Transition The First Ten Years - ekonomiks.weebly.comekonomiks.weebly.com/uploads/5/5/1/5/5515573/world_bank_2002.pdf · Transition The First Ten Years ... Learning from China?

Transition

The First Ten Years

Analysis and Lessons for Eastern Europeand the Former Soviet Union

Transition

The First Ten Years

Analysis and Lessons for Eastern Europeand the Former Soviet Union

THE WORLD BANKWashington, D.C.

© 2002 The International Bank for Reconstruction and Development/The World Bank1818 H Street, NWWashington, DC 20433

All rights reserved.

1 2 3 4 5 05 04 03 02

The findings, interpretations, and conclusions expressed here are those of the author(s) and do notnecessarily reflect the views of the Board of Executive Directors of the World Bank or the governmentsthey 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 thepart of the World Bank any judgment of the legal status of any territory or the endorsement or acceptanceof such boundaries.

Rights and Permissions

The material in this work is copyrighted. No part of this work may be reproduced or transmitted in anyform or by any means, electronic or mechanical, including photocopying, recording, or inclusion in anyinformation storage and retrieval system, without the prior written permission of the World Bank. TheWorld 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 theCopyright 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 Officeof the Publisher, World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422,e-mail [email protected]

ISBN 0-8213-5038-2

Library of Congress Cataloging-in-Publication Data has been applied for.

v

Foreword ..........................................................................................................................ix

Acknowledgments .............................................................................................................xi

Overview ........................................................................................................................xiiiThe Quest for Growth: Promoting Discipline and Encouragement ................................. xvThe Flipside: Protection and Discouragement ............................................................... xviiShading the Classification ............................................................................................. xviiNew Enterprises Spur Economic Growth ..................................................................... xviiiWhen Is Transition Over? ..............................................................................................xixDo Central Europe and the Baltics Point the Way Forward? .......................................... xixCan We Have It Both Ways: Encouragement without Discipline? ................................... xxLearning from China? ....................................................................................................xxiInstitutions Are Important, but So Are Policies ..............................................................xxiThe Political Economy of Discipline and Encouragement ............................................. xxiiShifting Policy Priorities to Account for Experience and New Conditions ................... xxivConclusions ............................................................................................................... xxviiiAnnex 1. Discipline and Encouragement: The Reform Agenda ................................. xxix

Part 1. The First Decade in Transition ................................................................. 1

1. How Did Transition Economies Perform? ..................................................................3Output Fell Sharply ..........................................................................................................3Industry Shrank—Services Grew....................................................................................... 5Private Enterprises Overtook the State Sector ...................................................................6Exports Rose—Moving Toward Industrial Countries ....................................................... 6Poverty Increased Sharply .................................................................................................6

1Contents

Contents

vi

2. Explaining Variation in Output Performance ...........................................................11Did Initial Conditions Affect Performance? .................................................................... 11External Economic Shocks Delayed Recovery .................................................................13Policies—Do They Matter? .............................................................................................13What Initial Conditions Matter and When Do They Matter? ......................................... 15What If Policies Themselves Are Endogenous? ...............................................................16Does the Speed of Reform Matter? .................................................................................16Annex 2.1. Summary of Cross-Country Empirical Literature on Growth in

Transition Economies ...............................................................................................16Annex 2.2. Additional Empirical Analysis .................................................................... 19

Part 2. Policy and Institutional Challenges Ahead ..........................................21

3. The Quest for Growth .............................................................................................23A Tale of Two Approaches .............................................................................................26The Associated Fiscal Adjustment… ...............................................................................29…And the Role of Labor Markets ..................................................................................30

4. Discipline and Encouragement .................................................................................33New Enterprises Drive the Transition .............................................................................39A Small Number of High-Productivity Small Enterprises Is Not Enough ........................ 41Can We Have It Both Ways, That Is, Protecting Old Inherited Enterprises and

Encouraging New Enterprises? .................................................................................45Annex 4.1. Assumptions for Small and New Enterprises .............................................. 48Annex 4.2. Implications of the Higher Productivity of SMEs ....................................... 50

5. Imposing Discipline..................................................................................................53Soft Budget Constraints Can Create Macroeconomic Crises ........................................... 53Nonpayments Weaken the Incentives for Efficiency and Restructuring ........................... 54Exit Mechanisms—Implement Now, Revise Later .......................................................... 56Competition Is Linked to Innovation and Growth .......................................................... 56

6. Extending Encouragement .......................................................................................59Corruption and Anticompetitive Practices Mar the Investment Climate ......................... 59Enterprises Lack Confidence in Legal and Judicial Institutions ....................................... 61Financial Deepening Is Slow but Progressing .................................................................. 62Privatization Attracts Foreign Direct Investment, and Positive Spillovers Follow ............ 67Tax Reform: Broadening the Base and Lowering Rates ................................................... 67Intergovernmental Fiscal Relations Supporting Discipline and Encouragement .............. 68

7. Privatization: Lessons and Agenda for the Future ....................................................71Traditional Privatization or Rapid Privatization? ............................................................73Why Countries Did What They Did ...............................................................................73Were Vouchers a Mistake and Other “What Ifs” ............................................................78Summarizing Lessons ......................................................................................................79

8. Supportive Social Policies .........................................................................................81Reforming Pension Systems ............................................................................................81

Contents

vii

Social Assistance Should Protect Children and the Most Destitute, Adding Moreas Budgets Allow ......................................................................................................83

Severe Cuts Have Compromised the Quality of Education ............................................. 84Containing 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 .................................91Who Wins and Who Loses? ............................................................................................92The Government Must Be Credible and Able to Constrain Oligarchs and Insiders ......... 94

10. Classifying Political Systems in Transition ...............................................................97Competitive Democracies Have High Political Contestability… ..................................... 99…and High Government Turnover ...............................................................................101

11. Political Systems Influence the Choice of Economic Reforms .................................103Political Systems Create Rent-Seeking Opportunities .................................................... 105How Do Political Systems Affect Economic Reform? ...................................................107

12. Confronting the Political Challenge .......................................................................111For Concentrated Political Regimes, Mobilizing Potential Winners .............................. 112For War-Torn Political Systems, Restoring Stability and Reducing Uncertainty............. 114For Noncompetitive Political Systems, Taking Advantage of State Capacity ................. 115For Competitive Democracies, Using Momentum to Build Coalitions for Reform ........ 115Conclusion ...................................................................................................................116

Selected Bibliographic Guide to the Political Economy of Transition .............................117

References ......................................................................................................................125

Boxes1 Increased Inequality ............................................................................................................. xiv1.1 Limits of GDP Statistics for Transition Economies .................................................................. 82.1 The Regional Impact of the Global Financial Crisis and Recovery ........................................ 133.1 The Business Environment and Enterprise Performance Survey ............................................. 243.2 The Problem of Tunneling .....................................................................................................274.1 Can the CIS Learn from China’s Reform Experience? ........................................................... 354.2 The German Experience ........................................................................................................374.3 Belarus ..................................................................................................................................465.1 External Debt and Fiscal Sustainability in the Low-Income CIS Countries ............................ 556.1 Reducing the Cost of Entry and Doing Business in Armenia ................................................. 607.1 Historical Counterfactuals: Mass Privatization in Russia ...................................................... 76

Figures1 Winners and Losers from Reform .......................................................................................xxii1.1 Changes in Real Output, 1990–2001 ......................................................................................41.2 Output Growth Rates, 1990–2001..........................................................................................42.1 Progress in Policy Reform, 1990s ..........................................................................................14

Contents

viii

3.1 Productivity Distribution of Old, Restructured, and New Enterprises ................................... 243.2 Performance of Old and New Enterprises, 1996–99.............................................................. 264.1 Private Sector Share in GDP, 1999 ........................................................................................404.2 Share of Employment in Small Enterprises, 1989–98 ............................................................414.3 Share of Value Added in Small Enterprises, 1989–98 ............................................................414.4 Value Added per Employee in Small Enterprises, 1998 .......................................................... 424.5 Index of GDP and Shares of Value Added and Employment Accounted for by

Small Enterprises, 1989–98 ............................................................................................434.6 Employment and GDP, 1990–98 ...........................................................................................444.7 Soft Budget Constraints and Employment in Small Enterprises, 2000 ................................... 48A4.1 Factor Price Frontier: SMEs and the Rest of the Economy .................................................... 50A4.2 Efficiency Gain from 10 Percent Factor Reallocation to the SME Sector ............................... 516.1 Insecurity of Property Rights in Transition Economies, 1999 ................................................ 616.2 Quality of Legal Drafting in Transition Economies, 1999 ..................................................... 626.3 Quality of Judiciary in Transition Economies ........................................................................636.4 Domestic Credit by Deposit Money Banks to the Private Sector, 1998 .................................. 646.5 Stock Market Capitalization and Per Capita Income, 1998 ................................................... 656.6 Operating Costs to Total Assets in the Banking Sector, 1997 ................................................ 656.7 Interest Rate Spreads, 1998 ...................................................................................................666.8 Cumulative Foreign Direct Investment Per Capita and Employment in Small

Enterprises, 1998 ............................................................................................................688.1 Public Expenditures on Education in Transition Economies, 1998 ........................................ 858.2 Health Expenditures, 1998 ...................................................................................................869.1 Winners and Losers from Reform .........................................................................................9310.1 Classifying Political Systems in Transition Economies, 1990–99 ........................................... 9810.2 Veto Points Index, 1989–99 ................................................................................................10010.3 Main Political Executive Turnovers, 1989–99 .....................................................................10211.1 Political Systems and Economic Reform Outcomes, 2000 ...................................................10411.2 State Capture Index, 1999...................................................................................................106

Tables1.1 The Transition Recession ........................................................................................................51.2 Composition of Output, 1990–91 and 1997–98 .....................................................................61.3 Private Sector Growth, 1990s..................................................................................................61.4 Export Growth and Destination, 1990s ..................................................................................71.5 Main Recipients of Foreign Direct Investment, 1992–99 ......................................................... 71.6 Average Poverty Rates, 1990 and 1998 ...................................................................................81.7 Changes in Inequality during the Transition, Various Years .................................................... 9A2.1 Regression of Average Growth on Initial Conditions, 1990–99 ............................................. 20A2.2 Regression of Annual Output Growth on Policies and Initial Conditions Allowing

for Differential Effects Early in Transition, 1990–99 ...................................................... 204.1 SMEs Have Higher Labor Productivity, 1998 ....................................................................... 42A4.1 Differences between New Enterprises and Small Enterprises, 1995 and 1998 ....................... 497.1 Methods of Privatization of Medium-Sized and Large Enterprises ........................................ 758.1 Reform Options for Social Protection Programs.................................................................... 838.2 Student-Teacher Ratios in Basic Education, 1990 and 1997 .................................................. 85

ix

1Foreword

This study looks at lessons to be drawn from the ten-year experience of the transition countriesin Eastern Europe and the former Soviet Union in the period 1991 to 2000. The World Bank’sWorld 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 andsustained reforms are important for recovery of growth and should be accompanied by social policiesdesigned to protect the most vulnerable groups until growth takes hold. It highlighted the need tocreate 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 growthperformance across countries has intensified. Poverty and income inequality in some countries haveincreased to levels not foreseen earlier. In many countries, reform efforts that started in the early1990s have been interrupted and in some cases even stalled. As a result, output recovery in some of thetransition 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. Atthe 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 thevariability in the recovery of output thereafter. Beyond this, important new lessons highlight some keytradeoffs 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- andmedium-size enterprises, in generating economic growth and in creating employment. The growth ofnew firms depends in part on direct policies to encourage entry—what this report calls the encourage-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 beenput 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-largeenterprises that remain from pretransition days. Soft budget constraints that allow these enterprises tonot 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.

