility and housing cation budgets and protection of employment in subsidies. Rather than offer across-the- board the sector are squeezing expenditures on text- price subsidies, some countries have introduced books, school supplies, and school maintenance. "lifeline" tariffs for utility services with metered In addition, poorer parts of many countries bear consumption, in which the price subsidy is re- a disproportionate share of the adjustment. In stricted to the initial block of consumption, called Georgia 43 percent of primary and secondary the basic need level. Lifeline tariffs are easier to schools in urban areas got textbooks for all chil- administer than income-tested schemes and are dren, compared with 27 percent in poorer rural less distortionary than generalized price subsidies. areas. In the Russian Federation richer regions Higher-income countries may be able to af- have spent more on education, while poorer re- ford a more generous safety net. Targeting gions have struggled to maintain the basic re- through some forms of means testing may be quirements (World Bank 1999b). 84 Supportive Social Policies By not spending enough to meet the basic needs of the school system, the state effectively FIGURE 8.1. shifts educational costs to families. Families con- Public Expenditures on Education in Transition tribute to school maintenance and education Economies, 1998 supplies. In some countries they are asked to supplement the salaries of education personnel Albania through under-the-table payments and various Armenia tutoring schemes-if they want their children to Azerbaijan pass. At the same time many countries have been Belarus slow to implement cost recovery measures for Bulgaria higher education, which tends to favor better- Croatia off families. Czech Republic Poorer countries need to take three sets of Estonia measures. The first is to ensure universal basic Macedonia, FYR education, of particular importance to Armenia, Georgia Azerbaijan, Georgia, and Tajikistan, where bud- Hungary getary resources for basic education do not pro- Kazakhstan vide universal coverage. The second is to ensure Kyrgyz Republic that expansion of tertiary education does not Latvia come at the expense of basic secondary educa- Lithuania tion. In some countries higher education enroll- Moldova ments are growing-appropriately reflecting the Poland , . , , , , , . ~~~~~~~~Romania _ human capital demands of these emerging econo- Rrai mies. This risks absorbing larger shares of edu- Russian Federation cation budgets, unless greater cost recovery, Slovak Repuic lower unit costs, and more encouragement of Turkmenistan private financing and delivery of higher educa- Ukrainek_ tion services follow. Uzbekistan The third set of measures is to shift resources from personnel (and energy) into repairing 0 20 40 60 80 100 schools and providing adequate educational Percentage of GDP material. As materials and maintenance expen- ditures were cut and student enrollments fell, the Source: World Bank data. number of teachers rose sharply. In Russia every region increased its number of teachers, for an overall increase of 25 percent between 1989 and TABLE 8.2. 1996 (World Bank 1999b). In Central Asia in- creases ranged up to 25 percent (Klugman 1999). Student-Teacher Ratios in Basic Education, 1990 and Student-teacher ratios are typically low for most 1997 of these countries and can be increased without (percent) compromising teaching quality or learning out- comes (table 8.2). Countries will need to ratio- Countries 1990 1997 nalize and consolidate their school infrastructure, Armenia 11.7 8.7 which is often overdimensioned due to falling Azerbaijan 10.5 9.9 birth rates and extremely low student-teacher Belarus 11.8 10.5 ratios. School consolidation would also enable Russian Federation 14.0 11.9 countries to reduce wage costs. Using OECD Turkmenisran 14.0 13.4 standards as a benchmark, up to a third of the teaching labor force can be reduced. That would Source: UNICEF-ICDC TransMONEE (1999). 85 Transition-The First Ten Years: Analysis and Lessons for Eastem Europe and the Former Soviet Union free budgetary resources to better compensate FIGURE 8.2. remaining teachers, to repair schools, and to pro- Health Expenditures, 1998 vide instructional materials. In addition, energy costs for education need Germany to be reduced. In some CIS countries energy ex- Croatiaa penditures consume between 30 and 50 percent Lithuania of education expenditures, reflecting the fast rise Sweden in energy prices in the past 10 years. In the me- Francea dium to long run, energy savings will be realized Slovenia as new schools replace old ones. Short-term mea- United Statesa sures-insulating school walls, double-glazing Czech Republic windows, and installing meters-can also help Canada reduce energy bills, but these will need up-front Netherlands investments from the budget. United Kingdom Finland Spain Containing Costs Will Make Health Care Macedonia, FYR Affordable for Those Who Need It Most Slovak Republic he public health achievements of the socialist Estonia T era are being undermined. Many countries Belarus _ are not spending enough on public health Moldovaa measures to confront the growing threat of HIV/ Portugal AIDS and drug-resistant tuberculosis. Russia Inadequate funding has led to under-the-table Poland payments. Unpaid or underpaid health workers Ukraine tap patients for side payments, including pay- Latvian ments for drugs and other items in short supply. Turkmenistana This is financially onerous for poor families who Uzbekistana have to sell assets or borrow to finance their Bulgaria health care. In the Kyrgyz Republic a third of all Armenia patients seeking inpatient care had to borrow Romaniab money. Similar figures are reported for Georgia, Albaniaa Tajikistan, and Ukraine. Kyrgyz Republic Often the inability to pay means that pa- Kazakhstan tients forgo health care. In Russia data for 1997 Tajikistanc suggest that 41 percent of all Russian patients Azerbaijana could not afford to purchase required drugs, Georgia I , I , and 11 percent could not afford any kind of 0 2 4 6 8 10 medical treatment (World Bank 2000b). Simi- larly, 37 percent of pregnant women in Percentage of GDP Tajikistan did not seek prenatal care because it was unaffordable, and almost a third of births a. Data are for 1997. b. Data are for 1996. occurred at home, a break from past practices c. Data are for 1999. of hospital births. In several countries a pri- c.~~~~~ ~ Dat areel forned 1999.te yse o eat Source: The World Bank's World Development Indicators vately financed, unregulated system of health database. care is in a public shell. In some countries, notably in the Caucasus and Central Asia, public expenditures are inadequate (figure 8.2). However, other countries are 86 Supportive Social Policies spending significant public resources without get- facilities and reallocation of savings on comple- ting the benefits of quality health care. The chal- mentary inputs. To reallocate expenditures most lenges facing most transition economies are simi- effectively, however, governments need to decide lar to those facing education: the rationalization what kind of health care system they want, in- of systems with an excess of personnel and cluding who provides what and who pays for what. 87 Part 3, The Political Economy of Discipline and Encouragement The Winners and Losers; from Discipline andl Encouragement W hy have some governments been able to enact policies to discipline the state sector and encourage new enterprises to create a foundation for sustainable growth? Why have oth- N lers maintained far less effective strategies of protecting inefficient enterprises and discour- aging new entry at considerable social cost? Can these government policy choices be systematically related to particular institutional characteristics of political systems in transition? Part 3 will try to answer these questions. To develop a better understanding of the political economy of reform in transition economies, we extend the framework developed in part 2 to examine the political dynamics of discipline and encour- agement. The economic reforms associated with discipline entail costs for the public in the short term. Imposing discipline on state enterprises requires substantial adjustment to correct decades of ineffi- cient investments and distorted policies. Such adjustment inevitably generates losers in the short terrn as a result of higher unemployment, higher prices, and lower provision of subsidized social services by enterprises. In contrast, the gains associated with the policies of encouragement accrue primarily over the long term as the institutions needed to promote entry and encourage competition-secure prop- erty rights, vigilant contract enforcement, good access to financing-cannot be created overnight. So, for the public, programs of comprehensive economic reform tend to bring substantial adjustment costs from discipline in the short term for the mere promise of future gains as encouragement promotes greater investment and growth. How can governments assure the public that those future gains will materialize? What guarantees that future governments will not backtrack on reforms before the promrr- ised gains materialize? What if the promised gains are directed to groups closely tied to existing elites and not widely distributed across society, as is common in so many highly inegalitarian countries? The politics of economic reform thus begins with a paradox: people are asked to support policies that impose clear costs in the short term for the promise of long-term gains inherently subject to high risks. To win broad support, the government must convince the public that the government will be able to sustain its commitment to reform to traverse what one author has called the "valley of transi- tion" until the "higher hills" of an efficiently functioning market economy are reached (Przeworski 1991). Recognizing the different time horizons of costs and benefits and the related risks that might 91 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Fonner Soviet Union lead rational individuals to discount those ben- limit insiders' ability to tunnel assets abroad. efits at the start of reform is central to the politi- New banks created in the liberalization of finan- cal economy of market-oriented transitions. The cial markets have fought to keep soft govern- government must be able to credibly commit to ment credits for their enterprise clients, which short-term losers that they will reap the gains could be recycled in volatile local bond and for- from reforms over the long term. eign exchange markets. Oil and gas exporters At the same time, a decade of transition ex- who gained from the opening of foreign trade in perience has shown that some groups realize sub- partially liberalized markets have struggled to stantial gains from the no man's land between build entry barriers to prevent competition from the command system and the market economy dissipating their rents. The ability of these groups (Hellman 1998). Uncertainty about property to preserve their extraordinary gains is based on rights before privatization allows insider man- their capacity to influence the political process, agers to tunnel assets from nominally state- and in the most extreme cases, to capture key owned enterprise to newly created spinoffs un- institutions of the state, highlighting the critical der managers' personal control. Partially interaction between politics and economics in liberalized prices spur arbitrage opportunities transition (Hellman, Jones, and Kaufmann 2000; between fixed-price and market-price sectors that World Bank 2000c). can generate enormous gains. Incomplete trade liberalization creates highly profitable monopoly Who Wins and Who Loses? rents, especially in economies rich in natural re- sources. These opportunities to take advantage he complex dynamics of the political of the distortions of a partially reformed T economyofreformcanbeexpressedgraphi- economy are available only to the select few, cally by tracing the paths of winners and losers namely those with control over nominally state- with respect to different levels of reform in the owned assets and those with close ties to politi- discipline and encouragement framework. A typi- cians able to award such advantages. As a re- cal economy at the start of the transition can be sult, these gains are highly concentrated. divided into three basic constituencies: In theory such gains should be short-term, because as transition progresses many of these * State sector workers. These are workers from temporary economic distortions will be eliminated state-owned enterprises without the skills to or competition will arise to dissipate the rents, become new entrants in the competitive mar- But experience shows that these short-term win- ket. They face significant losses initially be- ners of partial reform can convert a small share cause of discipline (unemployment, price in- of their gains into political influence that can be creases) and are unlikely to realize any gains used to restrict entry, undermine competition, and from encouragement. preserve the very distortions that generate these * Potential new entrants. These are workers rents. Such constituencies seek to freeze reform and new entrepreneurs, originally from into an equilibrium of liberalization without dis- state-owned enterprises, who have the skills cipline and selective encouragement, producing a to become new entrants in the competitive highly unequal pattern of costs and benefits of market. They face initial losses from disci- market-oriented transition over the long term. pline as they adjust to the decline in the state Examples of how the short-term winners of sector However, they are likely to see gains partial reform prevent further reforms that would from entry into the market if encouragement impose discipline and encourage new market is effectively implemented and sustained. entry and competition are well known through- * Insiders and oligarchs. These are actors who out the region. Enterprise insiders who gained begin the transition with substantial de facto minority stakes in privatized enterprises have control rights over state assets and close ties opposed improvements in corporate governance to the political elite inherited from the and the security of property rights that would previous command system. Insiders and olh- 92 The Winners and Losers from Discipline and Encouragement garchs benefit immediately from liberaliza- to promote and support new entry into the com- tion and privatization because they can con- petitive market. The oligarchs and insiders, by vert their existing control over state assets contrast, face an inverted U-curve income pat- into substantial gains. Moreover, insiders tern. Their concentrated gains in the early stages and oligarchs can reap further gains from of reform-associated with opportunities for ar- rent seeking, arbitrage, and asset stripping bitrage, rent seeking, and tunneling-are dissi- if liberalization and privatization are not pated as further reforms increase competition combined with discipline and encourage- and market entry. ment. As discipline is imposed and further For all these constituencies, gains and losses reforms encourage competition from new depend on how radical the first move in the re- entrants and the rule of law, these initial form process is at the start of the transition. gains are dissipated unless new barriers to The more radical this move, the greater the ini- entry can be erected. tial adjustment costs to both state sector work- ers and potential new entrants. Yet new entrants Figure 9.1 presents a stylized depiction of should begin to see greater gains at an earlier the income gains and losses for each of these point in the transition if such reforms lead to three groups under different levels of economic rapid development of the institutions that en- reform. The state sector workers face a drop in courage entry and competition. For the oli- income as a result of sector downsizing, with garchs and insiders, radical reforms generate little hope of substantial recovery as reform fewer distortions and imbalances for them to advances. The potential new entrants face a clas- extract rents and strip assets. So such reforms sic J-curve pattern of income, with adjustment reduce the high concentration of initial gains to costs at low levels of reform as they exit the state the oligarchs and insiders. sector and gains realized only once enough By contrast, less radical reform programs-- progress has been made in institutional reforms liberalization and privatization with weak disci- FIGURE 9.1. Winners and Losers from Reform Income gains New entrants Oligarchs and insiders O < R = Extent of reforms State sector workers Note: Ro = no reforms; R1 = point at which income gains of oligarchs and insiders are maximized; R2 = level of reforms that allows the winners of reforms beyond R, (new entrants) to compensate for or exercise enough political pressure to neutralize the resistance of oligarchs, insiders, and state sector workers. Source: Authors. 