Foreword

x

The combined encouragement-and-disciplinestrategy that this study proposes is crucial to re-allocate assets to more productive uses and toprovide economic space for new emerging firms.This perspective of encouragement and disciplineallows the report to shed light on issues such asprivatization and fiscal policy, which have beenmajor elements of economic reform in transitioncountries. Privatization is important to the ex-tent that it facilitates hard budget constraints onold enterprises and creates incentives for produc-tion and innovation rather than asset strippingand rent seeking in new enterprises.

A second lesson concerns the need to developor strengthen legal and regulatory institutions tooversee the management and governance of en-terprises, both those in the new private sector andthose remaining under state ownership. In coun-tries where direct sales of assets to strategic inves-tors—a preferred method of privatization—wasunavailable, policymakers were confronted withthe often difficult choice between privatization toineffective owners in a context of weak corporategovernance on the one hand and continued stateownership until strategic investors could be foundon the other. However, continued state ownershipdoes not lead to efficient stewardship of enter-prise assets unless there is a political commitmentto transparent privatization outcomes and a mini-mum institutional capacity to prevent asset strip-ping by managers during the intervening period.In either event, rules to protect minority share-holders; rules against insider deals and conflictsof interest; adequate accounting, auditing, and dis-closure standards; and takeover, insolvency, andcollateral legislation, together with developmentof enforcement capacity, are key to preventingasset stripping that reduces the true long-termvalue and competitiveness of a firm.

A third lesson involves the recognition thatwinners from the early stages of reforms such asliberalization and privatization may oppose sub-sequent reform steps when these reduce theirinitially substantial, but potentially temporarybenefits or rents. These winners will tend to re-sist reforms such as further trade liberalization;entry of new competitors, including foreign di-rect investment; and legislation protecting mi-nority shareholders and creditors to the extentthey reduce such rents. Furthermore, if the rentsare large as a fraction of total gross domesticproduct, which is usually the case in naturalresource- and energy-rich countries, these earlywinners may capture the state and force theeconomy into a trap of a low-level reform equi-librium. Understanding how such reform trapsarise and how to break out of them is an impor-tant area of inquiry addressed by this study. Tothe extent underlying political economy consid-erations permit a reform-minded team room tomaneuver, fiscal policy can play an importantrole here, by redirecting support away from ail-ing enterprises and toward worker training andseverance payments; by divesting social assetssuch as housing, child care, and health facilitiesfrom enterprises to governments; and by main-taining the high levels of human capital withwhich countries entered transition.

The experience of transition from centrallyplanned to a market economy is an historicallyunprecedented process, and one that is by nomeans finished in many countries in EasternEurope and the former Soviet Union. We hopethat this study will encourage further discussionamong policymakers and think tanks in the tran-sition countries and will assist their dialogue withexternal donors and advisers on how to bettersupport the transition process.

Johannes F. LinnVice PresidentEurope and Central Asia Region

xi

1Acknowledgments

This 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, IraLieberman, 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 ofCalifornia at Berkeley, Simon Johnson of the Massachusetts Institute of Technology, and Christian Mumssenand Thomas Richardson, both at the International Monetary Fund. Bruce Ross-Larson was the principaleditor. 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 andEconomic 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, CarlosCavalcanti, 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 JohannesLinn. Peer reviewers were Nicholas Stern, Alan Gelb, and Marek Dabrowski. The team thanks all ofthem 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 thatoriginally 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, VladimirA. Mau, Ivan Miklos, Viktor Pinzenik, Ion Sturza, and Marat Sultanov. Earlier versions of the reportwere 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 Educationof Charles University and the Economic Institute of Sciences of the Czech Republic in Prague, and at aconference on “Trade, Integration, and Transition” held in Bela Balassa’s memory in Budapest in Octo-ber 2001, where it benefited from the comments of János Kornai and Andras Nagy.

Book design, production, and dissemination were coordinated by the World Bank Publications team.

xiii

1Overview

Adozen years have elapsed since the world witnessed the euphoria greeting the fall of theBerlin wall. Ten years ago last summer, Boris Yeltsin claimed his place in history by climbingatop 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, everythingseemed possible (Balcerowicz 1995).

At the beginning of the new millennium, a profound divide lies between Central and SoutheasternEurope and the Baltics (CSB) and the Commonwealth of Independent States (CIS).1 In the CSB, offi-cially measured gross domestic product (GDP) bounced back from a transition recession, recovered toits 1990 level by 1998, and exceeded that level by 6 percent in 2000. However, in the CIS GDP in 2000stood at only 63 percent of its 1990 level. While GDP in Poland, the most populous country in theCSB, increased by more than 40 percent between 1990 and 1999, it shrank by 40 percent during thesame 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 povertyline.4 A decade before fewer than one in 25 lived in such absolute poverty. While absolute incomedeprivation at those levels is virtually nonexistent in many Central European countries, its incidence isas 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 havecome to rival the most unequal countries in the world (box 1).

A similar divide runs across the political landscape as well. Competitive democracies—underpinnedby 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 toparticipate in elections and constraints on civil liberties during at least some period of the transition haveconcentrated 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 forthe market economy as a result of corruption, weak public sector management, and in some cases war

Overview

xiv

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 transition economies and dramatically in some of them (see figure A). Countries such as Armenia, theKyrgyz 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 reformsand liberalization. But this is only part of the story. While in-equality has increased almost everywhere, the more advancedreformers show much more equal, rather than more unequal,outcomes, compared with less advanced reformers. This differ-ence cannot be solely explained by different conditions acrossthe countries at the start of transition.

Rather, a recent World Bank study (2000b) shows thatpositive developments largely explain the rise in inequality inthe CSB: rising returns to education, decompressing wages, andemerging returns to risktaking and entrepreneurship. Theseforces are welcome despite the increase in inequality, becausethey signal that the market is now rewarding skills and effort,as in more mature market economies. In the CSB, moreover,strong social transfers and redistribution mechanisms havedampened the rise in education premiums and wage disper-sion, in line with the demands these societies have placed ontheir governments for such measures.

The experience of the CIS is very different. Rising educa-tion premiums and wage dispersion explain very little of therise in inequality. In Armenia, Georgia, the Kyrgyz Republic,Moldova, and Russia income differences linked to educationalachievement explain less than 5 percent of inequality, com-pared with 20 percent in Slovenia and 15 percent in Hungaryand Poland. The causes of the huge rise in inequality lie:

• In the prevalence of widespread corruption and rent seek-ing. There is a strong correlation between higher corruptionand higher inequality (and higher poverty) in the region.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. Theseinterests have been able to limit competition and concentrate their economic power through such mechanisms as special licenses andmonopolies. They have undermined state institutions and blocked reforms that would serve the public good.

• In the resulting collapse of formal wages and income opportunities. Wages at old jobs have collapsed or are not paid, while new formal jobopportunities are stifled by the lack of competitive markets and by the pervasiveness of corruption. People, except for a privileged few, arelargely stuck in their low-paying (and sometimes nonpaying) jobs. To make ends meet they supplement their incomes with diverse forms ofself-employment, much of it subsistence agriculture in small household plots. Throughout the CIS, earnings from such small plots accountfor 40–70 percent of total household earnings. Access to connections and informal networks and an ability to pay are key to finding a joband getting ahead. This has led to highly unequal outcomes.

1987–90 1996–99

FIGURE A.

Changes in Income Inequality in SelectedTransition Economies

Source: World Bank (2000b).

Hungary

Poland

Czech Republic

Estonia

Slovenia

Latvia

LithuaniaCroatia

Kyrgyz RepublicGeorgia

Moldova

Armenia

Russian Federation

Ukraine

Tajikistan

Belarus

0.80.40.20

Gini coefficient

0.6

Overview

xv

and civil strife. Outside Central Europe and theBaltics the optimism pervading the beginning oftransition has been tempered by the harsh eco-nomic realities of its first decade.

However, the starkness of this binary pictureneeds to be softened; growth outcomes have var-ied significantly even within the two broadgroups of countries. Furthermore, all countrieswent through the transitional recession, whichcaused real GDP to dip from its 1990 levels bynearly 15 percent in the CSB and by more than40 percent in the CIS.

Of the CSB countries, Hungary, Latvia, Po-land, Slovenia, and to some extent Estonia andLithuania have enjoyed several years of uninter-rupted growth. By contrast, growth in Bulgariaand Romania was sharply interrupted by seri-ous macroeconomic crises brought on by insuf-ficient structural reform in the mid-1990s, andGDP in 2000 stood at four-fifths its 1990 level.The Czech Republic had a similar but less severeexperience; GDP declined during 1997-99 be-cause of a macroeconomic crisis with structuralorigins. In fact, the Czech Republic was the onlycountry in Central Europe that had not reachedits 1990 GDP level by 2000.

In the CIS such early reformers as Armenia,Georgia, and the Kyrgyz Republic, whose GDPfell steeply, and such nonreformers as Belarusand Uzbekistan, where the decline in GDP wassmaller, have been growing in the past five years.But Russia, barring a short-lived upturn in 1997,did not begin to grow until 1999, while Ukrainedid not return to growth until 2000.

The wide variation in transition across theregion raises questions:

• Why has the growth of some transitioneconomies been better than that of others?To what extent can these differences be as-cribed to economic policy choices rather thancircumstances at the start of transition orexternal economic shocks?

• Do the policy lessons from the countries thatenjoyed several years of rapid growth con-tinue to be relevant for the CIS and South-eastern Europe, which have made lessprogress with the transition? Do transition

economies have some common characteris-tics that make those lessons applicable today?

• If the advantages of economic reform are soobvious, why do countries mired in a noman’s land between centrally planned andmarket economies not adopt them? Howmight political support for reform be built inthose countries?

• In what key respects should policy advice totransition economies be modified to reflectexperience from the first decade and the newconditions prevailing today?

This report seeks answers to these questions.

The Quest for Growth: Promoting Disciplineand Encouragement

The focus on economic growth needs to beput into a broader perspective. For much

of the CIS and Southeastern Europe the restora-tion of sustained growth is a key priority. With-out it these countries will not generate income-earning opportunities for households. Nor willthey have the resources needed to provide basicpublic goods such as legal and judicial systems,secure property rights, and basic infrastructure;maintain essential investments in education andhealth; or set up social safety nets targeted tothe most vulnerable. In this respect transitioneconomies are no different from other economies.

Continued growth is also important for theleading reformers in Central Europe and theBaltics. While all the Central European coun-tries except the Czech Republic had surpassedtheir 1990 GDP by 2000, per capita incomes inthe three wealthiest countries aspiring to Euro-pean Union accession were still only 68 percentof the European Union average for Slovenia, 59percent for the Czech Republic, and 49 percentfor Hungary. However, an exclusive focus ongrowth, while providing basic public goods andprotecting the most vulnerable, is not enoughfor them. They need to consolidate the gains ofthe first decade of transition and address “sec-ond generation” reform issues. They have tosecure control over quasi-fiscal and contingentliabilities. They have to undertake reforms in

Overview

xvi

labor and financial markets to allow the ben-efits of growth to be more widely shared. Theyalso have to restructure social expenditures tomake them fiscally more affordable without im-pairing the effectiveness of the social safety net.Several of these reforms overlap with those re-quired to join the European Union.5

This report is primarily about economicgrowth. But the focus is not meant to be exclu-sive. Two companion reports on poverty andinequality and on anticorruption deal with is-sues particularly important in the transition(World Bank 2000b, c).

The common heritage of socialism impliedthat all countries in the region began their tran-sition with a production system adapted not toa competitive environment but to the exigen-cies of a command economy. External liberal-ization at the beginning of transition generatedsignificant productivity differences across sec-tors and enterprises in a production systembased on cheap energy and subsidized transport.For example, energy intensity, measured as theamount of energy used per unit of GDP, was0.95 tons of oil equivalent per US$1,000 of GDPin the Soviet Union in 1985, compared with0.50 tons of oil per US$1,000 of GDP in OECDcountries (IMF and others 1991). In April 1992,after Russia had adjusted the price of oil sev-eral times, its domestic price was still only 3percent of the world price (Tarr 1994). Manysectors and enterprises were not viable afterprice liberalization.