93 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Former Soviet Union pline and minimal measures to support compe- highly susceptible to state capture cause poten- tition-have the opposite effect. Such partial re- tial new entrants at the outset of transition to form programs at the start of the transition gen- substantially discount the potential gains from erate lower initial adjustment costs for both state any proposed radical economic reforms, leading sector workers and potential new entrants, as them to support partial reforms that offer lower state sector downsizing is limited and the flow initial costs-even though these partial reforms of subsidies continues. Oligarchs and insiders are more likely to lead eventually to barriers to enjoy the highest gains from liberalization and entry. privatization without discipline and encourage- The risk of ending up at a low level of re- ment. form (R1) is indeed high. Such partial reforms- Given these patterns of gains and losses, each liberalization without discipline and with lim- constituency prefers a different combination of ited encouragement of new entry-are the result reforms. State sector workers prefer the status of the joint pressure of oligarchs and state work- quo (Ro) and reject all reforms. Oligarchs and ers to prevent further reforms while the gains by insiders prefer to begin with a partial reform and new entrants are not high enough to allow them sustain the reform process through Rl, the point to exercise enough pressure for more compre- where their gains are maximized and beyond hensive return. Only a minimum commitment which implementation of policies of discipline of advancing reforms to at least R2 will generate and encouragement threatens to undermine gains enough support-where R2 is the level of reform from rent seeking and tunneling. For new en- that allows the winners of reforms beyond R1 trants, the process offers sacrifices in the begin- (new entrants) to compensate for or exercise ning for the promise of gains when the reforms enough political pressure to neutralize the resis- are advanced enough to create an environment tance of oligarchs, insiders, and state workers. conducive to new entry and competition. The political problem posed by initial win- For a government to secure the support of ners from the transition was not foreseen in the the potential new entrants for radical reforms at early literature on transition. Few recognized that the start of the transition, the government must oligarchs and insiders would be able to stall the be able to make a credible commitment that the transition in a state of partial reforms. Yet this reforms will be continued until at least R2. But has proven to be one of the most serious ob- the credibility of that commitment will depend stacles to economic reform, particularly in many on the strength of the oligarchs and insiders, who CIS countries. have an incentive to invest a share of their initial gains in capturing the state to stop the reform The Government Must Be Credible and Able to process at R1. Thus at the start of transition, the greater the risk that oligarchs and insiders will C capture the state in the future, the less likely that p artial reform-liberalization without disci- potential new entrants will support a radical re- P pline and selective encouragement-can be form program at the outset. a stable equilibrium of reform if government, Where the risk of capture by oligarchs and at the outset of the transition, lacks the cred- insiders is high, potential new entrants and state ibility to promise that narrow groups with close sector workers have an incentive to reject reforms links to the political elite will not capture the or to accept partial reform programs that reduce reform process. Public support for radical re- the initial adjustment costs. Yet it is precisely such forms, therefore, depends on perceptions of partial reforms that make state capture by oli- government credibility. garchs and insiders a self-fulfilling prophecy, as By recognizing that different combinations these reforms maximize opportunities for rent of reforms produce different configurations of seeking and theft. This has led to a so-called par- winners and losers, the discipline and encour- tial reform paradox in many transition econo- agement framework suggests two political chal- mies: governments that lack credibility and are lenges in promoting economic reform: 94 The Winners and Losers from Discipline and Encouragement * Securing up-front support from potential new greater risk of being captured by small, power- entrants for comprehensive reforms ful interests. In such systems politicians are held * Preventing early winners of liberalization and accountable to a narrower range of groups, in- privatization from trying to sustain a partially creasing the probability that those with access reformed equilibrium of rent seeking and to political power will expropriate or concen- theft by undermining further reforms. trate the gains from government policymaking. To meet these challenges, governments must To prevent the early winners from holding the have the capacity to project credibility to poten- economy in a partially reformed equilibrium, the tial new entrants and constrain the oligarchs and political system must constrain the power of any insiders. narrow group to capture the state. Expanding What influences the capacity of governments the range of social groups competing for influ- in transition economies to make credible com- ence over policymaking increases the costs to mitments that the promised gains from reform politicians of skewing reform in the interests cf will not be expropriated by early winners of re- a single group. So the government's capacity to form? Credibility and constraint are rooted in make a credible commitment to a wide range cf the nature of political institutions, which are groups is much greater in political systems that shaped by the cultural and historical legacies that promote competition and contestability. guided the exit from communism. Political in- These differences in political systems are stitutions designed with participation by com- shaped by a broad range of factors often loosely peting groups to foster political contestability referred to as "initial conditions," which incor- within an agreed set of rules are less likely to be porate historical, cultural, geographical, and captured by a corrupt political elite or narrow other variables that shape the transition path set of interest groups. The rigors of political (de Melo and others 1997). But understanding competition increase the costs to politicians of how different political systems shape the con- pleasing narrow constituencies and increase the figuration and choices of the winners and los- likelihood that broader interests will be repre- ers in reform provides important insights into sented in government over time. the relationship between democracy and In contrast, political systems designed to con- market-oriented reform. centrate power and restrict contestability are at 95 Classifying Political Systems in Transition T 10o determine the effects of different types of political systems on the capacity of governments to adopt and sustain economic reforms, we first differentiate the range of political systems across the transition economies. One of the most important features of this variation has been thle political contestability in the new regimes, that is, the extent to which key decisions of the political process-such as choosing political leaders, adopting laws, and making binding policy decisions-are subject to challenge by freely organized groups in and outside government. Political contestability can be determined along several different dimensions: * Political rights and civil liberties. The ability to challenge political decisions requires the rights to participate freely in the political process and to express one's views. The extent of such rights can be measured by indicators of political rights and civil liberties. * Veto points. The clearest institutional manifestation of political contestability is the right to veto political decisions. Different types of political systems have different numbers of "veto points.," that is, institutional actors who can veto political decisions. * Government turnover. Political contestability can also affect the turnover and tenure of govern- ments. Frequent government turnovers suggest a high degree of political contestability and shape the perceived competitive pressures on incumbent governments. Of course, excessively high gov- ernment turnover could be a reflection of underlying political instability. * War and political violence. The outbreak of war or political violence, often with ethnic or regional cleavages, indicates extreme contestability over the boundaries and basic organization of the po- litical process. Developing exact measures of these characteristics for transition economies is difficult given their rapid change. But a combination of indicators allows transition economies to be classified into four "ideal types" based on the extent of political contestability. Freedom House (various years) provides a widely accepted system of annual ratings of political and civil liberties (figure 10.1). On the basis of these ratings, the following country groups can be developed: 97 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Fonner Soviet Union FIGURE 10.1. Classifying Political Systems in Transition Economies, 1990-99 Czech Republic Slovenia Hungary Poland Competitive democracies Lithuania Estonia Latvia Average Slovak Republic Bulgaria Romania Ukraine Russia Concentrated political regimes Croatia Moldova Kyrgyz Republic Average Macedonia, FYR Armenia Albaniaa Georgia War-torn regimes Azerbaijan Tajikistan Average _................. .................................................... Belarus Kazakhstan Uzbekistan Noncompetitive political regimes Turkmenistan Average 0 2 4 6 8 Average rating of political and civil liberties Note: Ratings are based on the average scores for political rights and civil liberties ranging from 1 (free) to 7 (not free) by Freedom House from 1990 to 1999. The thresholds for determining the country groups are: competitive democracies: political rights < 2.0 and civil liberties < 2.5; concentrated political regimes: political rights or civil liberties > 2.5; noncompetitive political regimes; political rights or civil liberties > 5.0. Source: Freedom House (various years). Competitive democracies have from the start Estonia, Hungary, Latvia, Lithuania, Poland, of transition maintained both a high level of and Slovenia. political rights to compete in multiparty demo- * Concentrated political regimes conduct mul- cratic elections and an extensive range of civil tiparty elections, but for some period of the liberties. They include the Czech Republic, transition they have either curtailed full rights 98 Classifying Political Systems in Transition to participate in those elections or otherwise economies straddle different categories. Bulgaria limited political competition through con- and the Slovak Republic lie on the border between straints on civil liberties. The result has been a competitive and concentrated political regimes. concentration of political power, often in the Azerbaijan has characteristics of both war-torn executive branch of government, within the and concentrated political regimes. Moreover, framework of a multiparty electoral system. political systems are still evolving, so countries These countries include Bulgaria, Croatia, the have been shifting over time across different ideal Kyrgyz Republic, FYR Macedonia, Moldova, types. Croatia and the Slovak Republic have Romania, the Russian Federation, the Slovak moved sharply toward a more competitive demo- Republic, and Ukraine. The countries span a cratic system following critical elections that ended wide range of political systems. Bulgaria and the lengthy regimes of once powerful political leadl- the Slovak Republic have a high degree of ers. Romania has shown steady improvements in political contestability over the selection of extending political rights and civil liberties. Such governments, though some of these govern- changes suggest that countries are not locked into ments have in turn concentrated political any particular path of political transition. But power to limit contestability in policymaking. because this report summarizes the entire transi- * Noncompetitive political regimes constrain tion path, we rely on measurements averaged entry of potential opposition parties into the across the entire transition period. electoral process and sharply restrict politi- cal participation through the exercise of civil liberties. Such systems tend to have few in- CoetitHigh Political stitutionalized limitations to check the execu- Contestabilit-.. tive. These countries include Belarus, J)olitical contestability is a function of the Kazakhstan, Turkmenistan, and Uzbekistan. l rights to participate in the political system * In addition to political liberties and civil and the extent to which competing institutional rights, war and political violence define a actors have the power to influence political fourth group of transition economies, where decisionmaking. It is generally measured by thie external conflicts or extreme internal number of political actors whose approval is contestability has strained the state at certain necessary to adopt binding decisions on the pol- periods of the transition.1 War-torn political ity. As the number of political parties in a coa- regimes have engaged in prolonged wars or lition government increases, the number of po- civil conflicts over the past decade. Such con- tential veto points for policy decisions also flicts have generally been rooted in ethnic or increases. Similarly, presidential systems in territorial divisions. In these countries there which competing political parties dominate the is political contestability over the boundaries executive branch and legislature (that is, divided of the community to be governed or who has government) should also exhibit higher the right to select leaders and make binding contestability than those where the president's rules for that community. Such conflicts have party also controls the legislature. A simple placed severe strains on the capacity of the measure of political contestability based on state, resulting in some of the countries in a these veto points has been developed by Roubini prolonged loss of political order and control and Sachs (1989). The index measures the nun- and serious weaknesses in the provision of ber of competing political parties in governing basic public goods. They include Albania, coalitions in both parliamentary and presiden- Armenia, Azerbaijan, Bosnia-Herzegovina, tial systems (data for the transition economies Georgia, and Tajikistan.2 can be found in Frye and Hellman 2001; see also figure 10.2). Like all ideal types, these four categories are Competitive democracies have the highest not intended to be absolutes. Several transition number of veto points among countries in the 99 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Former Soviet Union FIGURE 10.2. Veto Points Index, 1989-99 Latvia Czech Republic Estonia Slovenia Competitive democracies Poland Hungary Lithuania Average Slovak Republic Bulgaria Romania Kyrgyz Republic Moldova Concentrated political regimes Russian Federation Ukraine Croatia Average Albania ~~~~.. .. ...... - - - - - - - - - - ........ .... Albania Macedonia, FYR Georgia Armenia War-torn regimes Tajikistan Azerbaijan Average Belarus Kazakhstan Turkmenistan 0 Noncompetitive political regimes Uzbekistan 0 Average I , . 0 1 2 3 4 Veto points index Note: The index for each country is based on the average monthly score from the onset of the transition through mid- 1999 on a scale of 0-4, defined as: 0 = One-party government with noncompetitive elections. 1 = One-party majority parliamentary government or united presidential government. 2 = Two-party coalition parliamentary government or divided presidential government. 3 = Three-or-more party coalition parliamentary government. 4 = Minority parliamentary government. Source: Frye and Hellman (2001). 100 Classifying Political Systems in Transition region. All of them are parliamentary or ... and High Government Turnover semipresidential systems governed by multiparty ountries with greater political contestability coalitions. Lithuania and Poland could be char- k also have, as might be expected, more fre- acterized as semipresidential systems, which com- , , ~quent changes of government (figure 10.3). In- bine parliamentary government with a directly deed, the countries most advanced in economic elected president. Indeed, one-party majority gov- reform have tended to have the most frequent ernments have been rare in most of these coun- changes in government, contrary to the conver- tries. Six of the seven competitive democracies tional view that such turnovers create an un- have had prolonged periods of coalition govern- certain environment that undermines reform. ments consisting of three or more political par- There have been nine governments in Poland, ties. In Estonia, Poland, and Slovenia coalition governments of up to five political parties have seven in EstoniA, an Hup the tite not been unusual. While the literature on politi- democracies have had a a verage of six gov- cal economy generally takes a negative view of ernm es havers an average of six So- coalition governments, the most reformist gov- viet bloc. This contrasts sharply with the other ernments in transition economies have tended * . . to be multiparty coalitions (Alesina and country groups. Concentrated political regimes, Rosenthal l995; Roubini and Sachs 1989). except borderline Bulgaria, have tended to have Conentrated p95 oubiticand regis generaly fewer government turnovers, averaging just Concfewentraete pointion aleregme(sgenefrally more than three for the group. In all of the non- have fewer veto pomits on average (see ilgure competitive countries, except Belarus, the leader 10.2). Five of the eight concentrated political at the time of the dissolution of the Soviet Union regimes are presidential systems. Bulgaria and has ruled continuously throughout the transi- the Slovak Republic, on the border between con- tin. Poli continuity thaou t tie af- centrated and competitive political systems, tin oiia otniyhsntpstvl f ctated multipan competitiongoverpolit stems, mfected the government's propensity to adopt have multiparty coalition governments more economic reforms. similar to a competitive system than a concen- trated one. Several of the presidential systems appear to have been ruled by divided govern- ments for much of the transition, mainly be- Notes cause presidents in the Kyrgyz Republic, Rus- 1. This report defines political violence as government sia, and Ukraine did not belong to any political turnovers or attempted turnovers through violence. party for extended periods. 2. This group includes all countries in the Europe Noncompetitive political regimes have the and Central Asia region that have been engaged in lowest scores on the veto points index. Powerful prolonged military conflicts, except Croatia and Rus- presidents in Belarus, Kazakhstan, Turkmenistan, sia. In Croatia the military conflict did not spark and Uzbekistan have essentially created one- domestic political violence that threatened the in- party systems with strong restrictions on oppo- cumbent regime of Franjo Tudjman. Russia engaged sition parties. Similar one-party systems have also in a long-term territorial conflict in the breakaway been developed in the war-torn political systems, republic of Chechnya and survived a violent attempt been developed in the war-torn political systems, to overthrow Boris Yeltsin. However, these two such as Tajikistan, but these countries have gen- events were unrelated and did not lead to disorder erally had much less stable governments as a re- and political fragmentation, as in other countries in sult of the conflicts. the group. 101 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Former Soviet Union FIGURE 10.3. Main Political Excecutive Turnovers, 1989-99 Poland Lithuania Estonia Latvia Competitive democracies Hungary Czech Republic Slovenia Average Bulgaria Romania Moldova Slovak Republic Kyrgyz Republic Concentrated political regimes Ukraine Croatia Russian Federation Average _ .. ........... .......... . . . . . . . . . . . . . . . . . Albania Azerbaijan Armenia Georgia War-torn regimes Macedonia, FYR Tajikistan Average Belarus Turkmenistan Kazakhstan Noncompetitive political regimes Uzbekistan Average _ l l l 0 2 4 6 8 10 Number of turnovers Note: This figure measures the number of times there has been a change of the country's lead executive-president in presidential systems and prime minister in parliamentary or semipresidential systems-since the country gained independence. Source: Frye and Hellman (2001). 102 Political Systems Influence the Choice of Economic Reforms T he simple classification of political systems in transition economies allows us to examine the relationship among different types of political institutions and the range of outcomes on the spectrum of discipline and encouragement discussed in part 2. Competitive democracies have proven to be among the most advanced economic reformers in the Europe and Central Asia region, pursuing policies that have promoted SMEs and maintaining hard budget constraints on both new and old enterprises. Concentrated political regimes have been more likely to sustain a pattern of partial reforms that protect old enterprises and create barriers to market entry. Yet in these countries the combination of liberalization and privatization with continued soft budget constraints and a weak rule of law have encouraged even new enterprises to focus their efforts on rent seeking and tunneling instead of productive entrepreneurship. Noncompetitive political regimes have been most likely to reject key elements of market transition, choosing instead to maintain greater continuity with the structures and practices of the previous corn- mand system. Though these regimes protect state enterprises and restrict the activities of new enter- prises, lack of any substantial liberalization or privatization has prevented the types of rent seeking and tunneling prominent in the concentrated political regimes, thus avoiding the sharp contractions common among other reform-minded countries. War-torn countries have been characterized by weak state capacity and a zig-zag pattern of economic reform, creating an environment that is not condu- cive to entry and investment. To measure the extent of economic reform, we rely on the European Bank for Reconstruction and Development's transition indicators (EBRD 2000), which evaluate annual progress in transition in eight different categories of market-oriented reform on a scale from 1 (little or no reform) to 4.3 (standards typical of advanced industrial economies). Competitive democracies have made the great- est progress in implementing market-oriented reforms, while the noncompetitive regimes have made the least (figure 11.1). Concentrated political regimes and war-torn regimes have made partial progress, advancing in some areas and lagging behind in others. The direction of the causality underlying these correlations is difficult to untangle completely. In- deed, despite a vast literature examining the relationship between levels of democracy and econonmic 103 Transition-The First Ten Years: Analysis and Lessons for Eastem Europe and the Fonner Soviet Union FIGuREI 1.1. Political Systems and Economic Reform Outcomes, 2000 Mean scores 1 2 3 4 Hungary Czech Republic Poland Estonia Slovenia Competitive democracies Lithuania Latvia Average Slovak Republic Croatia Bulgaria Romania Kyrgyz Republic Moldova Concentrated political regimes Russian Federation Ukraine Average _................... . ....... . ............ ..... . .. ....... ... . ----------------- Macedonia, FYR Georgia Albania Armenia War-torn regimes Azerbaijan Tajikistan Bosnia and Herzegovina Average Kazakhstan Uzbekistan Belarus Noncompetitive political regimes Turkmenistan Average l_l_l 0 0.4 0.8 1.2 Variance Note: The European Bank for Reconstruction and Development measures annual cumulative progress in transition each year in eight different categories of market-oriented reform on a scale from 1 (little or no reform) to 4.3 (standards typical of advanced industrial economies). Each bar represents the country average across all eight categories. The line represents the variance, which for each country measures the dispersion around the mean across all eight reform categories of the transition indicators. Source: EBRD (2000). 104 Political Systems Influence the Choice of Economic Reforms reform across the world, the results are still largely process-such as price and trade liberalization inconclusive (for a review of this literature, see with continued restrictions on entry, privatization Haggard and Webb 1993). Though the nature of with soft-budget constraints, and rapid creation the political system structures the incentives of of banks without a sufficient regulatory frame- politicians to adopt economic policy choices, the work for financial markets-create opportuni- reform choices themselves shape the configura- ties for rent seeking and theft. One crude indica- tion of social groups and the distribution of power tor of these imbalances comes from comparing that affect the structure and functioning of the the different rates of progress across the eight political system. For example, economic reforms different components of the European Bank for that enable new entry also strengthen the con- Reconstruction and Development's transition in- stituency of SMEs that build support for increas- dicators. Taking the variance (a measure of the ing political competition. Thus, the development dispersion around the mean) of the ratings for of the political system and progress of economic each country across these eight reform compo- reforms are closely inter-related. nents gives a rough measure of the imbalances Yet given the sharp break with communism in the reform that give rise to rent seeking (see and the disintegration of the Soviet Union, figure 11.1). choices about the structure of the political sys- The variances tend to be lowest in the most tem in transition economies were generally made reformist and least reformist countries. In corn- before decisions about the nature and pace of petitive democracies economic reforms have gen- economic reform. As a result, a stronger case can erally progressed across the board despite dif- be made for identifying the direction of causa- ferences in the sequencing of reform measures tion from political choices to economic choices. across countries. Similarly, noncompetitive po- Moreover, only a few countries (Belarus, Croatia, litical systems have generally made little or rio and the Slovak Republic) have seen a major progress in all of the key areas of economic re- change in political regime since the start of tran- form, maintaining substantial continuity with the sition. This suggests that while the pace and di- previous command system. In contrast, the con- rection of economic reform may have reinforced centrated political regimes and war-torn regimes initial choices about the structure of the politi- have tended to advance rapidly in liberalization cal system, economic reforms have yet to deci- and privatization with much slower progress in sively shift the course of political transition. the institutional reforms to support effectively However, after only a decade of transition, functioning markets, generating much higher these political and economic systems are still in variances in their reform scores. their infancy. As in all countries, one might ex- These asymmetries in the reform process cre- pect that over the long term, there will be a strong ate a wide range of arbitrage and rent-seeking interactions among economic reform, economic opportunities available to a small group, usu- performance, and political evolution (Przeworski ally with close ties to the government or the ex- 1991). Yet in the early stages of transition, po- isting state-owned sectors. The gap between litical choices appear to be the driving force of progress in liberalization and privatization and change in economic reform. development of a proper legal and regulatory framework also provides opportunities for theft Political Systems Crea,te Rent-Seeking of both state and private assets through expro- Opportunities priation of minority shareholders. The pace and symmetry of market-oriented A n important indicator of the opportunities reforms suggest variation across the different po- A for rent seeking in the transition process litical systems in rent seeking and asset stripping. comes not just from the overall measure of A good proxy measure of such phenomena can progress in market-oriented reform, but from be found in recent attempts to develop more sys- different progress across different components tematic, survey-based indices of corruption and of the reform agenda. Imbalances in the reform state capture. (State capture refers to the efforts 105 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Former Soviet Union of enterprises to influence laws, decrees, regula- 3.1) provides an index of state capture based on tions, and the like through private payments to the share of enterprises that reported a direct im- public officials.) By capturing state institutions pact on their business from private influence pay- enterprises seek to extract rents from the state ments to public officials (figure 11.2; see chap- througha myriad of preferences, exemptions, and ter 3).1 The extent of state capture can be anticompetitive practices. The Business Environ- compared across countries with different types ment and Enterprise Performance Survey (see box of political regimes. FIGURE 11.2. State Capture Index, 1999 Slovenia Hungary Estonia Lithuania Czech Republic ublic C mi ive democracies Poland Latvia Average Romania Slovak Republic Croatia Bulgaria Kyrgyz Republic Concentrated political regimes Ukraine Russian Federation Moldova Average Armenia Albania Georgia War-torn regimes Azerbaijan Average Uzbekistan Belarus Noncompetitive political regimes Kazakhstan Average _ l l l 0 10 20 30 40 50 Index Note: The state capture index is based on the Business Environment and Enterprise Performance Survey. Enterprises were asked to what extent their business was directly affected by private payments to affect the decisions of six institutions: the presidency, legislature, government apparatus, civil courts, criminal courts, and the central bank. The bar for each country represents the share of enterprises that reported a significant impact averaged across the six institutions. As the measurement of this index is subject to a margin of error, any efforts to rank countries would be inappropriate. The data were collected in 1999 and do not reflect the impact of reforms since that time. Source: Hellman, Jones, and Kaufmann (2000); World Bank (2000c). 