Two challenges had to be confronted:

• First, the imposition of market discipline oninherited enterprises so that they would facethe incentive to restructure and, in so doing,become more productive and able to com-pete at the new prices. Failure to do so shouldlead to closure.

• Second, encouragement of the creation ofnew enterprises willing and able to competein the marketplace without seeking specialfavors from the state.

Economic growth reflects the interplay be-tween old enterprises in need of state support,which reduce growth by absorbing more

resources than they produce, and restructuredand new enterprises, which increase growth. Thefall in growth is initially dominated by the dragof old enterprises, which leads to a period of de-cline. With time, if the business environment fa-vors production and innovation rather than rentseeking, restructured and new enterprises gainthe critical mass to overcome the negative effectsof old enterprises, leading to recovery andeconomywide growth.

The initial conditions of geography, history,and price and output distortions at the start oftransition and the external economic shocks aris-ing from the breakup of the Soviet Union, war,and civil strife were of course important. How-ever, analysis done for this report shows that ini-tial conditions were significant factors during theinitial period of output decline (1990-94), ratherthan throughout the full ten years of transition,even after accounting for differences across coun-tries in policy reform and the impact of externaleconomic shocks. That analysis also demon-strates that policy reforms have been significantfactors in differences among countries in thespeed of economic recovery, once due account istaken of differences among countries in initialconditions and the impact of external economicshocks. Furthermore, market-oriented policy re-form not only speeded up economic recovery andpromoted growth in the medium term, but it alsomitigated the effects of the transitional recessionin the short term. How effective policies havebeen in disciplining the old sector and encour-aging the new therefore holds the key to under-standing why growth has been better in sometransition economies than in others.

Discipline forces old enterprises to releaseassets and labor, which are then potentiallyavailable to restructured and new enterprises.It does this by hardening budget constraints,introducing competition in product markets,providing exit mechanisms, and monitoringmanagerial behavior to generate incentives forproduction and innovation (rather than for as-set stripping and theft).6 Discipline also pushesold enterprises to divest themselves of such so-cial assets as housing, health clinics, and kin-dergartens to local governments, shifting the

Overview

xvii

locus of social protection away from enterprisesto governments. The social safety net then needsto be strengthened to ensure that labor shed bycontracting enterprises and other losers fromreform do not fall into poverty, while not erod-ing these workers’ incentives to find employ-ment in new enterprises.

Encouragement entails policies to create anattractive and competitive investment climate inwhich restructured and new enterprises have in-centives to absorb labor and assets rendered in-expensive by the downsizing and to invest inexpansion. These policies include reducing ex-cessively high marginal tax rates, simplifyingregulatory procedures, establishing secure prop-erty rights, and providing basic infrastructure,while maintaining a level playing field amongold, restructured, and new enterprises. At thesame time the policy environment must provideincentives for wealth creation rather than rentseeking and asset stripping by new enterprises.This mode of adjustment broadly correspondsto the experience to date of economies in Cen-tral Europe and the Baltics.

The Flipside: Protection and Discouragement

The disposition of assets among old, restruc-tured, and new enterprises also provides a

useful perspective on the experience of many ofthe CIS countries and, to some extent, South-eastern Europe. These countries have tended toprotect rather than discipline old enterprisesthrough subsidies granted through the budget,energy consumption, and the banking sectors.Where institutions of public and corporate gov-ernance are not strong enough, asset stripping,theft, and other violations of property and share-holder rights become widespread. Entry of newenterprises is discouraged—or, at best, only se-lectively encouraged—because of oppositionfrom entrenched interests that would lose fromfurther liberalization of entry. Furthermore, be-cause of a poor investment climate where taxrates are high, licensing and registration proce-dures are open to abuse, and the legal and judi-cial system is weak, corruption becomes a seri-ous obstacle to the growth of new enterprises.

Support for the old sector is ultimately fi-nanced through taxes on households, new enter-prises, and enterprises that have restructured suc-cessfully to survive the market test. The socialsafety net is unable to prevent people from mov-ing into subsistence and low-productivity activi-ties to ensure their survival. Such a protect-and-discourage strategy creates an environment whereresource transfers tend to flow in a direction op-posite to that in a discipline-and-encourage envi-ronment. Transfers from efficient to inefficiententerprises and sectors undermine the credibilityof government policy, with detrimental conse-quences for the economy.

The logic of discipline and encouragementis intended to apply broadly to the productionof goods. But it may also be applied, with somemodification, to the banking system, an impor-tant part of the investment climate required toattract new enterprises. Hard budget con-straints on state banks and a credible threat ofexit for failed banks are essential to disciplinebanks and enterprises alike. However, encour-agement does not always imply free entry ofnew banks. Free entry could help make thebanking sector competitive, but potential en-trants must satisfy prudential norms, such asthose for minimum capital requirements andcapital adequacy. It is also important that ex-pansion of the banking system not outpace thecapacity for effective supervision and growthin the number of creditworthy borrowers.

Shading the Classification

T he juxtaposition of discipline-and-encouragement and protection-and-

discouragement highlights two contrastingmodes of adjustment. In reality, country out-comes span a range of intermediate possibilities,depending on whether liberalization was imple-mented, hard budget constraints imposed, andan enabling business environment promoted andin what order and how vigorously.

• The discipline of hard budget constraints andinstitutions of corporate governance to moni-tor managerial behavior and encouragement

Overview

xviii

through liberalization and a climate hospi-table to domestic and foreign investment areperhaps seen most clearly in Estonia, Hun-gary, and Poland.

• Even in the broad category of discipline andencouragement, however, softer budget con-straints and hence less discipline prevailed fora long time in the Czech Republic, Lithuania,and the Slovak Republic.

• Bulgaria, the Kyrgyz Republic, Moldova,Romania, Russia, and Ukraine liberalizedtheir economies, but for a long time failed tomaintain discipline through hard budget con-straints. They were also unable to containtunneling, the expropriation of assets andincome belonging to minority shareholders,and theft through either rule of law or ad-ministrative control. Though many of thesecountries did encourage new entry early inthe transition, the capture of the state by anarrow set of vested enterprises—old enter-prises and well-connected early entrants—dis-couraged further entry and created a poorinvestment climate, resulting in a pattern ofprotection and selective encouragement.

• Belarus, Turkmenistan, and Uzbekistan,which have undertaken some liberalization,but have not imposed hard budget constraints,strongly discourage new entry. Policies suchas access to foreign exchange and credit onspecial terms soften budget constraints forstate enterprises. But continuing reliance oncentralized political power and mechanismsof administrative control inherited from thecommand economy did limit extensive assetstripping and other forms of theft at the en-terprise level. That led to a situation incorpo-rating some elements of discipline in an oth-erwise strongly protective stance, togetherwith discouragement of entry.

New Enterprises Spur Economic Growth

The growth-enhancing effects of new enter-prises and the growth-restraining effects of

the old broadly suggest that new enterprises intransition economies are more productive thanold enterprises. This is supported by data from10 transition economies covering both the

leading and lagging reformers in the region, drawnfrom the World Bank’s database on small andmedium-size enterprises (SMEs). It is also sup-ported by a comparison between old and new en-terprises in the Business Environment and Enter-prise Performance Survey, conducted jointly bythe European Bank for Reconstruction and De-velopment and the World Bank in 1999 (see box3.1). The survey finds that new enterprises out-perform old enterprises in sales, exports, invest-ment, and employment (chapter 3). Thus a trans-fer of resources from old enterprises to new canbe a source of growth. Whether that potential isrealized depends on the discipline imposed on oldenterprises to shed resources and the encourage-ment extended to new enterprises to absorb them.

The interaction between old and new enter-prises is key to economic growth. The share oftotal employment and value added accounted forby small enterprises (defined as employing fewerthan 50 workers) as a proxy for new enterprisesdivides transition economies into two groups.7

In the Czech Republic, Hungary, Lithuania, andPoland, new enterprises grew very rapidly. Theynow account for 50 percent or more of employ-ment, the average for the European Union, andfor between 55 and 65 percent of value added.But in Kazakhstan, Russia, and Ukraine, whichhave seen modest or no growth in new enter-prises, the share of employment has stayed at orbelow 20 percent and the share of value addedhas stayed between 20 and 30 percent.

Hungary and Poland saw a sharp and earlydecline in employment and a rapid demise of theold sector, which initially made resources avail-able cheaply to the new sector. Such discipline isimportant but insufficient; encouragement is alsoneeded. Growth takes off only when the newsector evolves from a passive receptacle for ab-sorbing resources into an active competitor, rap-idly increasing its share of employment and at-tracting the most qualified workers. The evidencesuggests that new enterprises must reach a thres-hold of around 40 percent in their contributionto employment before they can become an en-gine of growth. In Russia and Ukraine, wherethe contribution of the new sector to employ-ment is well below the threshold, a large pro-portion of the labor force remains mired in old,

Overview

xix

unrestructured enterprises not generating in-creases in productivity. The new sector has notemerged as a source of growth.

New enterprises are more productive than oldones, but productivity differences diminish withtransition. The difference is greater in Kazakhstan,Russia, and Ukraine than in the Czech Republic,Hungary, Latvia, Lithuania, and Poland. Why?It is because closure and restructuring can raisethe productivity of factors in the old sector andbecause fast growth of enterprises and employ-ment can reduce the productivity of factors in thenew sectors. Thus a comparison of labor produc-tivity shows a difference in favor of new enter-prises of more than 100 percent in Ukraine, wherethe contribution of the new sector to total em-ployment is 17 percent. That difference narrowsto just more than 40 percent in Hungary, wherethe contribution of the new sector to total em-ployment is 55 percent.8

Creating a policy environment that disciplineslow-productivity old enterprises into releasingresources and encourages high-productivity newenterprises to absorb those resources and to un-dertake new investment, without tilting the play-ing field in favor of any particular type of enter-prise and while strengthening the social safetynet to protect the most vulnerable, is central toeconomic growth in transition economies. Thisis the main lesson from the successful reformersin Central Europe and the Baltics.

When Is Transition Over?

Do transition economies at different lev-els of gross national income per capita(ranging in purchasing power parity

terms from US$2,100 in Moldova in 1999 toUS$16,050 in Slovenia in 2000 [World Bank 2001])have anything in common that would make les-sons from economies in the region leading in re-form applicable to those lagging in reform? Thewide dispersion in the productivity of labor andcapital across types of enterprises at the onset oftransition and the erosion of those differences be-tween old and new sectors during reform providea natural definition of the end of transition.

Enterprises in a typical transition economycan be distinguished by history: are they new,

restructured, or old? They can also be distin-guished by economic performance: are they pro-ductive? Furthermore, history and performanceare related. New enterprises are expected to bemore productive than restructured enterprises,which are expected to be more productive thanold enterprises. As markets develop and resourcesare allowed to flow to their most valued uses,the role of history progressively weakens, anddifferences in productivity arising from member-ship in any of the categories tends to disappear,consistent with the evidence on the behavior ofdifferences in productivity.

This is not to suggest that differences in pro-ductivity across enterprises will disappear alto-gether. These differences always exist as a resultof technical innovation and new export marketpenetration, among other factors. However, thevariation in productivity could not be system-atically attributed to the enterprises’ historicallydetermined categories: old, restructured, andnew. When that distinguishing characteristic islost in a country, the transition can be taken tobe over. At that point, the economic issues andproblems policymakers must deal with are nolonger specific to transition. At what level of percapita income will this occur? The answer de-pends on the success of disciplining the old sec-tor and encouraging the new one. It also dependson the success of the business environment inattracting investment.