106 Political Systems Influence the Choice (of Economic Reforms There is a stark contrast in the extent and negotiations by popular fronts and a wide range impact of state capture across different political of other organized interests from trade unions systems. Concentrated political regimes exhibit to religious representatives. Guided partly by the consistently higher state capture, affecting on example of Western European democracies, the average more than twice as many enterprises as roundtables produced political institutions that in competitive or noncompetitive political sys- generally tended toward parliamentary systems, tems. Except Latvia, only a small share of enter- promoting party competition and constraining prises in competitive democracies report a sig- executive authorities. nificant impact from state capture. State capture The inclusive process for creating these po- is also low in noncompetitive political regimes, litical institutions and the broad range of politi- reflecting the weakness of the private sector rela- cal groups that could compete for power in the tive to a highly authoritarian state. War-torn new system enhanced the capacity of govern- countries have both high and low state capture: ments to make credible commitments that the low in countries with high domestic instability, promised gains of economic reform would not such as Albania, Armenia, and Bosnia and high be expropriated or otherwise restricted to par- in countries where powerful leaders have begun ticular vested interests. This contributed to a to consolidate political power, as in Azerbaijan wider social consensus on the main directions of and Georgia. The concentration of political reform, despite differences among parties on the power appears to be an important determinant sequencing and pace of reforms. It also led to a of the extent of state capture in transition econo- greater mobilization of organized interest groups mies (Hellman, Jones, and Kaufmann 2000). in civil society (such as independent labor unions) that would enhance political accountability How Do Political Systems AffectEconomicthroughout the transition. So governments mo- Refows? bilized broad public support for comprehensive reform programs early in the transition. T he data suggest clear links among different A key factor in building and sustaining this T types of political systems and alternative broader consensus on reform has been the his- paths of economic reform as defined in the disci- torical ties and geographical links of these coun- pline and encouragement framework. Of course, tries to Western and Northern Europe. The pull the correlation between political regimes and dif- of European accession generated strong incen- ferent economic reform paths need not imply cau- tives for a common institutional framework, both sation. There may be other factors, such as his- for the economy and the polity, creating a focal torical and cultural legacies, the structure of the point for the reform agenda that mitigated pos- economy, and even geographical position that af- sible disputes over alternative institutional cle- fect both the choice of political regime and the signs and regulatory frameworks. course of economic reform. Such factors are all By taking a comprehensive approach early on, highly correlated, making it impossible to disen- such programs generated far fewer opportunities tangle empirically the lines of causation. However, for rent seeking and theft, lowering the returns to the strong similarities in the transition process in such activities in the short term. This weakened countries with similar political systems suggest that the capacity of short-term winners to restrict com- political institutions affect choices about economic petition and preserve rent-generating market dis- reform. What are these similarities? tortions. Moreover, the regular succession of dif- ferent coalition governments created genuine competition among groups for political influence. In Competitive Democracies, Inclusion Promotes That led to equally fierce competition over rents, quickly dissipating any efforts to concentrate rent In the aftermath of popular revolutions against flows and preventing theft on a massive scale. communist rule, the political institutions in com- As already suggested, this competition oc- petitive democracies were forged in roundtable curred within a broad consensus on the direction 107 Transition-The First Ten Years: Analysis and Lessons for Eastem Europe and the Fonner Soviet Union and goals of the reform. Even the return to power In Concentrated Political Regimes, Oligarchs and of communist-successor parties in such countries Insiders Capture the State as the Czech Republic, Hungary, and Poland did not derail reform. Government credibility at the In concentrated political regimes the collapse of early stages of transition enabled a first move on communism was more a result of the contest the reform path that limited the rents from arbi- among competing elites than a broad social trage opportunities in distorted markets. It quickly movement. The new political regimes were not created new constituencies with an incentive to forged though roundtable negotiations among push for further reform. It also signaled to inves- potential competitors, rather, the regimes were tors the government's commitment to more long- designed by incumbent leaders to consolidate term structural reforms. Some groups gained im- their power. These new regimes tended to be mediate advantages from liberalization and presidential systems with power concentrated in privatization. But they could not convert the gains the executive branch. Political parties were weak into enough political influence to erect barriers to and did not represent a broad range of social competition and entry that would have preserved interests. Moreover, the old nomenklatura re- limited opportunities for rent seeking and theft mained strong, especially in the economy. along the way. Though comprehensive reforms were pro- From their first moves in the reform process, posed in some concentrated political regimes in new governments tended to focus on promoting the early stages of transition (Russia in 1991), new constituencies of winners, removing entry they were rarely adopted in full, and the ones barriers, and quickly tackling severe macroeco- adopted were poorly implemented. The regimes nomic instability. They also supported the losers lacked the credibility to build and sustain broad from the dislocations of the reform by maintain- popular support for such reforms. Instead, they ing adequate social protection (Orenstein 2001). tended to fall back on partial liberalization and Early efforts by enterprise insiders to spontane- privatization. The soft budgets and remaining ously privatize enterprises sparked a substantial barriers to entry generated tremendous oppor- political backlash in the Hungarian and Polish tunities for rent seeking and theft, especially in electoral arenas (and much later in the Czech economies rich in natural resources. elections), curbing the practice with varying ef- Such exits from communism also led to much fectiveness. As a result the concentration of eco- sharper deteriorations of state capacity than nomic power in the early stages of transition in among the competitive democracies of Central these competitive democracies was far less than Europe. In former Soviet states, secessions were in other parts of the region (evident in the data consciously intended to weaken federal control on inequality). As reforms progressed to promote by radically decentralizing authority over public entry and improve the enabling environment, bureaucracies, sparking a wave of asset stripping. they strengthened the constituencies with a stake In Bulgaria and Romania, for example, the col- in moving the reform agenda forward in the dif- lapse of incumbent regimes caused greater uncer- ficult areas of structural and institutional change. tainty than in the negotiated transitions of Cen- What enabled this virtuous circle? One key tral Europe, eroding state control. Furthermore, factor was the much higher state capacity than the deterioration of public administration pro- anywhere else in the region. The fairly peaceful foundly reduced the state's capacity to raise rev- exit from communism did not destroy key state enue, implement proposed economic reforms, and institutions. Facilitating more effective imple- build broader consensus around reform goals. mentation of reforms, the legacy of public ad- The partial reforms merely increased and ministration provided important preconditions concentrated rents and the opportunities for for promoting new entry, such as greater secu- tunneling and theft. Comprehensive price lib- rity of property and contract rights and better eralizations were often followed by a stream of public infrastructure. executive orders and decrees (with exceptions 108 Political Systems Influence the Choice of Economic Refonns for some sectors and goods). Licensing and limited institutional restraints to promote account- other regulatory barriers preserved trading mo- ability and weak intermediate organizations to nopolies. Explicit controls on foreign competi- channel the dissatisfaction of losers. Many con- tion were set. centrated democracies languished in an equilib- The winners from these partial reforms rium trap of partial economic reforms in which reaped spectacular gains. Those with control the concentration of both political and economic rights over valuable assets and political access power preserved market distortions generating to rent-seeking opportunities could easily priva- highly concentrated gains to narrow vested inter- tize those assets in an environment of poorly ests at considerable social cost. defined property rights, nonexistent corporate governance, and an inadequate regulatory frame- wr. As lon as th stc,frns a oe In Noncompetitive Political Systems, Leaders Savor Status Quo and Economic Stability pleted, these winners had a strong incentive to preserve their advantages, using their consider- In noncompetitive political regimes leaders from able resources to block reforms that threatened the Soviet era generally pursued economic sta- those advantages. bility, while securing their dominance of the post- Countervailing pressures from competing Soviet political system. Concerned about rival groups were weak, and without genuinely broad- sources of power in both the economy and the based social movements, no clear goals were ar- polity, they largely rejected market-oriented re- ticulated at the start of transition. Nor was there forms. They feared the opening to global ma.r- much social mobilization from the collapse of ket forces and the rise of oligarchs in partially communism. As a result, these countries em- reformed economies. Instead, they preferred to barked on transition without building a broad maintain continuity with key elements of the social consensus on the goals of reform and with- previous command system to maintain the state's out a means of organizing the public behind these (and the leaders') predominant role and to avoid goals. Effective political parties never material- any destabilizing adjustment costs associated ized to mobilize social support, and residual sup- with reform. port for the Communist Party remained high. In these cases economic reform was driven That set the stage for much greater political po- not by the winners or losers of reform, but by larization over economic reform, and that po- authoritarian political leaders trying to maintain larization in many cases became a pretext for political control and ensure some economic sta- further centralization of political power and lim- bility. Limited economic reform went hand in its on political competition (EBRD 1999). hand with limited political reform, as incumbent Lacking strong social support and a solid po- leaders severely restricted political opposition. litical base, incumbent politicians in concentrated The political imperative to maintain state democracies sought alliances with powerful in- power and the lack of economic reform avoided cumbent enterprises for funding and support. This the disintegration of state capacity that plagued naturally made the state highly subject to capture. many other transition economies. So although The early winners from the mass of arbitrage op- the state in these countries does not provide the portunities had a strong interest in using their po- institutional foundation for the market economy, litical influence to preserve rent-generating dis- it still provides public goods-but at levels simni- tortions that intensified the winners' economic lar to the communist era. This suggests that if a power. Direct barriers to entry solidified their new government introduced comprehensive eco- gains, weakening new constituencies that might nomic reforms it would have stronger capacity counterbalance these powerful incumbents. to implement them and would begin with a less The social costs of state capture have been high daunting decline in living standards. in these countries. But the direct costs to politi- Despite these possible advantages, incumbent cians of poor policy choices were low thanks to political leaders face little pressure to pursue 109 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Former Soviet Union reforms. By maintaining dominant state ownership of Nagorno-Karabakh, Albanians in the Kosovo in the economy and considerable discretionary ca- region of Yugoslavia, Abkhaz and Meshketians in pacity to intervene in economic affairs, these lead- Georgia, fundamentalist Muslims in Tajikistan- ers ensure their positions remain powerful. Overt all are cases of minority groups seeking to pre-empt restrictions on political competition reduce even through violent means the creation of a state domi- further the risk of challenges to incumbents' power. nated by the ethnic majority. Governments in such In these systems few viable constituencies have environments could not make credible commit- the capacity to push for economic reforms. The ments about the distribution of future gains from economic elites tend to be closely allied with (or reforms or about not expropriating wealth. co-opted by) incumbent political leaders. There During periods of peace and relative stabil- is no critical mass of new actors with an interest ity, governments in these countries have tried to in reforms. Trade unions, business associations, adopt comprehensive reform programs, but they and other civil groups that might represent continue to be undermined by severe credibility broader social interests are circumscribed or tied problems. War and instability radically shorten to the state. In addition, the lack of reform