Do Central Europe and the Baltics Point theWay Forward?

The striking diversity in challenges and cir-cumstances among countries that have not

proceeded far in the transition—in particular,countries in the CIS and Southeastern Europe—raises the question of whether 10 years after thedissolution of the Soviet Union these countriescan learn from the successful reforms in CentralEurope and the Baltics.

Countries in the CIS and Southeastern Eu-rope continue to face significant productivitydifferences across old, restructured, and newenterprises characteristic of the transition andare therefore a long way from the end of transi-tion. Hence, the framework of discipline and

Overview

xx

encouragement and its associated policy and in-stitutional implications remain relevant for un-derstanding what needs to be done to restoregrowth and protect the most vulnerable in theseeconomies (the policy and institutional reformsassociated with discipline and encouragementare summarized in annex 1).

However, the political context for pursuingreform policies has changed greatly in a decade.At the beginning of transition, implementing re-forms focused on overcoming the resistance ofthe nomenklatura, whose political and economicprivileges fueled support for the prevailing eco-nomic system, and on building support for re-form among the newly mobilized public. But inthe past decade power over economic resourceshas shifted, often in a highly concentrated pat-tern, from state bureaucrats to the private sector,even in much of the CIS and in Southeastern Eu-rope. That has made it easier for narrow specialinterests to capture the state and block furtherreforms that may undermine short-term rents.

Other problems, specific to individual coun-tries or subgroups of countries, have little to dowith the transition but demand urgent resolu-tion. Securing peace and inaugurating the pains-taking task of nation building in the SouthCaucasus and the Balkans, wracked by war andcivil strife, are priorities. So is controlling thespread of tuberculosis and HIV/AIDS, whichthreaten millions of lives. But these challengesare outside the scope of this report.

Can We Have It Both Ways: Encouragementwithout Discipline?

The implementation of policies associatedwith both discipline and encouragement

presents a challenge. Is it possible to downsize theold sector slowly while encouraging the new sec-tor and thus to avoid the pain of liquidation andrestructuring until a cushion has been put in place?Encouragement without discipline will not workif old enterprises absorb resources that would oth-erwise flow to new enterprises. For example:

• Protection of state-owned enterprises and farmcollectives through the banking sector in Bul-garia and Romania led to a sharp increase in

nonperforming loans as a share of total bank-ing sector loans in the 1990s (in Romania, 34percent in 1998). These loans prevented theexpansion of bank credit to new, small, andpolitically less-connected enterprises. They alsotriggered banking and macroeconomic crisesthat called for stabilization and a tighteningof credit, hurting new enterprises.

• Protection of the old industrial sector inBelarus and Uzbekistan through specially fa-vorable foreign exchange regimes, directedcredit, and high trade protection has meantthat whatever credit and foreign exchange re-main are available only to new smaller enter-prises at prices several times higher than whatwould have been paid in unified markets. InUzbekistan small enterprises have to pay threetimes more for foreign exchange to financetheir imports than do large state enterprises.

• Protection through tax and utility arrears incountries such as Georgia, the Kyrgyz Repub-lic, Moldova, Russia, Romania, and Ukrainemeant that new and more energy-efficiententerprises were charged more to compen-sate for revenue losses from nonpayment byold, less energy-efficient enterprises to utili-ties in the energy sector. Tax exemptions forlarge enterprises and agricultural collectivesin Ukraine, negotiated offsets to pay taxes inRussia, and tax avoidance in exchange forbribes by large enterprises in Georgia typi-cally worked to the disadvantage of new andsmaller enterprises, which ended up payinghigher prices and bribes as a proportion oftheir annual revenue.

The lack of a vibrant emerging private sec-tor, because of a policy of discouragement, lim-its the outside options available to those in oldenterprises. These limited private sector job op-tions increase the social cost of restructuring theold enterprises, resulting in the need for addi-tional protection for the old sector. The comple-mentary relationship between discipline and en-couragement also sheds light on why a relaxationin discipline—brought about, for example, byspecial treatment for powerful lobbies—is asso-ciated with selective rather than complete encour-agement. It also helps explain why policy reform

Overview

xxi

must cover both discipline and encouragement,thus proceeding along an ambitiously broadfront, and therefore why there are no magic bul-lets in transition.

Learning from China?

The success of encouraging entry of new en-terprises in China (where GDP per capita

grew 8 percent per year from 1978 to 1995, lift-ing 200 million people out of absolute poverty)without imposing significant discipline on stateenterprises raises a question of the applicabilityof China’s reform to the transition economies ofEastern Europe and the former Soviet Union.Through different channels China modulated thetradeoff between encouraging new enterprisesand not imposing hard budget constraints onexisting state enterprises. The country reapedspectacular gains from liberalizing repressed sec-tors such as agriculture, which had surplus la-bor, and rural industries and from a massive in-flow of foreign direct investment.

Part of these gains, helped by a high savingsrate, could be transferred through the bankingsystem to finance loss-making state enterprises,which were far less important in China than inmost countries in Eastern Europe and the formerSoviet Union. For example, only 19 percent ofthe Chinese labor force worked in the state sec-tor and were thus entitled to a range of socialbenefits, compared with 90 percent in Russia.Furthermore, tight political control over assetstripping, arbitraging between controlled andmarket prices for private gain and corruption,together with some state capacity to manage pub-lic assets allowed China to move its loss-makingstate enterprises more slowly to market condi-tions at the same time that explosive growth ofnew enterprises took place. If a substantial in-flow of resources, for example, from liberaliz-ing a previously repressed sector or from foreigndirect investment, and state capacity to managethe process allow a country to follow such aphased transition, it is less likely to experience aperiod of contraction in output.

These conditions were largely absent in mosttransition economies covered by this report. Butthe costs of soft budget constraints on China’s

state enterprises remain to be fully recognized.The share of nonperforming loans in the bank-ing system, which served as a conduit for assis-tance to state enterprises, is between 30 and 40percent of annual GDP. Addressing this prob-lem is likely to pose a major fiscal challenge. Insum, the transition economies of Eastern Europeand the former Soviet Union did not have theresources for a phased transition for state enter-prises, but they would be well advised to drawfrom China’s experience the importance of en-couraging new enterprises as a basis for wealthcreation and economic growth.

Institutions Are Important, but So Are Policies

Early on, Fischer and Gelb (1991) flaggedthe role of institutional reforms in the transi-

tion. Given that it is now argued that a key de-ficiency of the transition has been insufficientattention to building a market-friendly institu-tional framework, reflecting on the experienceof East Germany’s unification with West Ger-many in 1990 is instructive. It illustrates, amongother things, how inappropriate policy choicescan undermine performance even in the mostfavorable of institutional environments. EastGermany was able to adopt all the institutionsof West Germany without delay. It also receivedmassive financial transfers, which averaged be-tween 40 and 60 percent of East Germany’sGDP over the period 1991 to 1997, and whichcontinue at levels exceeding 4 percent of WestGermany’s GDP per year. Furthermore, Germanreunification conferred on East Germany auto-matic membership in the European Union.

Despite these considerable advantages, EastGermany suffered an initial decline in GDP thatwas deeper than its transition economy neigh-bors’ declines, and it has experienced one of theslowest GDP growth rates in Europe, disappoint-ing expectations that it would catch up rapidlywith West Germany. This occurred, first, becausethe one-to-one conversion rate between WestGerman and East German marks led to a sub-stantial overvaluation of the East German cur-rency and, second, because attempts to bring EastGerman wages in line with West German wages,in the face of much lower labor productivity in

Overview

xxii

the East, undermined East Germany’s competi-tiveness. As a result, a much larger part of theinherited capital stock was rendered unproduc-tive than would have been the case had moreappropriate macroeconomic policies been cho-sen. Thus, while institutional change is impor-tant, so too is policy reform, and it is essentialthat they proceed hand in hand.

The Political Economy of Discipline andEncouragement

Many transition economies outside CentralEurope and the Baltics are stuck in a no

man’s land between plan and market. If the ad-vantages of economic reforms are so obvious,why doesn’t every country adopt them? Can eco-nomic policy choices be systematically relatedto particular institutional characteristics of po-litical systems in transition?

The political economy of reform within theframework of discipline and encouragement canbe expressed graphically by tracing the pathsof winners and losers from the transition. Fig-ure 1 depicts the gains and losses in income ac-

cruing to three different constituencies at dif-ferent doses of reform in a typical transitioneconomy.

• State sector workers, employed in state en-terprises and lacking the skills to becomenew entrants in the competitive market, facea sharp drop in income as discipline callsfor downsizing the sector, with little hopeof any substantial recovery with the inten-sification of reform.

• Potential new entrants, workers in state en-terprises and new entrepreneurs with skills tobecome new entrants in the competitive mar-ket, have a classic J-curve pattern of income.They face significant adjustment costs at lowlevels of reform as they exit the state sector. Inaddition, they realize gains only when enoughprogress has been made with policy and insti-tutional reforms to promote and support newentry into the competitive market.

• Oligarchs and insiders begin the transitionwith substantial de facto control rights overstate assets and close ties with the politicalelite inherited from the previous command

R0

Oligarchs and insiders

R = Extent of reforms

State sector workers

New entrants

R1 R2

Income gains

0

+

–

FIGURE 1.

Winners and Losers from Reform

Note: R0 = no reforms; R1 = point at which income gains of oligarchs and insiders are maximized; R2 = level of reformsthat allows the winners of reforms beyond R1 (new entrants) to compensate for or exercise enough political pressure toneutralize the resistance of oligarchs, insiders, and state sector workers.

Source: Authors.

Overview

xxiii

system. However, because of limited skillsto compete in the market economy, they facean inverted U-curve of income gains. Theyare the immediate beneficiaries of liberal-ization and privatization, as de facto con-trol rights over state assets can be convertedinto de jure control and cash flow rights.They reap concentrated gains in the earlystages of reform from the opportunities forarbitrage, rent seeking, and tunneling thatarise if liberalization and privatization arenot combined with discipline and encour-agement. But these gains dissipate as furtherreforms lead to increasing competition andmarket entry.

Given these patterns of gains and losses, eachconstituency prefers a different combination ofreforms. State sector workers prefer the statusquo R0 and reject all reforms. Oligarchs and in-siders prefer a partial reform and sustain the re-form process through R1, the point where theirgains are maximized and beyond which furtherimplementation of policies of discipline and en-couragement threaten to undermine gains fromrent seeking and tunneling. For potential newentrants, the reform process offers sacrifices atthe beginning for the promise of gains when thereforms are further advanced.

Where the risk of oligarchs and insidersblocking anything more than partial reform ishigh, potential new entrants and state workerswill either reject reform or support only partialreform, because the latter, by limiting thedownsizing of the state sector and maintainingthe flow of subsidies, imposes lower adjustmentcosts. Yet it is precisely such partial reforms—liberalization without discipline and with selec-tive encouragement—that make capture of thestate by oligarchs and insiders a self-fulfillingprophecy. This has led to a so-called partial re-form paradox in many transition economies inwhich governments lack credibility and are highlysusceptible to state capture. This leads potentialnew entrants at the outset of transition to dis-count substantially the potential gains from anyproposed radical reforms and instead supportpartial reforms that offer lower costs early in thereform process, even though they are more likely

to lead to barriers to entry. Public support forradical reforms therefore depends on perceptionsof government credibility in its commitment tofollow through with such reforms.

The risk of “getting stuck” at a low level ofreform (R1) characterized by liberalization with-out discipline and limited encouragement of newentry is high. As both insiders and state sectorworkers face declining incomes after R1, thesegroups have a strong incentive to join forces tooppose further economic reforms. It is onlywhen reforms reach a critical threshold (R2) thatthe added gains to new entrants are enough toallow these winners to either compensate thelosses of the other groups or to generate enoughpolitical pressure to neutralize opposition tocontinued reform.