iso- time horizons, so ethnic groups are not prepared lates these countries from the global economy, to accept short-term sacrifices for the promise leading to autarky that undermines what exter- of long-term gains. nal constituencies might do in promoting reform. Prolonged conflict also sharply reduces out- The extent of economic reform is thus highly put, living standards, and the resources of the dependent on the political leader's preferences. state. Physical and human capital deteriorate. Although many authoritarian leaders in other Poverty increases. Even the capacity of the state regions adopted comprehensive economic re- to provide the most basic public goods falls apart. forms, this has not yet happened in any of the War also can concentrate economic power transition economies. Why? Authoritarian lead- in the hands of smugglers, arms traders, and ers in the region have tended to inherit their power paramilitary groups, who use their power in and support from the surviving structures of the peacetime to secure their positions in the former communist system. This has created a economy. These "winners" use the instability of strong link between authoritarian political power war to centralize control over state assets and and command administrative methods in the eco- distribution networks, particularly energy. They nomic sphere in transition economies. also tend to maintain close relationships with political leaders, who depend on them to fund and supply the war. Given the weakness of In War-Torn Regimes, Violence and Lack of an supytewr.ie h ekeso Credibilit Prevent ReVorm countervailing interests, constraining the power of such groups during peacetime is difficult. In war-torn political systems, efforts to promote Manipulating privatization to enhance their con- comprehensive reforms at the early stages of tran- trol over assets, the winners use their influence sition were thwarted by contestability over who to preserve exemptions and other preferences that had the rights to make binding decisions for the undermine competition and to weaken the de- community and over the definition of the com- velopment of the rule of law. munity. Societies with deep-rooted ethnic divi- This combination of political instability, a sions face a high risk that governments domi- weak state, and powerful economic groups nated by one group will seek to expropriate rooted in illicit and opaque trade has undermined wealth and resources from rivals, while encod- reform in most war-torn regimes. ing advantages for themselves into the develop- ing institutional structure. Note Facing such threats, minority ethnic groups in n ~~~~~~~1. H4ellman, Jones, and Kaufmann (2000) present several countries in the region have violently chal- evidence from the Business Environment and Enter- lenged the legitimacy of the state. Croats in Bosnia- prise Performance Survey to show that enterprises that Herzegovina, Armenians in the Azerbaijani enclave engage in state capture get substantial rents. 110 Confronting the Political Challenge A fter a decade of transition, the political challenges of pressing ahead with the remaining reform agenda differ substantially in each of the four groups of countries. To build a founda- tion of public support for economic reform, governments should focus on smoothing the curves of winners and losers in the short term (see figure 9.1). That entails lowering the short-term adjustment costs for the potential new entrants and the high concentration of gains for the short-term winners such as oligarchs and insiders. Also needed are political changes to enhance the government's credibility and capacity to constrain the power of constituencies seeking to sustain partial reforms regardless of the social costs. To smooth the short-term loser's curve, governments need to mitigate the adjustment costs asso- ciated with comprehensive economic reforms by preserving a social safety net that cushions the dislocations of the downsizing state sector. Such support is also an important way for the govern- ment to signal its commitment to defending broad public interests in reform, which will enhance government credibility. To smooth the short-term winners' curve, governments need to reduce the incentives that lead oligarchs and insiders to block reform midstream through taxation or other redistributive schemes. But where the state is highly susceptible to capture by these groups, the likelihood of implementing such schemes is small. So reducing the concentration of gains to oligarchs and insiders must be based on changes in the structure of political power. Mobilizing collective action by a broader range of social groups that will gain from further reforms could pressure politicians to reduce state capture. Reforrn- ing the political system-to increase participation and enhance political competition-can break the link between highly concentrated economic power and political power. Increasing transparency and accountability in government increases the costs to politicians of skewing economic reform in the interest of narrow constituencies. Of course, the oligarchs and insiders will oppose such political reforms, recognizing the threat to their advantaged positions. Consequently, successful political reforms in states captured by such groups are likely only after significant changes in the underlying correlation of power in the system. Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Fonner Soviet Union Not all the transition economies battle these subsidized inputs, and monopoly production and same problems. A decade of variation in reform distribution rights. Commercial banks have taken paths and political developments has left coun- advantage of macroeconomic volatility, ineffi- tries with very different political challenges. cient financial markets, and lax regulatory struc- tures to reap gains from arbitrage. Politicians and bureaucrats have used their discretionary pow- ForilCoinc Potentrated Pitical ers to intervene in the economy to extract bribes and other advantages from enterprises and B reaking the political economy equilibrium households. Enterprise insiders have manipulated B underlying state capture and partial reforms unclear property rights and weak corporate gov- is the most important and difficult challenge in ernance to divert enterprise assets and cash flow advancing the transition in countries with con- into offshore companies and other subsidiaries centrated political regimes. The vested interests under their direct ownership. These practices underlying soft budget constraints, barriers to have been largely at the expense of unprotected entry, opaque regulatory frameworks, and land small shareholders. reform have accumulated considerable economic The winners from these market distortions and political power, raising more barriers to po- and inefficiencies have powerful incentives to litical entry for the losers of such policies. State preserve them and the associated rent flows, re- capture at all levels of the political system and gardless of the social costs. Large exporters lobby the lack of accountability for politicians and for entry barriers. Commercial banks oppose sta- public officials make it difficult for even the most bilization programs and proposals to enhance committed reformers to overcome the powerful the central bank's regulatory powers. Insider- vested interests against reform. In addition, weak owners undermine efforts to clarify corporate state capacity has limited the state's autonomy governance and to introduce greater transpar- to tackle these groups effectively. ency into the distribution of property rights. Reforms have proven most difficult politically Public officials resist deregulation and efforts to where the rents have been most concentrated, limit discretionary interventions. Although the particularly in energy and finance. In nearly all gains from such distortions tend to be highly the concentrated democracies with substantial concentrated, the losses are dispersed among natural resources, energy lobbies have been among consumers, savers, minority shareholders, start- the most formidable opponents of reforms. Their up companies, small and medium-size businesses, political power, often exceeding that of major and foreign investors. political parties, cannot be underestimated. This pattern of dispersed losses and concen- The political leverage of the power and energy trated gains holds the key to designing politi- sectors comes not just from their resources, but cally feasible strategies of economic reform. In through the vast network of nonpayments. As stud- many concentrated democracies, reformers of- ies of the virtual economy suggest, dominantpower ten attempt to overcome political obstacles by and energy monopolies are the key sources of value augmenting executive power to counter the in the complex web of nonpayments because the power of vested interests. Yet given the state monopolies continue to provide oil, gas, and capture in these countries, such efforts tend to electricity to loss-making enterprises, often in ex- fail. At various times the presidents of both change for overvalued barter goods (Gaddy and Russia and Ukraine were granted extraordinary Ickes 1999). Being the main conduits of the state's decree-making powers to push through eco- soft-budget constraint gives them considerable po- nomic reforms that did not break the stalemate litical influence over the state. on further reforms. Instead, concentrated gains Many other groups could be considered win- and widely dispersed losses suggest the mobili- ners from the distortions of partial reforms. zation of the losers in the existing low-level equi- Natural resource extraction and export enter- librium through greater political inclusion. How prises have gained from distorted domestic prices, such a strategy should be designed depends on 112 Confronting tbe Political Challenge factors specific to each country. But some gen- than 20 percent of the enterprises reported memn- eral approaches have worked in other transi- bership in business associations, and less than tion economies. 10 percent relied on them for resolving prob- lems with the government (World Bank 2000c). Mobilize the electorate. Electoral appeals to The unofficial sector is a large and poten- mobilize the losers of partial reform have built tially influential constituency that has not been support for macroeconomic stabilization and for effectively mobilized in the concentrated democ- banking reform. Given the wide and generally racies. Enterprises in the unofficial economy gen- regressive impact of high inflation, political par- erally suffer most from the weakness in the en- ties in several transition economies have mobi- abling environment, especially in the opportunity lized enough electoral support for macroeconomic costs associated with limitations on their capac- stabilization to overcome the opposition of pow- ity to expand operations. Tax reforms that lower erful commercial banks and other actors that marginal rates can promote entry from the un- gained from macroeconomic volatility. In addi- official to the official economy. This would pro- tion, banking crises in the Czech Republic and mote growth and possibly crystallize an impor- Hungary sparked electoral appeals to disgruntled tant political constituency to remove obstacles savers, helping to break the stalemate over bank- in the business environment that work only Ito ing privatization and regulatory reform. the advantage of a narrow group of powerful enterprises. Given the size of the informal sector Mobilize excluded enterprises. Another im- in many of these countries, the potential poliri- portant strategy for weakening the opposition cal consequences of mobilization could seriously of narrow vested interests to reforms is mobiliz- shift the balance of power away from incurn- ing collective action among enterprises excluded bent vested interests to a much broader collec- from these concentrated gains. SMEs, new en- tion of economic actors. However, there are few terprises, and second tier enterprises suffer most examples of rapid shifts from the informal to from existing weaknesses in the enabling envi- the formal sector, or of the political mobiliza- ronment, from discretionary taxation and regu- tion of such economic actors. lation and from anticompetitive barriers. How- One way to promote new enterprises and ever, they lack vehicles of collective action and those in the unofficial sector is to align the in- influence with the government. In Central Eu- centives of local governments to increase entry. rope business associations have strengthened the Tax-sharing schemes between central and local voice of this tier of the economy, constraining governments can be modified so that property public officials and checking influential enter- taxes on small business are assigned exclusively prises and financial-industrial groups. In the con- to local governments, encouraging them to re- centrated democracies, such associations have form the enabling environment. remained weak, and political parties have not sought strategic alliances with such actors as an Ensure-and use-a free media. The emer- alternative basis for support and funding. gence of a free press and broadcast media can The Business Environment and Enterprise overcome the coordination dilemmas associated Performance Survey (see box 3.1) showed vast with reforms that generate concentrated winners differences across transition economies in how but highly dispersed losers in many sectors. The many enterprises were members of business as- media can promote collective action among these sociations and how they used these associations dispersed losers of partial reforms and increase to seek remedies to problems with the govern- the political pressure on those who gain from ment. In Hungary more than 75 percent of the the general equilibrium underlying these econo- enterprises surveyed reported membership in a mies. The challenge for reformers in concentrated trade association, and 60 percent said they would democracies is to use the media to make clear rely on such associations first in handling prob- the links between the rents from partial reforrms lems with the government. But in Russia fewer and the direct costs to society. 113 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Former Soviet Union But the problem in many of the concentrated strengthening political parties and other organi- political regimes is that ownership of the media is zations representing the collective interests of al- closely intertwined with powerful financial- ternative constituencies, and developing institu- industrial groups. In other concentrated political tional restraints will expand political access regimes the media remains largely under the con- beyond vested interests and increase the costs to trol of the state. This naturally limits what the politicians of maintaining partial reforms. But media can do in breaking the partial reform trap. breaking the equilibrium of partial reforms is a challenge that can be addressed only through si- Make obvious what is hidden. Tax arrears multaneous economic and political reforms. and nonpayments need to be linked in the pub- Though exogenous shocks often create windows lic mind with delayed public sector wages and for decisive action by committed reformers, their pensions and poor provision of social services. capacity to spur profound changes in the struc- The complex web of hidden subsidies to power- ture of political and economic power simulta- ful business needs to be exposed, making clear neously is highly constrained, and their agenda that such subsidies tend to benefit incumbent highly overloaded. managers (often through offshore accounts) Our analysis suggests that promoting entry rather than workers. Barter and arrears in the and competition on the margin, whenever and energy sector need to be linked to domestic power wherever possible, creates the necessary precon- and fuel shortages and outages that plague many ditions for a gradual move out of the partial re- of these countries. form trap. As more enterprises enter the market Converting hidden and discretionary subsidies economy, the competition for rents should also to enterprises into budgetary subsidies to support promote competition for political influence, which worker training, severance schemes, and better lo- weakens the capacity of powerful enterprises and cal services in communities affected by downsizing sectors to capture the state and oppose reforms is vital for discipline and encouragement. that might undermine their advantages. Promot- The potential for mobilizing these dispersed ing entry and competition on the margin will not losers is particularly high in countries whose so- accomplish radical changes in the short run, but cial sectors need resources, but whose high- it may be the most effective means of creating profile conglomerates in key sectors enjoy a range demand for reforms in the long run. of explicit tax and duty exemptions and main- tain high tax arrears. This is particularly true in For War-Tom Poitical Systems, Restoring some of the small, fairly open economies, such Stability and Reducing Uncertainty as Georgia or Moldova, where trade flows are a major source of the tax base and powerful inter- hough the magnitude of the challenges faced ests in control of those flows are the main tax T by these countries in light of the extreme evaders. Explaining to the public the costs of degradation of state capacity and concomitant evasion should be an important element of any sharp decline in living standards may be among strategy to mobilize support for further reforms. the most serious in the region, the precondition Having made clear who benefits from partial for success is clear: without resolving the un- reforms-and how those benefits come directly derlying divisions behind political fragmenta- at the expense of large but dispersed domestic tion and violence, further reforms are unlikely constituencies-reformers can begin to build to be successful. political coalitions to marshal the potential win- Once some measure of stability is restored and ners from further reforms. uncertainty reduced, these countries need to re- build basic state capacity and restore public goods Allow political competition and economic for the functioning of the market. Attention has competition to reinforce each other. Enhancing to go to strengthening the state's capacity to col- political contestability by mobilizing civil society, lect tax revenue to meet the considerable fiscal 114 Confronting the Political Challenge challenge. Rapid liberalization in the immediate 1997). Incumbent parties can win support by ex- aftermath of conflict is critical to eliminating the panding fiscal expenditures before key elections, rents associated with controls on prices and dis- possibly requiring sharp contractions after the tribution networks that fuel the winners of the elections. Some competitive democracies (Hun- wartime economy. In addition, comprehensive gary and Poland) have faced or are now grap- structural reforms, focused on privatization and pling with high budget deficits, with possibly de- demonopolization, are needed to prevent these stabilizing consequences. wartime winners from solidifying their position Some competitive democracies have com- through state capture in the postwar economy. bined budget deficits with high current account Given deteriorated state capacity in most of deficits. The rapid inflow of foreign capital has these countries, direct assistance and participa- risks and benefits. Banking systems in many of tion by bilateral and multilateral donors will be these countries are not sufficiently developed and critical in generating the resources, providing the regulated to handle these inflows. A key chal- necessary technical assistance, and buttressing the lenge is to prevent a mismatch between demand political commitment for these fundamental tasks for private sector borrowing and the capacity of of state building. the domestic banking system to make credit al- location decisions, monitor borrowers, and en- .. .. . ~~~~~force discipline on delinquent borrowers. The For Noncompetitive Political Systems, Taking foc dicpln ondlnun.orwr.Ti Advantage of State Capacity risk of political interference in banking systems remains and must be strenuously avoided. Pre- A change in political regime could create op- serving and enhancing the independence of regu- A portunities for implementing a comprehen- latory and supervisory bodies in the financial sive reform strategy. Regime change is often ac- system is crucial to preventing recurrent crises. companied by a resurgence of political competi- Success should not breed complacency in tion, a strengthening of political parties, and a meeting the remaining challenges of structural rejuvenation of civil society that can pressure new reform, particularly in the public sector and in political leaders to pursue policy innovation and politically sensitive areas of the economy. Main- improve economic performance. As suggested taining the main pillars of the social safety net earlier, these regimes can take advantage of as a cushion in the beginning of reform was im- higher levels of state capacity to implement re- portant in the success of these countries. Hovv- forms. This suggests that there might be some ever, it has also left them with highly overstaffed advantages of tardiness (in the spirit of Alexander bureaucracies and high public sector wages, Gershenkron's famous phrase) that would en- dampening their growth potential. able these countries to shift from a minimal re- Strong political resistance to downsizing and form equilibrium to a more comprehensive set wage cuts can be expected from public sector of reforms without the same deterioration of state workers. Local governments will oppose any capacity that marked countries that began the policies that reduce government control over transition with partial reforms. the provision of health services and education. Resistance from these constituencies will be dif- ficult to overcome, just as in many advanced Momentum to Build Coalitions for Refom industrial democracies. Countering this oppo- sition means developing mechanisms to mobi- A s advanced democracies around the world lize politically the new and rapidly expanding A have amply demonstrated, multiparty com- private sector, especially SMEs, whose interests petition can pose periodic risks to macroeco- are directly affected by the discretionary power nomic stability often aligned with the electoral of public sector bureaucracies. cycle, or what is referred to as the political- Beyond the public sector, many of the com- business cycle (Alesina, Roubini, and Cohen petitive democracies still need to restructure 115 Transition-The First Ten Years: Analysis and Lessons for Eastem Europe and the Former Soviet Union politically sensitive sectors, such as agriculture, of political power, reforming political institutions coal, mining, railways, shipbuilding, and steel. can alter the incentives of public officials to limit The dynamics of political coalition-building that or even reverse these imbalances. Such changes foster a broader consensus on the course of eco- are crucial for shifting the incentives of old enter- nomic reforms also tend to give these sectors con- prises and promoting the proper development of siderable power to demand subsidies, protection- new ones. ist measures, and other advantages as a condition Initial conditions and political institutions for their political support. The agricultural lobby, affect the likelihood that some countries will fol- in particular, is a powerful obstacle to reform. low particular reform paths. But these structural As the reform agenda loses urgency over time, factors can never predetermine outcomes in such government capacity to take on powerful lob- complex and multifaceted processes as transition. bies will diminish. But the lack of reforms in these A decade of transition shows the critical role of important sectors holds back growth in the new political leadership in shaping reform. A thor- private sector, weakening the overall perfor- ough political economy analysis of winners and mance of these economies. One continuing losers from reform can set the parameters for source of pressure for reform in some of these understanding the likely pressure points in any sectors is European Union accession. Again, it is system and provide guidance for crafting feasible important to mobilize domestic constituencies- reform strategies. But it cannot predict the qual- especially the vibrant new sector, which bears ity or strength of the leadership of the reform the brunt of the costs of postponed reforms-to process that will motivate the pace and direc- build a coalition with the strength to overcome tion of reforms. the opposition to further reforms. Political economy analysis has an inherent status quo bias (Fernandez and Rodrik 1991). Conclusion But experience from around the world shows that talented political leaders can maneuver T he key challenge of the political economy countries out of so-called reform traps and shift T of reform is to create the conditions that equilibrium paths. Critical elections can break will generate incentives for new market entry and long-term stalemates on reform. New leaders shift the emphasis of enterprises from rent extrac- can mobilize alternative coalitions and spark tion to entrepreneurship and productive invest- collective action that tips the balance of power ment. Though initial conditions cannot be between the potential winners and losers from changed, measures to compensate for the struc- further economic reforms. Clever reformers can tural peculiarities of different economies can be devise win-win strategies that co-opt their op- suggested. Though the "first move" of the reform ponents to build support for reform. We can- process cannot be undone, coalitions of winners not predict the "quality of reform-mongering," and losers can be fostered to weaken the grip of to use Albert Hirschman's phrase (1963, p. vested interests and deconcentrate monopoly 225), either within or across countries. How- power. Though the concentration of economic ever, it is the essential ingredient in understand- power tends to be supported by a concentration ing the politics of economic reform. 116 Selected Bibliographic Guide to the Political Economy of Transition T he simultaneous political and economic transitions in Eastern Europe and the former Soviet Union have spawned an enormous literature across a wide variety of disciplines. Though numerous references are provided in the text, the list below highlights some key contribu- tions. Many of the writings have informed the arguments in this paper. The Socialist System: Reform and Collapse . 1990. We the People: The Revolution of '89 Witnessed in Warsaw, Budapest, Ber- Aslund, Anders. 1991. Gorbachev's Struggle for lin &E Prague. Cambridge, MA: Granta Economic Reform, 2nd ed. Ithaca, NY: B Poks. Cornell University Press. Books. Easterly, William, and Stanley Fischer. 1995. IMF (International Monetary Fund), World "The Soviet Economic Decline." World Bank, OECD (Organisation for Eco- Bank Economic Review 9(3): 341-71. nomic Co-operation and Development), and EBRD (European Bank for Recon- Ericson, Richard. 1991. "The Classical Soviet- struction and Development). 1991. f Type Economy: Nature of the System and Study of the Soviet Economy. vol. 3. Implications for Reform." Journal of Eco- Washington, D.C. nomic Perspectives 5(4): 11-27. Fischer, Stanley, and Alan Gelb. 1991. "The Pro- Kornai, Janos. 1992. The Socialist System: The Political Econom o fCommunism. cess of Socialist Economic Transforma- y f tion." Journal of Economic Perspectives Princeton, NJ: Princeton University Press. 5(4): 91-105. Murrell, Peter, and Mancur Olson. 1991. "The Garton Ash, Timothy. 1983. The Polish Revolu- Devolution of Centrally Planned Econo- tion: Solidarity 1980-82. London: mies." Journal of Comparative Econom- Jonathan Cape. ics 15(2): 239-65. 117 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Fonner Soviet Union Comparative Economic Performance Transition Economies in Europe and The Nature and Causes of the Transitional Recession Asia." Post-Soviet Geography and Eco- nomics 37(5): 265-85. Blanchard, Olivier, and Michael Kremer. 1997. de Melo, Martha, Cevdet Denizer, and Alan "Disorganization." Quarterly Journal of Gelb. 1996. "From Plan to Market: Pat- Economics 112(4): 1091-127. terns of Transition." World Bank Eco- Calvo, Guillermo, and Fabrizio Coricelli. 1993. nomic Review 10(3): 397-424. "Output Collapse in Eastern Europe: The de Melo, Martha, Cevdet Denizer, Alan Gelb, Role of Credit." International Monetary and Stoyan Tenev. 2001. "Circumstances Fund Staff Papers 40(1): 32-52. and Choice: The Role of Initial Condi- Ericson, Richard E. 1999. "The Structural Bar- tions and Policies in Transition Econo- rier to Transition Hidden in Input-Output mies." World Bank Economic Review Tables of Centrally Planned Economies." 15(1): 1-31. Economic Systems 23(3): 199-224. EBRD (European Bank for Reconstruction and Gomulka, Stanislaw. 1998. "Output: Causes of Development). Various years. Transition the Decline and Recovery." Working Pa- Report. London. pers Series No. 8, Center for Social and IMF (International Monetary Fund). 2000. Economic Research, Central European World Economic Outlook 2000: Focus on University, Warsaw, Poland. Transition Economies. Washington, D.C. Kornai,Janos. 1993. "Transformational Recession: World Bank. 1996. World Development Report A General Phenomenon Examined through 1996: From Plan to Market. Washington, the Example of Hungary's Development." D.C. Economie Appliquee 46(2): 181-227. Rodrick, Dani. 1992. "Making Sense of the Soviet Trade Shock in Eastern Europe: A Framework and Some Estimates." In General Mario Blejer, Guillermo A. Calvo, Fabrizio Coricelli, and Alan H. Gelb, Aslund, Anders. 1995. How Russia Became a eds., Eastern Europe in Transition: From Market Economy. Washington, D.C.: Recession to Growth? World Bank Dis- Brookings Institution. cussion Paper No. 196, World Bank, Blanchard, Olivier. 1997. The Economics of Washington, D.C. Transition in Eastern Europe. Oxford, Roland, Gerard, and Thierry Verdier. 1999. England: Clarendon Press. "Transition and the Output Fall." Econom- Blanchard, Olivier, Rudiger Dornbusch, Paul ics of Transition 7(1): 1-28. Krugman, Richard Layard, and Lawrence Summers. 1991. Reform in Eastern Eu- rope. Cambridge, MA: MIT Press. Explaining Variation in Economic Refonn Outcomes Clague, Christopher, and Gordon C. Rausser, Bokros, Lajos. 2000. "Visegrad Twins' Diverg- eds. 1992. The Emergence of Market ing Path to Relative Prosperity. Compar- Economies in Eastern Europe. Cambridge, ing the Transition Experience of the Czech MA: Blackwell Publishers. Republic and Hungary." Conference at the . . 1 r | T) ^ | ~~~~~Conicelli, Fabrizio. 1998. Macroeconomic Poli- Czech National Bank, Prague. September. cies and the Development of Markets in de Melo, Martha, and Alan Gelb. 1996. "A Transition Economies. Budapest: Central Comparative Analysis of Twenty European University Press. 118 Selected Bibliographic Guide to the Political Economy of Transition Dabrowski, Marek, Stanislaw Gomulka, and Economic Research Macroeconomic An- Jacek Rostowski. 2000. "Whence Reform? nual 1994. Cambridge, MA: MIT Press. A Critique of the Stiglitz Perspective." Cen- Aslund, Anders, Peter Boone, and Simon tre for Social and Economic Research, Johnson. 