By recognizing that different combinationsof reforms produce different configurations ofwinners and losers, the framework of disciplineand encouragement suggests two political chal-lenges in promoting economic reform:

• Securing the support of potential new en-trants for comprehensive reforms until widerefficiency gains from discipline and encour-agement are realized.

• Preventing the early winners from liberaliza-tion and privatization from undermining fur-ther reforms that would impose discipline andencourage new entry and competition andthus reduce their rents.

To meet these challenges, governments mustappear credible to potential new entrants in itscommitments to follow through with the longand difficult process of economic reform. Gov-ernments must also be able to constrain oli-garchs and insiders from using their initial ad-vantages in the reform process to derail furtherreforms that would create a more competitivemarket economy.

Credibility and constraint are rooted in po-litical institutions shaped by the cultural and his-torical legacies that guided the exit from com-munism. In many countries in the CIS andSoutheastern Europe, where the state has beencaptured by narrow private interests, the collapseof communism was rooted in a contest amongcompeting elites rather than in any broad social

Overview

xxiv

movement. The new political arrangements inthese “concentrated political regimes” were de-signed by incumbent leaders, often as a way toconsolidate their power. They lacked the cred-ibility to build and sustain broad popular sup-port for a comprehensive reform program.

As a result, these countries embarked on tran-sition without a broad social consensus on thegoals of reform and without a way of organiz-ing the public behind these goals. Instead, in-cumbent politicians sought alliances with pow-erful incumbent enterprises. In addition, thepoliticians continued partial liberalization andprivatization in the context of soft budgets andbarriers to entry that created tremendous oppor-tunities for rent seeking by old and new enter-prises, especially in economies rich in naturalresources. Countervailing pressures from com-peting groups were weak and the disaffection andapathy of the “losers” minimized the direct coststo politicians of poor policy choices. As a result,countries with concentrated political regimeshave tended to languish in an equilibrium trapof partial economic reforms. Political and eco-nomic power has been used to preserve marketdistortions that benefit narrow vested interestsat considerable social cost.

This partial reform equilibrium can be con-trasted with the situation in the “competitive de-mocracies” prevailing in Central Europe and theBaltics. In the aftermath of popular revolutionsagainst communist rule, political institutions inmost of these countries emerged from roundtablenegotiations among broadly representative popu-lar fronts and a wide range of other organizedinterests. This, together with the close ties of thesecountries to Western and Northern Europe andthe pull of potential European Union accession,contributed to a wider social consensus on themain directions of reform and broad public sup-port for comprehensive reform programs in theearly stages of transition.

New governments in competitive democra-cies tended to focus first on promoting new con-stituencies of “winners” by removing entry bar-riers, quickly tackling severe macroeconomicinstability (with its high costs to the public), andusing social protection to support the “losers”from the dislocations of reform. A legacy of

strong public administration allowed for greatersecurity of property and contract rights and bet-ter public infrastructure, important preconditionsfor promoting new entry. As reform progressedto promote entry and improve the enabling en-vironment, constituencies with a stake in advanc-ing reform grew stronger, and the emergence ofpowerful insiders and oligarchs diminished. Thiscombination allowed these countries to imple-ment and sustain comprehensive reforms.

Political developments and economic reformsare closely interrelated. Political systems affectthe incentives of politicians to make certain eco-nomic policy choices; reform choices shape theconfiguration of social groups and the distribu-tion of power, which affects the structure andfunctioning of the political system. For example,economic reforms that facilitate new entry alsostrengthen the constituency of SMEs, which buildsupport for increasing political competition.

Nevertheless, given the sharp break with com-munism and the disintegration of the SovietUnion, choices about the structure of politicalsystems in the transition economies were gener-ally made before decisions about the nature andpace of economic reform. Moreover, in all but acouple of countries—Croatia and the SlovakRepublic—the nature of the political regime hasnot changed much since the start of transition.This suggests that while the pace and directionof economic reforms may have reinforced initialchoices about the structure of the political sys-tem, they do not appear to have decisively shiftedthe course of political transition. As a result, astronger case can be made for identifying thedirection of causation from political choices toeconomic choices, thus providing part of theexplanation for why some countries have beenunable to move beyond partial reform.

Shifting Policy Priorities to Account forExperience and New Conditions

Much economic policymaking is endog-enous from the broader perspective of

political economy. Designing effective reformstrategies must therefore take into account thepolitical incentives and constraints that blockprogress in transition. But although initial

Overview

xxv

conditions and political institutions influence re-form paths, these factors cannot wholly predeter-mine outcomes in a process as complex and mul-tifaceted as transition. Experience from across theworld demonstrates that talented political lead-ers can maneuver countries out of so-called re-form traps. Critical elections or external shockscan break long-term stalemates on reform. Newleaders can mobilize alternative coalitions andspark collective action that tips the balance ofpower between the potential winners and losersfrom further economic reforms. Clever winnerscan devise win-win strategies that co-opt theiropponents to build support for reform.

In what ways should policy advice duringthese extraordinary opportunities for reform re-flect the experience of the past decade and today’sconditions? Three broad areas can be identified.

From Privatization and Restructuring toPromoting Entry

Policy needs to shift its emphasis fromprivatization and restructuring of assets to creat-ing wealth through new enterprises. The earlyemphasis on rapid privatization entailed remov-ing ownership of enterprises from the state, cre-ating a constituency for private ownership to helpguarantee the irreversibility of reform, and stop-ping abuses such as asset stripping by enterprisemanagers and other forms of “spontaneousprivatization.” Although much remains to bedone, particularly in privatizing medium-sized andlarge enterprises, these past concerns weigh lessheavily on policymakers today. To this must beadded the suggestion from the empirical litera-ture that privatization has promoted restructur-ing in the CSB, but not in the CIS (see Djankovand Murrell 2000 and chapter 7 in this volume).

New enterprises are important to promotinggrowth. In both the leading and lagging reform-ers, new enterprises enjoy a productivity advan-tage over old enterprises. So transferring re-sources from old enterprises to new ones is asource of growth. Although causality cannot beinferred from the evidence, countries that havereturned to sustained growth have relied on avibrant new sector to absorb labor and otherresources released by the downsizing of the old

sector and to provide a major share of employ-ment (50 percent) and value added (55–65 per-cent) in the economy. By contrast, in countrieswhere restoring sustained growth has provedmore elusive, new enterprises account for a lowshare of employment (10–20 percent) and valueadded (10–20 percent).

That is why encouraging an investment cli-mate attractive for new entrants and meetingthe policy and institutional challenges of en-couragement should be the highest prioritiesfor policymakers in transition economies. Re-member, though, that encouragement cannotgo very far without discipline. Therefore, theemphasis on encouragement is more effectivethe more it is accompanied by hard budget con-straints, exit mechanisms, product market com-petition, and stronger institutions for monitor-ing managerial behavior.

From Depoliticizing Enterprises to MonitoringManagers

The lack of restructuring in privatized enterprisesin the CIS, together with tunneling and theft,renew interest in the question of what institu-tions are needed to encourage managers to be-come effective stewards of enterprise assets. Al-though developing these institutions of corporategovernance has been on the reform agenda sincethe start of transition, the difficulty of doing soin countries without recent market experiencewas probably underestimated.

International experience suggests that with-out effective legal protection, suppliers of financedo not enter into contracts with enterprises toensure that they get a return on their invest-ment—the essence of the corporate governanceproblem—even if such arrangements are in theinterest of both parties. Concentrated ownership,by providing enhanced monitoring of managersby shareholders, can overcome some of the cor-porate governance problems that plague transi-tion economies lacking such legal protection. Butthe type of concentrated ownership matters aswell. Enterprises controlled by strategic inves-tors, particularly if they are foreign, have per-formed much better than those controlled byholding companies or other financial institutions.

Overview

xxvi

The selection of strategic investors matters too.Enterprises sold through transparent tenders orauctions have generally attracted better owners,outperforming enterprises sold directly to politi-cally connected parties, frequently at highly sub-sidized prices. Without such safeguards, concen-trated ownership does not avoid the risk ofexpropriation of assets and income belonging tominority shareholders.

In countries where the preferred method ofprivatization—direct sales through transparenttenders or auctions to strategic investors—wasunavailable, the relevant comparison in assess-ing privatization is not between the actualmethod chosen and the ideal method, but be-tween the actual method and continued stateownership until strategic investors were found.Countries where mass privatization usingvouchers originally dispersed ownership andwhere secondary trading has not led to trans-parent consolidation of shares witnessed expro-priation of assets and income of minority share-holders by those that were able to gain controlover the enterprise in the first stages of theprivatization program. But the success of con-tinued state ownership is not assured unlessthere is a political commitment to transparentprivatization outcomes and a minimum insti-tutional capacity to prevent asset stripping byenterprise managers in the interim period. In-deed, navigating between continued state own-ership with eroding control rights on the onehand and a transfer to ineffective new privateowners with an inadequate institutional frame-work on the other hand was one of the mostdifficult challenges confronting policymakers incharge of privatization. Irrespective of the al-ternative chosen, governments need to enshrineinvestor protection in the legal system andsupplement it with a system of regulation forfinancial intermediaries, such as investmentfunds and brokers (Johnson and Shleifer 2001).

Developing laws and institutions to protectinvestors and monitor managerial behavior andthus facilitate the development of bank and non-bank financial intermediation in countries withno recent market experience is far more diffi-cult when opposed by early winners from tran-sition. Further reforms would dissipate the rents

accruing to the early winners. In Russia, forexample, powerful insiders with a stake in weakcorporate governance have frequently hamperedthe work and enforcement efforts of the Secu-rities and Exchange Commission. In environ-ments of high state capture, privatization hasnot created enough demand for the enforcementof property rights. Indeed, quite the opposite istrue. While privatization is positively associatedwith public governance (the latter summariz-ing the state’s capacity to provide key publicgoods) in low-capture environments, the asso-ciation is negative in high-capture environments(EBRD 1999).

Several of these issues were not foreseen atthe beginning of the transition. One was the ap-parent stability of such partial reform equilibri-ums. Another was the unexpectedly perverse re-lationship between privatization and the qualityof governance in such environments. That in-creased the challenge of enhancing creditors’ andshareholders’ rights; promoting internationallyrecognized accounting and auditing standards;and enforcing takeover, insolvency, and collat-eral legislation in the face of opposition from anarrow set of entrenched private interests.

The need to strengthen corporate governance,despite opposition from oligarchs and insiders,is an important lesson from the first decade oftransition. This is, however, a time-consumingprocess, during which policymakers still need tomake choices about the appropriate stewardshipof state assets, including privatization. The fol-lowing broad principles should guide a programof privatization:

• Privatization should be part of an overallstrategy of discipline and encouragement.

• Small enterprises still owned by the stateshould be sold directly to new ownersthrough an open and competitive auction andwithout restrictions on who may bid for theshares.

• In general, medium-size and large enterprisesshould be privatized to strategic outside in-vestors, who, with a concentrated controllingstake, will best use enterprise assets. Althoughseveral transaction methods may be used, in-cluding negotiated sales, the evidence suggests

Overview

xxvii

this can be brought about most effectivelythrough competitive case-by-case methods,which are more deliberative than voucherschemes or rapid, small auctions. They useindependent financial advisors who both pre-pare the enterprise for sale and act as salesagents on behalf of the state. Rapid privati-zation to insiders or through mass privatiza-tion should be avoided. In countries such asBelarus, Turkmenistan, and Uzbekistan, eachof which has a substantially unfinished agendaof privatization of medium-sized and large en-terprises, but where the state retains the ca-pacity to provide public goods, the state shoulduse its administrative capacity to control thedisposition of public assets as transparentlyas possible, while developing institutions ofcorporate governance. Transparency would beenhanced, for example, if decisions regardingpublic assets were to be reviewed by indepen-dent boards of directors and accompanied bypublic disclosure.

• Privatization should be accompanied by in-creasing competition in the market for theproducts sold by the enterprise in question andvigorously enforced by the competition policyauthority. This can help discipline managerswhen corporate governance is weak.

• Divestiture of enterprises in sectors charac-terized by a natural monopoly or oligopoly(becoming rarer with advances in technology)must proceed with caution, if at all. Estab-lishing an efficient regulatory regime is a pre-requisite to protect the public interest, lestdivestiture transform an inefficient publicmonopoly into a poorly regulated or unregu-lated private monopoly.

• The state’s property and cash flow rightsshould be clarified and strengthened in en-terprises in which the state continues to holda stake.

Mobilizing the Winners from Further Reform

Breaking the political economy equilibrium un-derlying partial reforms is the most important anddifficult challenge in advancing transition in manycountries of the region, particularly in the CIS.Where the state is already susceptible to influence

by powerful vested interests in the new privatesector, granting extraordinary decree-makingpowers to the executive branch to dissipate rentsand level the playing field has not won againststrong opposition from insiders and oligarchs.

What is needed instead is to mobilize throughgreater political inclusion and coordination allconstituencies that lose from partial reform andthat stand to gain from further progress towarda more competitive market economy. For ex-ample, given the wide and generally regressiveimpact of high inflation, political parties in sev-eral transition economies mobilized enough elec-toral support for macroeconomic stabilizationto overcome the opposition of powerful com-mercial banks and other actors that gained fromeconomic volatility. Similarly, banking crises inBulgaria, the Czech Republic, and Hungarysparked electoral appeals to disgruntled saversthat helped break the stalemate over such issuesas banking privatization and regulatory reform.

Business associations could serve as vehiclesof collective action by SMEs, new enterprises,and “second-tier” enterprises that suffer froman unlevel playing field, discretionary taxationand regulation, and anticompetitive barriers. Incountries with concentrated political regimes,such associations are weaker than those in thecompetitive democracies of Central Europe,which have more voice. So political parties haveyet to seek strategic alliances with such actors asan alternative base of support and funding.

Overcoming the coordination dilemmas ofmobilizing the highly dispersed winners of fur-ther reform is not easy. A major challenge forthe reformist team that comes to power duringa period of extraordinary politics in countrieswith concentrated political regimes is to makeclear the links between rents from partial re-form and the direct costs to society. Tax arrears,tax and duty exemptions for high-profile con-glomerates, and nonpayments need to be linkedin the public mind to delayed public sectorwages and pensions and the poor provision ofsocial services. The complex web ofnontransparent subsidies to powerful businessesneeds to be uncovered, revealing that such sub-sidies tend to benefit incumbent managers ratherthan workers.

Overview

xxviii

To advance reforms, governments should fo-cus on smoothing the curves of the winners andlosers at the initial stages of reform (see figure1). This means lowering the adjustment costs forpotential new entrants and reducing the highconcentration of gains to oligarchs and insiders.One way to do this is by strengthening the pro-vision of basic public goods, such as secure prop-erty rights and a legal and judicial system. An-other way is by reducing excessively highmarginal tax rates and broadening the tax basethat promotes entry of enterprises from the un-official to the official economy. This can breakthe vicious cycle of informalization, lower taxrevenue, and further intensification of tax rateson a shrinking base. Developing a rule-based taxadministration to enforce efficient taxation ofthe new private sector is also important.

To align the incentives of local governmentsto identify with small business and increase en-try, taxes on small enterprises should be allocatedto local government. Simplifying entry and li-censing arrangements for new enterprises is criti-cal. These measures, by encouraging the emer-gence of new enterprises, offer a stable outsideoption to state workers, creating opportunitiesfor them to become potential entrants into theburgeoning sector of new enterprises and less-ening their opposition to reform.

As entry occurs gradually at the margin, theseactors become an effective constituency demand-ing reforms to remove weaknesses in the invest-ment climate over the long run. Furthermore, agradual reallocation of public expenditures fromnontransparent and discretionary subsidies toworker training, severance payments, and grantsfor improving services in communities affectedby downsizing can create further momentum forreform. More broadly, strengthening the socialsafety net and divesting such social assets as hous-ing, child care, and health facilities shifts theburden of social protection from enterprises togovernments, thus facilitating the restructuringthat will foster a return to growth. Fiscal policytherefore has the potential to smooth the curvesdescribed in figure 1 and redistribute a part ofthe reform dividend to those who would other-wise bear its costs. It is thus a key element insupporting comprehensive reform.

Conclusions

Analysis of the first ten years of transition inEastern Europe and the former Soviet

Union highlights the following lessons, whichcould be applied in future to economies that havemade limited progress with reform.

• While the initial conditions that prevailed atthe beginning of transition were critical forexplaining the output decline that occurredinitially in all countries, market-orientedpolicy reforms have played a significant rolein promoting subsequent economic growth.Creating an environment that disciplines oldenterprises into releasing assets and labor andencourages new enterprises to absorb thoseresources and undertake new investment,without tilting the playing field in favor ofany particular type of enterprise, is centralto economic growth.

• Policymakers cannot postpone the pain of liq-uidating and restructuring the old sector untilthe cushion provided by new enterprises is inplace. The success of the encouragement strat-egy requires simultaneous application of dis-cipline, because a lack of discipline underminesthe level playing field between different kindsof enterprises. Furthermore, the practice ofallowing old and large enterprises to avoidpaying taxes and social security contributionsand to avoid repaying bank debts has been atthe root of macroeconomic crises.

• Developing legal and regulatory institutionsto oversee enterprise management, thoughtime-consuming, is important. In the mean-while, where direct sales of state assets to stra-tegic investors—a preferred method ofprivatization—is not feasible, policymakersface a difficult choice between (i) privatizationto ineffective owners in a context of weakcorporate governance, with the risk of ex-propriation of assets and income of minorityshareholders by those who gained controlover the enterprise, and (ii) continued stateownership in the face of inadequate politicalcommitment to transparent privatizationoutcomes and limited institutional capacityto prevent asset stripping by incumbent en-terprise managers.

Overview

xxix

• Breaking out of low-level equilibrium trapsin which the immediate beneficiaries of lib-eralization and privatization have capturedthe state and oppose measures of encourage-ment such as competition and free entry thatwould reduce their rents requires mobilizingsmall, medium-sized, and new enterprises andsecond-tier businesses that suffer as a resultof the uneven playing field and stand to gainfrom further reform. Fiscal policy has an im-portant role to play here, by redirecting sup-port away from ailing enterprises and towardworker training and severance payments, andby divesting social assets such as housing,child care, and health facilities from enter-prises to governments.

Annex 1. Discipline and Encouragement:The Reform Agenda

What policy and institutional reforms areneeded to create an environment favor-

able to discipline and encouragement? While noindividual policy can be assigned to a single out-come in an interrelated system, it helps to thinkof policy packages as those primarily directed atdisciplining the old sector and those primarilydirected at encouraging the new sector withouttilting the playing field in favor of any particulartype of enterprise.

Discipline

In an environment of price and trade liberaliza-tion, discipline requires imposing hard budgetconstraints on enterprises, providing exit mecha-nisms for insolvent enterprises, monitoring andinfluencing managerial behavior to reward effi-cient stewardship of assets and to discouragetunneling and theft, increasing product marketcompetition, transferring social assets from en-terprises to local governments, and using thesocial safety net as a cushion for displaced work-ers and other losers from reform.

Imposing hard budget constraints requires:

• Eliminating tax exemptions, fiscal and finan-cial subsidies, budget and tax offsets, anddirected credits.

• Controlling fiscal risks arising from implicitand contingent liabilities on account of state-owned enterprises, banks and pension sys-tems, guarantees for projects and balancesheets of special purpose agencies, and thefiscal stance of subnational governments.

• Implementing bankruptcy laws to facilitate exitthrough a formal process and to create incen-tives for closure through informal mechanisms.

• Reforming the budget process so the state canmeet its expenditure obligations in cash andon time, with a view to eliminating arrearsand noncash payments.

Monitoring and influencing managerial be-havior requires:

• In reforming economies, where the bulk ofassets are in private hands, improving insti-tutions of corporate governance by strength-ening legal protection for minority sharehold-ers and creditors, bringing in managementby concentrated owners or strategic investors,promoting internationally recognized ac-counting and auditing standards, and work-ing, as capital markets grow, to develop amarket in corporate control.

• In nonreforming economies, where assets re-main largely in the public sector and wherethe state retains the capacity to provide pub-lic goods, the state should use its administra-tive capacity to control asset disposition astransparently as possible, while developinginstitutions of corporate governance.

Promoting competition in product marketsis an important ingredient of discipline, especiallybecause of competition’s effect on the behaviorof enterprise managers. In the CIS, for example,competition can compensate to some extent forweak monitoring of managers by shareholdersand creditors. This requires:

• Opening markets and promoting free entry,including trade liberalization

• Enforcing competition laws through a gov-ernment agency vested with the requisiteauthority.

Transferring responsibility for social assets suchas housing, utilities, clinics, and kindergartens from

Overview

xxx

enterprises to local governments will allow enter-prise restructuring to proceed in countries that haverelied heavily on enterprises as instruments of so-cial policy. This requires:

• Clarifying the roles and responsibilities ofsubnational governments

• Giving those governments resources to ful-fill their assigned responsibilities

• Reforming communal services (including cen-tral heating and gas) and moving tariffs tocost-recovery levels

• Replacing across-the-board housing and util-ity subsidies with targeted social assistanceto the poorest households, sometimes com-bined with lifeline tariffs.

Social insurance programs that cover pen-sions and unemployment insurance and socialassistance programs make it more feasible to dis-cipline old enterprises into shedding labor andto help create a constituency for discipline. Re-form of social insurance programs requires:

• Moving pension reform in Central Europeand the Baltics to multipillar systems, with aminimum poverty-based benefit. Becausestructural unemployment is falling in thesecountries, they can implement unemploymentinsurance programs.

• Reforming pay-as-you-go systems in Bul-garia, Romania, Russia, and Ukraine to putthem on a firmer fiscal footing, with a mini-mum poverty-based benefit. Unemploymentinsurance might involve a flat benefit or sev-erance payment.

• Moving to a flat benefit structure in theresource-constrained, low-income countriesof the CIS, to protect the poorest elderly.

The reform of social assistance programs re-quires:

• Moving to a means-tested cash benefit assis-tance program in Central Europe and the Baltics

• Moving to categorical cash benefits, eitheruniversal or targeted by category, in Bulgaria,Romania, Russia, and Ukraine and usingmeans-tested benefits only where local insti-tutions are strong

• Improving targeting of limited cash benefitsin the low-income CIS countries throughgeographical targeting, community-basedidentification, or self-targeting through someform of public works scheme.

Encouragement

In addition to liberalizing prices and trade, im-proving the investment climate for domestic andforeign investors is key to encouraging new en-terprises. This requires establishing secure prop-erty and contract rights and providing basic in-frastructure, reducing excessive marginal taxrates, simplifying regulatory and licensing pro-cedures, and developing a competitive and effi-cient banking system.

Ensuring the adoption of laws best suited tosecuring property rights and contracts requiresemphasis on two areas:

• The process for drafting laws: enterprisesshould have input in their design and be in-formed beforehand of changes in rules thatwill affect their operations

• The effectiveness of the judiciary: its fairness(bias, honesty, and consistency) and the like-lihood of enforcement.

Tax reform should:

• Raise the turnover threshold for becoming avalue-added taxpayer high enough to excludesmall enterprises, which should instead besubject to a small-enterprise tax regime. Thisregime should be simple enough to lightenthe administrative and reporting burden ontaxpayers and reduce interactions betweenthe taxpayer and the tax authority.