1996. "How to Stabilize: Lessons Warsaw. Processed. From Post-Communist Countrles." Klaus, Vaclav. 1997. Renaissance: The Rebirth Brookings Papers on Economic Activity of Liberty in the Heart of Europe. Wash- 26(1): 217-313. ington, D.C.: Cato Institute. Dewatripont, Mathias, and Gerard Rolancd. Kornai, Janos. 1990. The Road to a Free 1992a. "Economic Reform and Dynamic Economy. Shifting from a Socialist System: Political Constraints." Review of Economi c The Example of Hungary. New York: W. Studies 59(4): 703-30. W Norton. . 1992b. "The Virtues of Gradualism and . Forthcoming. "Ten Years After 'The Road Legitimacy in the Transition to a Market to a Free Economy,' The Author Self-Evalu- Economy." Economic Journal 102(411): ation." In Boris Pleskovic and Nicholas 291-300. Stern, eds., Annual World Bank Conference on Deelopent Eonomcs. Wshinton,. 1995. "The Design of Reform Packages onD.C:lorld Ba onk under Uncertainty." American Economic D.C.: World Bank. Review 85(5): 1207-23. McMillan, John, and Barry Naughton. 1992. "Howllan, tohReform Barry Planne onoy Les- Fischer, Stanley, Ratna Sahay, and Carlos A. "How to Reform a Planned Economy: Les- Vegh. 1996a. "Stabilization and Growth sons from China." Oxford Review of Eco- in Transition Economies: The Early Expe- nomic Policy 8(1): 130-43. rience. " Journal of Economic Perspectives Poznanski, Kazimierz Z., ed. 1993. Stabilization 10(2): 45-66. and Privatization in Poland. Boston, MA: . 1996b. "Economies in Transition: The Kluwer Academic Publishers. Beginnings of Growth." American Eco- Qian, Yuan, Gerard Roland, and Chenggang Xu. nomic Review 86(2): 229-33. 1999. "WhyisChinaDifferentFromEast- Kornai, Janos. 1990. The Road to a Free ern Europe: Perspectives from Organiza- Economy. Shifting from a Socialist System: tion Theory." European Economic Review The Example of Hungary. New York, NY: 43(4-6): 1085-94. Norton. Sachs, Jeffrey. 1996. "The Transition at Mid Lipton, David, and Jeffrey D. Sachs. 1990. "Cre- Decade." American Economic Review Pa- ating a Market in Eastern Europe: The Case pers and Proceedings 86(2): 128-33. of Poland." Brookings Papers on Economic Stiglitz, Joseph E. 1999. "Whither Reform? Ten Activity 20(1): 75-147. Years of the Transition." Keynote address, McKinnon, Ronald. 1992. The Order of Eco- World Bank Annual Bank Conference on nomic Liberalization: Financial Control in Development Economics, Washington, the Transition to a Market Economy. Bal- D.C. timore, MD: John Hopkins University Press. Speed and Sequencing of Reforms Murphy, Kevin M., Andrei Shleifer, and Robert Aghion, Philippe, and Oliver Blanchard. 1994. W. Vishny. 1992. "The Transition to a "On the Speed of Transition in Central Market Economy: Pitfalls of Partial Re- Europe." In Stanley Fischer and J. form." Quarterly Journal of Economics Rotemberg, eds., National Bureau of 107(3): 889-906. 119 Transition-The First Ten Years: Analysis and Lessons for Eastem Europe and the Forner Soviet Union Murrell, Peter. 1992a. "Conservative Political Djankov, Simeon. 1999. "The Restructuring of Philosophy and the Strategy of Economic Insider Dominated Firms." Economics of Transition." East European Politics and Transition 7(2): 467-79. Societies 6(1): 3-16. Earle, John S., Roman Frydman, and Andrzej . 1992b. "Evolutionary and Radical Ap- Rapaczynski, eds. 1993. Privatization in the proaches to Economic Reform." Econom- Transition to a Market Economy. Budapest: ics of Planning 25(1): 79-95. Central European University Press. Sachs, Jeffrey. 1993. Poland's Jump to the Mar- Frydman, Roman, Cheryl W Gray, and Andrzej ket Economy. Cambridge, MA: MIT Press. Rapacszynski, eds. 1996. Corporate Gov- ernance in Central Europe and Russia, vols. . 1994. "Life in the Economic Emergency nBudapest: Central European Univer- Room." In John Williamson, ed., The Po- . litical Economy of Policy Reformn. Wash- sity Press. ington, D.C.: Institute for International Frydman, Roman, and others. 1993a. The Economics. Privatization Process in Central Europe. Budapest: Central European University Press. P t a r e r. 1993b. The Privatization Process in Rus- sia, Ukraine, and the Baltic States. Budapest: Aghion, Philippe, and Olivier Blanchard. 1998. Central European University Press. "On Privatization Methods in Eastern Eu- Lipton, David, and Jeffrey D. Sachs. 1990. rope and their Implications." Economics "Privatization in Eastern Europe: The Case of Transition 6(1): 87-99. o Transition 6(1): 87-99. of Poland." Brookings Papers on Economic Amsden, Alice, Jacek Kochanowicz, and Lance Activity 20(2): 293-341. Taylor. 1994. The Market Meets Its Match: Nellis, John. 1999. "Time to Rethink Restructuring the Economies of Eastern Privatization in Transition Economies?" Europe. Cambridge, MA: Harvard Univer- Finance and Development 36(2): 16-9. sity Press. Roland. Gerard. 1994. "On the Speed and Sequenc- Balcerowicz, Leszek, Cheryl Gray, and Iraj n fPiaiainadRsrcuig"Eo Hoh. ed. 198 nters Exi Prcsssig of Privatization and Restructuring." Eco- Hohi ed. 198 EnepieEi.rcse nomic Journal 104(426): 1158-68. in Transition Economies. Budapest: Cen- tral European University Press. Roland, Gerard, and Thierry Verdier. 1994. Black, Bernard S., Reinier Kraakman, and Anna "Privatization in Eastern Europe: Irrevers- Tarassova. 2000. "Russian Privatization ibility and Critical Mass Effects." Journal and Corporate Governance." Stanford of Public Economics 54(2): 161-83. Law Review 52(6): 1731-808. Stark, David, and Laszlo Bruszt. 1998. Postsocialist Pathways: Transforming Politics and Prop- Blasi, Joseph R., Maya Kroumova, and DouglaseryiEatCnaluop.amid,MA Kruse. 1997. Kremlin Capitalism: The ertymnEastCentralEurope.Cambidge,MA: Privatization of The Russian Economy. Ithaca, NY: Cornell University Press. Boycko, Maxim, Andrei Shleifer, and Robert Empirical Assessments of Privatization Vishny. 1995. Privatizing Russia. Cam- Blasi, Joseph, and Andrei Shleifer. 1996. "Cor- bridge, MA: MIT Press. . . porate Governance in Russia: An Initial . 1996. "A Theory of Privatization." Eco- Look." In Roman Frydman, Cheryl W. nomicJournal 106(435): 309-19. Gray, and Andrzej Rapacszynski, eds., 120 Selected Bibliographic Guide to the Political Economy of Transition Coporate Governance in Central Europe EBRD Working Paper No. 37, Europeain and Russia, vol. 2. Budapest: Central Eu- Bank for Reconstruction and Development, ropean University Press. London. Brown, David, and John Earle. 2000. "Privatization Gaddy, Clifford, and Barry Ickes. 1999. Russia's and Enterprise Restructuring in Russia: New Virtual Economy. Washington, D.C.: Evidence from Panel Data on Industrial En- Brookings Institution. terprises. " Working Paper; Russian-European Ivanova, Nadezhda, and Charles Wyplosz. 1999. Centre For Economic Policy, Moscow. "Arrears: The Tide that is Drowning Rus- Djankov, Simeon, and Peter Murrell. 2000. The sia." Working Paper, Russian-European Determinants of Enterprise Restructuring Centre for Economic Policy, Moscow. in Transition: An Assessment of the Evi- Johnson, Simon, Daniel Kaufmann, John dence. Washington, D.C.: World Bank. McMillan, and Christopher Woodruff. 2000. Earle, John, and Saul Estrin. 1998. "Privatization, "Why Do Firms Hide? Bribes and Unofficial Competition, and Budget Constraints: Dis- Activity After Communism." Journal ofPub- ciplining Enterprises in Russia." Working lic Economics 76(3): 495-520. Paper No. 128, Stockholm Institute of Tran- McKinnon. Ronald I. 1991. The Order of Eco- sition Economics, Stockholm. nomic Liberalization: Financial Control in Frydman, Roman, Cheryl W. Gray, Marek the Transition to a Market Economy. Balti- Hessel, and Andrzej Rapaczynski. 1997. more, MD: John Hopkins University Press. "Private Ownership and Corporate Perfor- Tanzi, Vito, ed. 1992. Fiscal Policies in Econo- mance: Evidence from Transition Econo- T i, it ed.n1992. FiscalgPolicies.i Ecn- mies." EBRD Working Paper No. 26, Eu- mies in Transition. Washington, D.C.: In- ropean Bank for Reconstruction and Development, London. Woodruff, David. 1999. Money Unmade: Bar- _____ 1999. "When Does Privatization Work? ter and the Fate of Russian Capitalism. The Impact of Private Ownership on Cor- Ithaca, NY: Cornell University Press. porate Performance in the Transition Economies." Quarterly Journal of Eco- Social Safety Nets and Reform nomics 114(4): 1153-92. Lieberman, Ira W., Stilpon S. Nestor, and Raj M. Commander, Simon, and Fabrizio Coricelli, eds. Desai, eds. 1997. Between State and Mar- 1995. Unemployment, Restructuring, and ket: Mass Privatization in Transition Econo- the Labor Market in East Europe and Rus- mies. Washington D.C.: The World Bank sia. Washington, D.C.: World Bank. and the OECD (Organisation for Economic Cook, Linda, Mitchell Orenstein, and Marilyn Co-operation and Development). Rueschemeyer, eds. 1999. Left Parties and Megginson, William E., and Jeffry M. Netter. Social Policy in Postcommunist Europe. 2001. "From State to Market: A Survey of Boulder CO: Westview Press. Empirical Studies on Privatization." Jour- Graham, Carol. 1995. "The Politics of Safety nal of Economic Literature 39(2): 321-89. Nets." In Larry Diamond and Marc ]F. Plattner, eds., Economic Reform and De- mocracy. Baltimore, MD: Johns Hopkins Barter, Arrears, and Noncash Payments University Press. Commander, Simon, and Christian Mumssen. Kornai, Janos. 1997. "Reform and the Welfare 1998. "Understanding Barter in Russia," Sector in the Post-Socialist Countries: A 121 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Former Soviet Union Normative Approach." In Joan Nelson, Technical Paper No. 494, World Bank, Charles Tilly, and Lee Walker, eds., Trans- Washington, D.C. forming Post-Communist Political Econo- Kaminski, Bartlomiej. 1999a. "Hungary: Foreign mies. Washington, D.C.: National Acad- Trade Issues in the Context of Accession emy Press. TaeIse nteCneto ceso to the EU." World Bank Technical Paper Kramer, Mark. 1997. "Social Protection Policies No. 441, World Bank, Washington, D.C. and Safety Nets in East-Central Europe: Dilemmas of the Post-Communist Trans- .etmntb n "Thre PofiFoesin DrcIn- formation." In Ethan Kapstein and Michael Madlamd ueTransi- Accession to the European Union." World Mandelbaum, eds., Sustaining the Tas-Bank Technical Paper No. 442, World tion: The Social Safety Net in Bank Washington, D.C. Postcommunist Europe. New York, NY: B Council on Foreign Relations. Nabli, Mustapha. 1999. "Financial Integration, Milanovic, Branko. 1998. Income, Inequality, Vulnerabilities to Crisis, and EU Accession and Poverty during the Transition from in Five Central European Countries." Planned to Market Economy. Washington, World Bank Technical Paper No. 439, D.C.: World Bank. World Bank, Washington, D.C. Nunberg, Barbara. 2000. "Ready for Europe: World Bank. 2000. Making Transition Work for Public Administration Reform and Euro- Everyone: Poverty and Inequality in Eu- rope and Central Asia. Washington, D.C. pean Union Accession in Central and East- ern Europe." World Bank Technical Paper No. 466, World Bank, Washington, D.C. Accession to the European Union Petkova, Ivanka, Elitsa Markova, Irena Mladenova, Begg, lain. 1999. "EU Investment Grants Re- and Ruslan Stefanova, eds. 2000. Regional view." World Bank Technical Paper No. Cooperation in Central and Eastern Europe. 435, World Bank, Washington, D.C. Sofia: Economic Policy Institute. Drulak, Petr, ed. 2001. National and European Tang, Helena, ed. 2000. Winners and Losers of Identities in EU Enlargement: Views from EU Integration: Policy Issues for Central Central and Eastern Europe. Prague: In- and Eastern Europe. Washington, D.C.: stitute of International Relations. Bertelsmann Foundation and the World European Commission and the World Bank. Bank. 1998. European Union Accession: The Ulgenerk, Esen, and Leila Zlaoui. 2000. "From Challenges for Public Liability Manage- Transition to Accession: Developing Stable ment in Central Europe. Washington, D.C. and Competitive Financial Markets in Bul- _____ 2000. European Union Accession: Op- garia." World Bank Technical Paper No. portunities and Risks in Central European 473, World Bank, Washington, D.C. Finances. Washington, D.C. World Bank. 1997. "Poland-Reform and Funck, Bernard, and Lodovico Pizzati, eds. 200 1. Growth on the Road to the EU." Country Labor, Employment, and Social Policies in study, Washington, D.C. the EU Enlargement Process. Washington, . 1998. "Slovak Republic-A Strategy for D.C.: World Bank. Growth and European Integration." Coun- Garibaldi, Pietro, Mattia Makovec, and try study, Washington, D.C. Gabriella Stoyanova. 2001. "From Tran- _ . 1999a. "Slovenia-Economic Transfor- sition to EU Accession: The Bulgarian La- mation and EU Accession." Country study, bor Market during the 1990s." World Bank Washington, D.C. 122 Selected Bibliographic Guide to the Political Economy of Transition . 1999b. "Hungary-On the Road to the Compared. Budapest: Central European European Union." Country study, Wash- University Press. ington, D.C. Hellman, Joel S. 1998. "Winners Take All: The . 1999c. "Estonia-Implementing the EU Politics of Partial Reform in Postcommunist Accession Agenda." Country study, Wash- Transitions." World Politics 50(2): 203-34. ington, D.C. Offe, Claus. 1991. "Capitalism by Democratic . 1999d. "Czech Republic-Toward EU Design? Democratic Theory Facing the Accession." Country study, Washington, Triple Transition in East Central Europe." D.C. Social Research 58(4): 893-902. . 2000. "Progress Toward the Unification Przeworski, Adam. 1991. Democracy and the of Europe." Paper for the World Free of Market: Political and Economic Reforms in Poverty series, Washington, D.C. Eastern Europe and Latin America. Cam- .____ 2001. "Bulgaria-The Dual Challenge of bridge, MA: Cambridge University Press. Transition and Accession." Washington, D.C. Roland, Gerard. 2000. Transition and Econom- ics: Politics, Markets, and Firms. Cam- The Politics of Economic Reform in Transition bridge, MA: MIT Press. Economies Shleifer, Andrei, and Daniel Treisman. 200C0. Balcerowicz Leszek 1994"UndeWithout A Map: Political Tactics and Eco- Balcerowicz, Leszek. 1994. "Understanding nmcRfr nRsi.Cmrde A Postcommunist Transitions." Journal of miT Press. Democracy 5(4): 75-89. MIT Press. .1995. Socialism, Capitalism, Transforma- tion. Budapest: Central European Univer- Corruption, State Capture, and the Rule of Law sity Press. Frye, Timothy. 1997. "Contracting in the Bunce, Valerie. 1999. "The Political Economy Shadow of the State: Private Arbitration of Post Communism." Slavic Review 58(4): Commissions in Russia." In Jeffrey D. 756-93. Sachs and Katharina Pistor, eds., The Rule Ekiert, Grzegorz, and Jan Kubik. 1998. "Con- of Law and Economic Reform in Russia. tentious Politics in New Democracies: Hun- New York, NY: Westview Press. gary, the former East Germany, Poland, and Frye, Timothy, and Andrei Shleifer. 1997. "The Slovakia." World Politics 50(4): 547-81. Invisible Hand andthe GrabbingHand." The Elster, Jon, Claus Offe, Ulrich K. Preuss, Frank American Economic Review 87(2): 354-8. Boeuher, Ulrike Goetting, and Friedbert W Hay, Jonathan R. and Andrei Shleifer. 1998. Rueb. 1998. Institutional Design in Post- "Private Enforcement of Public Laws: A communist Societies: Rebuilding the Ship Theory of Legal Reform." American Eco- at Sea. Cambridge, MA: Cambridge Uni- nomic Review 88(2): 398-403. versity Press. Hendley, Kathryn, Peter Murrell, and Randi Fish, M. Stephen. 1998. "The Determinants of Ryterman. 2000. "Law, Relationships, and Economic Reform in the Post-Communist Private Enforcement: Transactional Strat- World." East European Politics and Soci- egies of Russian Enterprises." Europe-Asia eties 12(1): 31-78. Studies 52(4): 627-56. Greskovits, Bela. 1998. The Political Economy Hellman, Joel S., Geraint Jones, and Daniel of Protest and Patience: East European Kaufmann. 