• Allocate small business taxes and propertytaxes to subnational governments to helpthem identify with new emerging enterprises,which are typically a source of job creation,rather than with large, bankrupt enterprisesto save jobs that should be cut.

Streamlining the business licensing and reg-istration requirements that govern entry of newenterprises is a high priority. Addressing theseissues requires:

Overview

xxxi

• Simplifying and making transparent entryand licensing procedures

• Reducing the scope for arbitrary decision-making and abuse of power.

Insecure property rights are more of a con-straint on investment for new enterprises thanthe availability of bank finance, particularly inthe CIS, which is less far along in the transition(Johnson, McMillan, and Woodruff 2000). Butas the transition proceeds and legal and judi-cial reforms strengthen property rights, finan-cial deepening and development will be essen-tial to support the growth of a private sectorled by new enterprises.

Developing a competitive banking sector re-quires a strategy to deal with the exit of failedbanks, the entry of new banks, and bankprivatization and restructuring. Strengthening theregulatory authority for bank supervision is ex-tremely important for performing these activi-ties effectively. This will require:

• The resolution of failed banks, which domi-nate the banking system in transition econo-mies and have large interbank exposures andloans to loss-making enterprises, thereforeposes special problems. Liquidation and re-structuring options should be assessed care-fully in the event of systemic risks to the fi-nancial sector and the need to impose hardbudget constraints on banks and enterprises.

• The fostering of a competitive and efficientbanking system requires encouraging the entryof new banks that satisfy prudential criteria forminimum capital requirements and capital ad-equacy. Because supervisory responsibility re-sides mainly with the home country regulatoryauthority, entry by foreign banks and acquisi-tion of stakes in existing banks by foreign banksis a quick way of importing managerial andgovernance expertise and improving bank regu-lation in a transition economy. In any event,the expansion of the banking system shouldoccur only in line with the growing capacityfor bank regulation and growth in the numberof creditworthy borrowers.

• The privatization of banks to strategic in-vestors whenever possible. If foreign, they

can help upgrade managerial and supervi-sory standards. The alternative—privatizingbanks to concentrated owners—should oc-cur only if there is a clear separation betweenshareholders and borrowers. The pace ofprivatization should not outrun the devel-opment of adequate supervision authority.

Notes

1. The CSB comprises Albania, Bosnia andHerzegovina, Bulgaria, Croatia, the Czech Republic,Estonia, Hungary, Latvia, Lithuania, the formerYugoslav Republic of Macedonia, Poland, Romania,the Slovak Republic, and Slovenia. The Federal Re-public of Yugoslavia is not included in the aggregatesfor the CSB countries because no data are availablebefore 1998. The CIS comprises Armenia, Azerbaijan,Belarus, Georgia, Kazakhstan, the Kyrgyz Republic,Moldova, the Russian Federation, Tajikistan,Turkmenistan, Ukraine, and Uzbekistan.2. The terms “Russia” and “Russian Federation” areused interchangeably and refer to the same country.

3. Conceptual and measurement problems plague theGDP data (see box 1.1). However, these problems donot modify the qualitative thrust of the statementspresented here.

4. These estimates are based on 1993 purchasingpower parity rates (World Bank 2000a).

5. The agenda of accession to the European Unionlooms large for countries in Central and Eastern Eu-rope. The World Bank has been engaged in a majorproject that examines country-specific and subregionalissues arising from European Union accession. Theprincipal outputs of the project are listed in the bib-liographic guide; therefore, that subject will not becovered in this report.

6. Kornai (1986) introduced the term soft budgetconstraint (the opposite of a hard budget constraint)to characterize the environment faced by state enter-prises in Hungary in the 1980s.7. The analysis is based on the assumption that datafor small enterprises can be used to approximate newenterprises. The approximation is not accurate inas-much as the set of small enterprises includes small oldenterprises. Annex 4.1 presents an upper bound forthe error committed by this assumption in the CSBcountries. While the estimate for the upper bound issubstantial in 1995, it shrinks significantly under rea-sonable assumptions about the mortality rate of en-terprises by 1998, the year for most of the data in thisreport. The terms “new enterprises” and “small en-terprises” are henceforth used interchangeably.

Overview

xxxii

8. Annex 4.2 describes the circumstances underwhich comparisons based on labor productivity (forwhich data are available) correspond to comparisonsbased on total factor productivity, which is the rel-evant concept for this report.

Part 1

The First Decade in Transition

3

1How Did Transition

Economies Perform?

The 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 mostevident 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, whatwas 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 inproductivity (figure 1.2). Even for these countries, though, the initial fall was larger than anticipated.1

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 RussianFederation the next year.

Only now is there evidence of growth being restored in this group of countries. The magnitude andduration of the transition recession was, for all countries, comparable to that for developed countriesduring the Great Depression, and for most of them it was much worse (table 1.1). The CIS had anaverage of 6.5 years of declining output, resulting in the loss of half the initial level of measuredoutput. 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 amongthe CSB. In this, they are much closer to the average of the CIS than to other CSB countries. In theCIS 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 hadmild declines.

Transition—The First Ten Years: Analysis and Lessons for Eastern Europe and the Former Soviet Union

4

The transition recession is now over—Ukraine,the only country with 10 consecutive years ofoutput decline, registered growth in 2000. How-ever, the recovery has not been smooth in all coun-tries. Three CSB countries (Bulgaria, the CzechRepublic, and Romania) had at least two years of

output decline after their initial recovery. Fourothers had one-year recessions: Albania in 1997,as the collapse of a pyramid scheme led to a fi-nancial crisis and civil disturbances, and Croatia,Estonia, and Lithuania in 1999, largely becauseof the Russia crisis in August 1998.

FIGURE 1.1.

Changes in Real Output, 1990–2001

Note: Europe and Central Asia is the average of the CSB and the CIS. All aggregates are population-weighted. Valuesfor 2001 are projected.

Source: World Bank country office data.

Index (1990 = 100)120

110

100

90

80

70

60

501990 20011991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Central and SoutheasternEurope and the Baltics

Commonwealth ofIndependent States

Europe and Central Asia

FIGURE 1.2.

Output Growth Rates, 1990–2001

Note: Europe and Central Asia is the average of the CSB and the CIS. All aggregates are population-weighted. Valuesfor 2001 are projected.


Annual rate (percent)8

4

0

–4

–8

–12

–161990 20011991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Central and SoutheasternEurope and the Baltics


Europe and Central Asia

How Did Transition Economies Perform?

5

Industry Shrank—Services Grew

Both the CSB and CIS started the transitionwith a much larger industrial sector and a

much smaller service sector than market

TABLE 1.1.

The Transition Recession

Consecutive years Cumulative output Real GDP, 2000Countries of output decline decline (percent) (1990 = 100)

CSBa 3.8 22.6 106.5

Albania 3 33 110Bulgaria 4 16 81Croatia 4 36 87Czech Republic 3 12 99Estonia 5 35 85Hungary 4 15 109Latvia 6 51 61Lithuania 5 44 67Poland 2 6 112Romania 3 21 144Slovak Republic 4 23 82Slovenia 3 14 105

CISa 6.5 50.5 62.7Armenia 4 63 67Azerbaijan 6 60 55Belarus 6 35 88Georgia 5 78 29Kazakhstan 6 41 90Kyrgyz Republic 6 50 66Moldova 7 63 35Russian Federation 7 40 64Tajikistan 7 50 48Turkmenistan 8 48 76Ukraine 10 59 43Uzbekistan 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).

economies with comparable per capita incomes.During the transition, the industrial sectorshrank, falling to about a third of the economy,and the share of services grew to about half (table1.2). Perhaps less expected, the increase in


6

services in the CIS took place at the expense ofindustry and agriculture, which both declined byabout 9 percent of GDP. These sectoral shifts andthe decline in total output mean that output inagriculture and manufacturing is now 40–45 per-cent of its pretransition level.

Private Enterprises Overtook the State Sector

The most basic transformation was movingresources from the state to the private sec-

tor. In 1999 the private sector in most countriesin the region produced more than half of GDP(table 1.3). The share was much larger, more than

70 percent, for such advanced reformers as theCzech Republic, Estonia, and Hungary. In viewof the rapid growth of GDP in the CSB in recentyears and the slow privatization of large stateenterprises, the increase in share testifies to a dra-matic increase in new private sector activity. (Therole of new enterprises in promoting economicgrowth is discussed in part 2.)

Exports Rose—Moving Toward IndustrialCountries

Countries more advanced in their recoveryhave been more successful in increasing

their exports and reorienting them to the indus-trial countries. However, even in countries whoseoutput did not grow between 1993 and 1998, suchas Moldova, Russia, and Ukraine, real exportsincreased (table 1.4).

Important in the recovery of output was di-rect investment from abroad. These flows areimportant not only as a source of capital and newtechnology to modernize industries and extractnatural resources, but also as a signal of confi-dence in the transition to a market economy. Dur-ing 1996–99 more than US$70 billion in directinvestment came to the region, most of it to theCSB (table 1.5). In the CIS foreign direct invest-ment was largely confined to the energy-rich coun-tries, with Azerbaijan, Kazakhstan, and Russiareceiving 75 percent of the total. Russia’s sharewas even lower than several of the CIS countries’shares, despite the considerably greater size of itseconomy and resource endowment.

Poverty Increased Sharply

The foregoing discussion suggests that theinitial fall in output may overestimate the

decline in living standards during the transition(box 1.1). However, the period was still one ofextreme hardship for many people (World Bank2000b). Although extreme poverty is still lowerin the transition economies than in other devel-oping countries, it increased sharply during thedecade (table 1.6).2 In 1998 one of every 20people in the transition economies had per capitaincomes below US$1 a day, up from fewer thanone in 60 a decade before. Moreover, the increase

TABLE 1.2.

Composition of Output, 1990–91 and 1997–98

Percentage of GDP

Regions and periods Agriculture Industry Services

CSB1990–91 13.7 45.1 41.21997–98 13.9 33.0 53.1

CIS1990–91 27.5 39.7 32.81997–98 18.7 31.2 50.1


TABLE 1.3.

Private Sector Growth, 1990s

Percentage of GDP

Countries 1990 1994 1999

CSB 11 50 68Czech Republic 12 65 80Estonia 10 55 75Hungary 18 55 80Romania 17 40 60

CIS 10 20 50Armenia 12 40 60Belarus 5 15 20Russian Federation 5 50 70

Source: EBRD (2000).


7

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 67.5Albania 22.0 62.9 94.1Bulgaria 4.3 55.1 59.0Czech Republic 10.4 29.9 69.3Estonia 10.8 25.9 71.3Hungary 11.1 67.4 81.5Macedonia, FYR 7.3 22.2 65.5Poland 12.9 71.6 75.5Romania 8.7 44.3 71.0Slovak Republic 6.9 15.9 59.2Slovenia 5.7 33.7 70.7

CIS 3.2 28.0 29.0Armenia –8.6 9.4 34.9Azerbaijana 14.0 4.2 20.0Belarus –3.2 15.3 11.0Georgiaa 10.3 2.3 25.9Kazakhstan 3.4 43.8 29.6Kyrgyz Republic –2.4 24.7 44.0Moldova 4.8 6.2 31.3Russian Federation 4.7 59.3 49.4Ukraine 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 US$ 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 4.6Estonia 647 3.9 1,050 5.2Hungary 9,399 5.7 6,979 3.8Poland 2,540 0.6 17,096 2.9

CIS 8,272 1.0 22,001 2.5Azerbaijan 237 4.2 3,222 20.9Kazakhstan 2,357 2.7 4,971 6.4Russian Federation 3,965 0.3 8,412 0.7Turkmenistan 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.


8

BOX 1.1.