2000. "Seize the State, Seize and Latin American Transformations the Day: State Capture, Corruption, and 123 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Fonner Soviet Union Influence in Transition." Policy Research Party Cooperation. Cambridge, MA: Cam- Working Paper No. 2444, World Bank, bridge University Press. Washington, D.C. Linz, Juan J., and Alfred C. Stepan. 1996. Prob- Johnson, Simon, Daniel Kaufman, and Andrei lems of Democratic Transition and Con- Shleifer. 1997. "The Unofficial Economy solidation: Southern Europe, South in Transition." Brookings Papers on Eco- America, and Post-Communist Europe, nomic Activity 27(2): 159-239. Part IV. Baltimore, MD: Johns Hopkins Johnson, Simon, Daniel Kaufmann, and Pablo University Press. Zoido-Lobaton. 1998. "Government in Roeder, Philip G. 1994. "Varieties of Post-So- Transition: Regulatory Discretion and the viet Authoritarian Regimes." Post-Soviet Unofficial Economy." American Economic Affairs 10(1): 61-101. Review 88(2): 387-92. Johnson, Simon, John McMillan, and C. Woo- G druff. 2000. "Entrepreneurs and the Or- overance dering of Institutional Reform-Poland, Elster, Jon, Claus Offe, and Ulrich K. Preuss. Slovakia, Romania, Russia, and Ukraine 1998. Institutional Design in Post-Com- Compared." Economics of Transition munist Societies. New York, NY: Cam- 8(1): 1-36. bridge University Press. La Porta, Rafael, Andrei Shleifer, Robert W. Offe, Claus. 1991. "Capitalism by Democratic Vishny, and Florencio Lopez-de-Silanes. Design? Democratic Theory Facing the 1998. "Law and Finance." Journal of Po- Triple Transition in East Central Europe." litical Economy 106(6): 1113-55. Social Research 58(4): 865-92. Sachs, Jeffrey D., and Katharina Pistor, eds. 1997. Przeworski, Adam, and others. 1995. Sustain- The Rule of Law and Economic Reform able Democracy. Cambridge, MA: Cam- in Russia. Boulder, CO: Westview Press. bridge University Press. Shleifer, Andrei, and Robert Vishny. 1998. The Grabbing Hand: Government Pathologies GanTir Cures Ca,rmbig, Ma:harvard Nationalism, Ethnic Conflict, and Secession and Tneir Cures. Cambridge, MA: Harvard University Press. Fearon, James D., and David D. Laitin. 1996. "Ex- Volkov, Vladimir. 1999. "Violent Entrepreneur- plaining Interethnic Cooperation." Ameri- ship in Post-Communist Russia." Europe- can Political Science Review 90(4): 715-35. Asia Studies 51(5): 741-54. Laitin, David D. 1998. Identity in Formation. World Bank. 2000. Anticorruption in Transition: Ithaca, NY: Cornell University Press. A Contribution to the Policy Debate. Offe, Claus. 1997. "Ethnic Politics in East Eu- Washington, D.C. ropean Transitions." Varieties of Transi- tions: The East European and East Ger- man Experience. Cambridge, MA: MIT Classifying Political Regimes in Transition Press. Kitschelt, Herbert, Zdenka Mansfeldova, Vujacic, Veljko. 1996. "Historical Legacies, Na- Radoslaw Markowski, and Gabor Toka. tionalist Mobilization, and Political Out- 1999. Post-Communist Party Systems: comes in Russia and Serbia: A Weberian Competition, Representation, and Inter- View." Theory and Society 25(6): 763-801. 124 References The word "processed" describes informally reproduced works that may not be commonly available through library systems. Alesina, Alberto, and Howard Rosenthal. 1995. Berg, Andrew, Eduardo Borensztein, Ratna Partisan Politics, Divided Government, Sahay, and Jeromin Zettelmeyer. 1999. and the Economy. New York, NY: Cam- "The Evolution of Output in Transition bridge University Press. Economies: Explaining the Differences." Alesina, Alberto, Nouriel Roubini, and Gerald IMF Working Paper No. WP/99173, In- Cohen. 1997. Political Cycles and the ternational Monetary Fund, Washington, Macroeconomy. Cambridge, MA: MIT D.C. Press. Campos, Nauro F, and Fabrizio Coricelli. 2000. "Growth in Transition: What We Know, Aslund, Anders. 2001. "The Myth of Output What We Don't, and What We Should." Collapse after Communism." Carnegie Centre for Economic Policy Research, Endowment Working Paper No. 18, London. Carnegie Endowment for International Pacne,ie Wmng t fon,aDlC. Carlin, Wendy, Steven Fries, Mark Schaffer, and Paul Seabright. 2001. "Competition and Aslund, Anders, Peter Boone, and Simon Enterprise Performance in Transition Johnson. 1996. "Howto Stabilize: Lessons Economies: Evidence from a Cross- from Post-Communist Countries." Country Survey." EBRD Working Paper Brookings Papers on Economic Activity No. 63, European Bank for Reconstruction 26(1): 217-311. and Development, London. Balcerowicz, Leszek. 1995. Socialism, Capital- Castanheira, Micael, and Vladimir Popov. 1999. ism, Transformation. Budapest: Central "Framework Paper on the Political European University Press. Economy of Growth in Russia and 125 Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Forner Soviet Union Central and Eastern European Countries." Available at http://www.europa.eu.int/ Paper for the Global Development Net- comm/eurostat. work conference, World Bank, Washing- Fernandez, Raquel, and Dani Rodrik. 1991. "Re- ton, D.C. November. ton, D.C. November. sistance to Reform: Status Quo Bias in the Caves, R. 1998. "Industrial Organization and Presence of Individual-Specific Uncer- New Findings on the Turnover and Mobil- tainty. " American Economic Review 81 (5): ity of Firms." Journal of Economic Litera- 1146-55. ture 36(4): 1947-82. Fischer, Stanley, and Alan Gelb. 1991. "The Pro- de Melo, Martha, and Alan Gelb. 1996. "A Com- cess of Socialist Economic Transforma- parative Analysis of Twenty Transition tion." Journal of Economic Perpectives Economies in Europe and Asia" Post-Soviet 5(4): 91-106. Geography and Economics 37(5): 265-85. Fischer, Stanley, Ratna Sahay, and Carlos A. de Melo, Martha, Cevdet Denizer, and Alan Veight. 1998. "How Far is Eastern Europe Gelb. 1996. "From Plan to Market: Pat- from Brussels?" IMF Working Paper No. terns of Transition." World Bank, Wash- WP/98/53, International Monetary Fund, ington, D.C. Washington, D.C. de Melo, Martha, Cevdet Denizer, Alan Gelb, Frye, Timothy, and Joel Hellman. 2001. "Political and Stoyan Tenev. 1997. "Circumstances Data-Base of Post-Communist Countries." and Choice: The Role of Initial Conditions World Bank, Washington, D.C. Processed. and Policies in Transition Economies." Freedom House. Various years. Freedom in the Policy Research Working Paper No. 1866, World: The Annual Survey of Political World Bank, Washington, D.C. Rights and Civil Liberties. New York. Dethier, Jean-Jacques, Hafez Ghanem, and Edda Available at www.freedomhouse.org. Zoli. 1999. "Does Democracy Facilitate the Gaddy, Clifford, and Barry Ickes. 1999. Russia's Economic Transition?" World Bank, Wash- Virtual Economy. Washington, D.C.: ington, D.C. Processed. Brookings Institution. Djankov, Simeon, and Peter Murrell. 2000. The Haggard, Stephan, and Steven Webb. 1993. Determinants of Enterprise Restructuring "What Do We Know about the Political in Transition: An Assessment of the Evi- Economy of Economic Policy Reform?" dence. Washington, D.C.: World Bank. World Bank Policy Research Observer Easterly, William. 2000. "It's Been a Rough Ride, 8(2): 143-67. Toto, but I think We're Back in Kansas." Havrylysyhyn, Oleh, and Ron van Rooden. World Bank, Washington, D.C. Processed. 1999. "Institutions Matter in Transition, Eichengreen, Barry, and Christof Ruehl. 2000. but So Do Policies." Fifth Dubrovnik Con- "Financial Sector Developments in Transi- ference on Transition, Dubrovnic. June. tion Economies." University of California Hellman, Joel. 1998. "Winners Take All: The at Berkeley. Processed. Politics of Partial Reform." World Politics EBRD (European Bank for Reconstruction and 50(2): 203. Development). 1999. Transition Report Hellman, Joel S., Geraint Jones, and Daniel 1999. London. Kaufmann. 2000. "Seize the State, Seize the _____ 2000. Transition Report 2000. London. Day: An Empirical Analysis of State Cap- ture and Corruption in Transition Econo- Eurostat. 1998. New Enterprises in CEC in mies." Policy Research Working Paper No. 1997. Studies and Research, theme 4. 2444, World Bank, Washington D.C. 126 References Heilman, Joel S., Geraint Jones, Daniel Ukraine Compared." Economics of Tran- Kaufmann, and Mark Schankerman. 2000. sition 8(1): 1-36. "Measuring Governance Corruption and Johnson, Simon, Peter Boone, Alasdair Breach, State Capture: How Firms and Bureaucrats and Eric Friedman. 2000. "Corporate Shape the Business Environment in Tran- G i t sition Economies." Policv Research Work- . i Journal of Financial Economics 58(1-2): ing Paper No. 2312, World Bank, Wash- 141-86. ington D.C. Johnson, Simon, Rafael La Porta, Florencio Hernandez-Cata, Ernesto. 1997. "Growth and Lop,dSim ane R an Anrei hloenifr . . ~~~Lopez-de-Silanes, and Andrei Shleifer. Liberalization during the Transition from ,, ~~~~~~~2000. "Tunneling." Harvard Institute of Plan to Market." IMF Staff Papers 44(4): Economic Research Discussion Paper No. 405-29. 1887, Harvard Institute of Economic Re- Heybey, Berta, and Peter Murrell. 1999. "The search, Cambridge, MA. Relationship between Economic Growth Klugman, Jeni. 1999. "Financing and Gover- and the Speed of Liberalization during nance of Education in Central Asia." Transition." Journal of Policy Reform 3(2): MO CT-MOST. Economic Policy in Tran- 121-37. sitional Economies 9(4): 423-42. Hirschman, Albert 0. 1963. journeys Towards Kornai, Janos. 1986. "The Soft Budget Con- Progress. New York, NY: The Twentieth straint." Kyklos 39(1): 3-30. Century Fund. Levine, Ross, and Sara Zervos. 1998. "Stock IMF (International Monetary Fund), World Mre,Bns n cnmcGot. Bank, OECD (Organisation for Economic Market, Banks, and Economic Growth." Co-opratio and eveloment) andAmerican Economic Review 88(3): 537-- Co-operation and Development), and 5 EBRD (European Bank for Reconstruction 58. and Development). 1991. A Study of the Maddison, Angus. 1982. Phases of Capitalist Soviet Economy, vol. 3. Washington, D.C. Development. Oxford, U.K.: Oxford Uni- Jensen, Michael C., and William H. Meckling. versity Press. 1976. "Theory of the Firm: Managerial Martin, Ricardo. 2000. "Revisiting Regression Behavior, Agency Costs, and Ownership Analysis: Initial Conditions, Policies, andl Structure." Journal of Financial Econom- Growth." Background paper, World Bank, ics 3(4): 305-60. Washington D.C. Johnson, Simon, and Andrei Shleifer. 2001. Orenstein, Mitchell. 2001. Out of the Red:Build- "Privatization and Corporate Gover- ing Capitalism and Democracy in Post- nance." Paper prepared for the 12th An- Communist Europe. Ann Arbor, MI: Uni- nual East Asian Seminar on Economics, versity of Michigan Press. June 28-30, National Bureau of Economic Pinto, Brian, Vladimir Drebentsov, and Research, Cambridge, MA. Alexander Morozov. 2000. "Give Macro- Johnson, Simon, Daniel Kaufmann, and Andrei economic Stability and Growth in Russia Shleifer. 1997. "The Unofficial Economy a Chance: Harden Budgets by Eliminating in Transition." Brookings Papers on Eco- Non-Payments." Economics of Transition nomic Activity 27(2): 159-239. 8(2): 297-324. Johnson, Simon, John McMillan, and Christo- Popov, Vladimir. 1998. "Will Russia Achieve Fast pher Woodruff. 2000. "Entrepreneurs and Economic Growth?" Communist Econo- the Ordering of Institutional Reform-Po- mies and Economic Transformation 10(4): land, Slovakia, Romania, Russia, and 421-35. 127 Transffion-The hrst Ten Years: Analysis and Lessons ior Eastern Europe and the Former Soviet Union . 1999. "Shock Therapy versus Gradual- Harvard University and University of isrn: The End of the Debate." Carleton Michigan. Processed. University, Ottawa. Processed. Tarr, David. 1994. "The Terms of Trade Effects Przeworski, Adam. 1991. Democracy and the of Moving to World Prices on Countries Market: Political and Economic Reforms of the FSU." Journal of Comparative Eco- in Eastern Europe and Latin America. New nomics 18(1): 1-24. York, NY: Cambridge University Press. UNICEF-ICDC (United Nations Children's Fund- Roubini, Nouriel, and Jeffrey Sachs. 1989. "Po- International Child Development Centre). litical and Economic Determinants of Bud- 1999. "TransMONEE." Database. get Deficits in the Industrial Democra- World Bank. 1998. World Development Report cies." European Economic Review 33(5): 1998/99: Knowledge for Development. 903-33. Washington, D.C. Ruehl, Christof, and Viatcheslav Vinogradov. . 1999a. Czech Republic: Capital Market 2001. "A Simple Model of the Transition Review. Washington, D.C. from Central Planning." Center for Eco- nomic Research and Graduate Education .1999b."Russia:RegionalEducationStudy. of Charles University and the Economic Human Development Sector Unit, Europe Institute of the Academy of Sciences of the and Central Asia Region, Washington, D.C. Czech Republic, Prague. Processed. .2000a. Poverty Reduction and the World Selowsky, Marcelo, and Ricardo Martin. 1998. Bank: Progress in Fiscal 1999. Washing- "Policy Performance and Output Growth ton, D.C. in the Transition Economies." American . 2000b. Making Transition Work for Ev- Economic Review-Papers and Proceedings eryone: Poverty and Inequality in Europe 87(2): 350-53. and Central Asia. Washington D.C. Shleifer, Andrei, and Robert Vishny. 1997. "A . 2000c. Anticorruption in Transition: A Survey of Corporate Governance." Jour- Contribution to the Policy Debate. Wash- nal of Finance. 52(2): 737-83. ington, D.C. Shleifer, Andrei, and D. Wolfenson. 2000. "In- . 2001. World Development Indicators, vestor Protection and Equity Markets." 2001. Washington, D.C. 128 N\li- has economic growth in some transition economies of Eastern Europe and the former Soviet IJnimn been stronger than in others? If economic reforms bring clear ', ncfi t for countries in transition, why have some go \ 1cuii nient, been so reluctant to 1i,ccct them? How 1hotld policy advice o ttiv V cd to these countries be modified in the light of experience and today's conditions? Transition- Tfhe First Ten Years draws on the World Bank's operational experience and the extensive literature on transition to help address these questions. This report looks at the policy and institutional conditions that encourage the growth of new firms in transition economies while imnposing financial discipline on the old firms inherited from the socialist past, without granting special favors to either. While emphasizing the importance of market- oriented policy reforms, the report also examines political strategies to push the reform process forward in different transition countries. The report is aimed at policymakers and think tanks in transition countries, external donors and advisors, and all those in the international JI ii community interested in the policy and institutional th. iL.A nw fWrn- the countries of Eastern 7i : and the former Soviet Union. "Successful transition economies of Eastern a.nd the forner Soviet Union simiultaneously followed two key sets of .-. -1 - 1- ' - - constraints o¢n old firms and encounr.ic-ir o enitry of new trmsn swithiouit, however, provtiid iog, T favors for the latter. One without the otlher- is not a. '. A ciriti< al r1al:iMRui dose of both is also needed to - !' winners fLCI tdhe early stages of Lc r,i frlyl derailing further progress. Policies to 1 t ci, nn-e a-- snc as key social services and transfers and .e S training, are critical to -I' l.1. . ,!' _the To t---,i uenrv -1r a-- i a- of being stuck ii a low-level reform-.L l. r1x-1-`(-vC (JI e which countries that have not made muah -l- I J t advised to consider." Federation i a- SLZ' i in Transiteio?-. iV#lvor:couv. THE WORLD BANK 1818 H Street, Nl.W. D' :I lW | K i Wial.thiI i ittI, D.C. 20433 U.S.A. ,. i.0 ii i I Il d Ipim uic: l' 2 '-477-1234 i Facsimile: 202-477-6391 -- . Internet: www.wvorldbank.org E-mail: I Ccd 1IN lit d 1-i.l 11 li l 1SBN 0-8213-5038-2