Limits of GDP Statistics for Transition Economies

Estimating GDP and using it as an indicator of living standards have some well-recognized problems. These range from data collection andaggregation issues to the omission of nonmarketed goods (pollution and family services) and depletions of exhaustible resources. Official GDPalso has special limits as a welfare indicator for transition economies, particularly when comparing output performance to the pretransitionperiod. These limits fall into three groups: index number problems, omission of informal activities, and effects of the changes on the compositionof output.

Real GDP is an aggregate index constructed by weighting the outputs of individual products according to their respective prices. To beprecise, the aggregation is done using value added (that is, the output price minus the cost of intermediate inputs) rather than prices. Whenrelative 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 verysensitive to the base period for the aggregation. Moreover, the standard relations among different indexes do not hold when the initial prices arenot 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 theinitial changes in output are accurate measures of changes in welfare. Eliminating queuing, for example, with the same amount of real outputclearly improves welfare. However, the increase in the relative price of consumer goods (say, for housing and utilities) means that the sameoverall “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 thatcoming from the private sector. This, compounded by the need to go from net material product to GDP, which includes services, meant thatofficial output statistics did not capture the rapid growth of the informal sector. Subsequently, tax evasion and pressure from regulations andpublic sector bureaucracy provided incentives to operate business in the informal economy. Estimates of the size of the informal sector, usingvarious methods, suggest that its share of GDP varies enormously across the region, from 6 to 60 percent of GDP.

In addition, during the transition there were sharp changes in the composition of output, reducing the proportion of goods from whichconsumers derived 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 standardof living.

TABLE 1.6.

Average Poverty Rates, 1990 and 1998(percent)

Population living on lessthan US$1 a day

Regions 1990 1998

Eastern Europe and Central Asia 1.5 5.1East Asia and Pacific 28.2 15.3Latin America and the Caribbean 16.8 15.6Middle East and North Africa 2.4 1.9South Asia 43.8 40.0Sub-Saharan Africa 47.0 46.4

Total 20.0 17.1

Source: World Bank data.

in poverty was much larger and more persistentthan many expected at the start of the process.Even in the most successful countries—such asPoland, where poverty came down steadily fromits peak in 1994—poverty rates were still higherin 1998 than in 1991 (World Bank 2000b).

Poverty increased not just because of the fallin output, but because of greater inequality inthe distribution of income. Inequality increasedin all transition economies, with great variationacross the region (table 1.7). In some cases, theincrease was modest—as in Hungary, where theGini coefficient for per capita income rose from0.21 in 1987 to only 0.25 a decade later. Even inthe Czech Republic and Slovenia, where inequal-ity rose more, the distribution of income remainsfairly egalitarian. Yet in the CIS and elsewhere,


9

the increases in inequality have been unprec-edented. In Armenia, Russia, Tajikistan, andUkraine, the level of inequality as measured byGini coefficients has nearly doubled.

Notes

1. For example, one of the most authoritative earlyreports on the state of the socialist economies, A Studyof the Soviet Economy (IMF and others 1991), antici-pated an early recession that would end by mid-decade at the latest (or earlier, under a “radical reforms”scenario). The report projected an annual averagegrowth rate for the 1990s of 1.1 percent (3.3 percentunder the radical reform scenario), which only twocountries, Poland and Slovenia, were able to achieve.

2. International comparisons of poverty rates arefraught with problems of estimation and interpreta-tion. In transition economies in Europe and CentralAsia, harsh winters, deteriorating housing stocks, utili-ties designed for pretransition conditions, subsidizedenergy prices, and similar conditions mean that fami-lies with US$2 a day are likely to have a lower stan-dard of living than families with the same income inother regions. Furthermore, incomes do not alwaysmatch people’s assessment of their economic welfare,which tends to depend also on their deprivation rela-tive to other people and their previous condition.

TABLE 1.7.

Changes in Inequality during the Transition, VariousYears

Gini coefficient of income per capita

Countries 1987–90 1993–94 1996–98

CSB 0.23 0.29 0.33Bulgaria 0.23 0.38 0.41Croatia 0.36 — 0.35Czech Republic 0.19 0.23 0.25Estonia 0.24 0.35 0.37Hungary 0.21 0.23 0.25Latvia 0.24 0.31 0.32Lithuania 0.23 0.37 0.34Poland 0.28 0.28 0.33Romania 0.23 0.29 0.30Slovenia 0.22 0.25 0.30

CISa 0.28 0.36 0.46

Armenia 0.27 — 0.61Belarus 0.23 0.28 0.26Georgia 0.29 — 0.43Kazakhstan 0.30 0.33 0.35Kyrgyz Republic 0.31 0.55 0.47Moldova 0.27 — 0.42Russian Federation 0.26 0.48 0.47Tajikistan 0.28 — 0.47Turkmenistan 0.28 0.36 0.45Ukraine 0.24 — 0.47

— Not available.a. Median of countries with data.Source: World Bank (2000b).

11

The search for explanations of economic outcomes—causes, differences in magnitude, varia-tions in speed and sustainability—has spawned a large literature. The explanations focus onthe 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 alsobeen examined to explain why economies may be trapped in situations of partial reform—in a noman’s land between plan and market, where the early gainers from reform vigorously oppose furtherprogress toward a market economy.

Did Initial Conditions Affect Performance?

Several 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 developmentunder 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 intothree categories: structure, distortions, and institutions.1

Structure encompasses such variables as the share of industry, the degree of urbanization, the shareof trade with the socialist block, the richness of the natural resource endowment, and the initialincome. These can be described as follows:

• Share 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-incomecountries are generally more urbanized. The proportion of people in urban areas in 1990 ranged

2Explaining Variation in

Output Performance


12

from around 70 percent (Estonia, Latvia,Russia) to around 30 percent (Albania,Kyrgyz Republic, Tajikistan).

• Trade dependence. High trade dependence onother communist countries (measured by theratio of Council of Mutual Economic Assis-tance exports and imports to GDP) reflectedthe level of industrialization under centralplanning, which favored large plants and re-gional interdependence.2 Inter-republic flowswere especially large for the smaller repub-lics of the Soviet Union, which had little tradeoutside the area.

• Natural resource endowment. Several coun-tries in the region—Azerbaijan, Kazakhstan,Russia, and Turkmenistan—have rich butunderdeveloped deposits of oil and gas. Thisgave them the potential for rapid growth, butrequired large investments to make produc-tion and transportation possible. Some energyexporters have tended to delay reforms—withdeleterious effects on growth.

• Income. Incomes (in 1989 dollars adjustedfor purchasing power parity) were generallyhigher in Central Europe and the Europeanpart of the Soviet Union, ranging fromUS$1,400 per capita in Albania to US$9,200in Slovenia.

Distortions in the economy refer to such fac-tors as repressed inflation, black market ex-change rates, trade shocks arising from the dis-solution of the Soviet Union, the extent of prioreconomic reform within the centrally plannedsystem, and the degree of economic stagnationprior to the transition.

• Repressed inflation. Most transition econo-mies had repressed inflation, measured hereas the difference between the increase in realwages and real GDP from 1987 to 1990. Thisindex was high for the Soviet Union, pro-pelled by the federal government’s gradualweakening of control over wages and regionalbudgets associated with the partial liberal-ization of the Gorbachev reforms.

• Black market exchange rates. This variable isdefined as the difference between the blackmarket exchange rate and the official exchangerate, indicating the rationing of foreign

exchange as well as a subsidy to imports andtax on exports. For the Soviet Union, the blackmarket exchange rate reached as high as 1,800percent in 1990.

• Terms of trade loss for the CIS. Trade flowswithin the Council of Mutual Economic As-sistance took place at administrative pricesthat were not directly linked for the most partto world prices. This meant large changes inthe terms of trade after trade was liberalized.The indicator measures the terms-of-tradeloss as a share of GDP, as calculated by DavidTarr (1994). Small energy importers, such asMoldova and the Baltic states, suffered thelargest proportional losses (more than 10percent of GDP). Countries that were netenergy exporters generally had gains.

• Reform history. Some countries (Hungary,Poland, Federal Republic of Yugoslavia, andto a lesser extent, Bulgaria) introduced someelements of market-based reforms before thecollapse of the Soviet Union. The indicatorcapturing reform history is the World Bank’sIndex of Liberalization in 1989 (de Melo,Denizer, and Gelb 1996).

• Pretransition growth rate. Looking at theaverage growth in 1985–89, the more ma-ture countries stagnated in the late 1980s,while the poorest countries on average hadhigher growth.

Institutions encompass such variables asyears under central planning, location in rela-tion to Western markets, and experience withnationhood.

• Market memory. Some countries could drawon their market experience before the Sovietperiod in the design of an institutional-legalframework supporting markets at the startof the transition.

• Location. Countries in Central Europe, par-ticularly those bordering on the West, hadmore extensive trade links with marketeconomies and enterprises and institutionswere more exposed to competitive pressures.Individuals had more freedom of travel, al-lowing more exposure to Western markets.

• New states. Countries with little experienceas independent nation states may have had

Explaining Variation in Output Performance

13

more difficulty creating efficient political in-stitutions and achieving political consensus.

External Economic Shocks Delayed Recovery

The onset of transition was accompanied bysevere shocks. The collapse of the institu-

tional and technological links of the Soviet cen-trally planned system disrupted the supply of in-puts for production and the delivery of outputs,posing new challenges for enterprises. The loss ofbudget transfers from the center and the elimina-tion of subsidized energy imports were severeblows, particularly to some of the newly indepen-dent states of the CIS. The broader external eco-nomic environment was also less favorable in the1990s, and the transition thus coincided withlower growth rates in other developing countries.3

The various financial crises of the 1990s—Mexico, East Asia, and particularly Russia—alsocontributed to delaying or interrupting the recov-ery of output (box 2.1). War and civil strife—inArmenia, Azerbaijan, and Tajikistan in 1992–94,in Georgia and Moldova in 1992, and in Croatiaand FYR Macedonia in 1991–94—took a majortoll on lives, infrastructure, and the state, under-mining the political consensus on reforms neededfor successful transition. (The political economyof reform in war-torn countries is examined inpart 3 of this volume.)

Policies—Do They Matter?

The shift from planned to market economiesis a social and economic transformation of

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. Formany countries the first effect was disrupted trade with Russia—as demand contracted and trade finance and payments system arrangementswere interrupted. CIS countries, such as Armenia, Azerbaijan, Kazakhstan, Tajikistan, and Ukraine, saw their exports to Russia decline up to 70percent in the nine months after the crisis. The global crisis also deepened the recession in Western Europe, hurting the export performance ofcountries in Central and Eastern Europe. Mounting unemployment in Russia and administrative controls to stem capital flight severely curtailedworkers’ remittances from Russia, whose real value also declined as the ruble depreciated. This was particularly damaging for Russia’s smallerneighbors. Between 1997 and 1999 Russian transfers abroad fell from US$771 million to US$493 million. Russian foreign direct investmentabroad 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 1998and 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 manycountries privatization prospects were dampened. Despite initial fears, however, net foreign direct investment and portfolio equity flows actuallyincreased in 1998 and 1999 over 1997.

The crisis exposed vulnerabilities in the financial sector of most countries, even those implementing strong adjustment programs before thecrisis. 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 becausethe 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 exchangeexposures 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 wasalready 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 percentin end-1998, and to 56 percent in mid-1999.

Since mid-1999 the massive real devaluation of the ruble and higher oil prices have fueled a rapid recovery of exports and economic activityin Russia, helped by world demand growth. Oil prices are expected to remain firm in the near term, 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.


14

blueprints. Although the specific design of poli-cies, their sequencing, and their speed of imple-mentation are still subject to debate, there was abroad consensus that reforms should include:

• Macroeconomic stabilization• Price and trade liberalization• Imposition of hard budget constraints on

banks and enterprises• Enabling environment for private sector de-

velopment• Reform of the tax system and restructuring

of public expenditure

• Legal and judicial reform• Reform of public sector institutions.

The question of whether these reform poli-cies matter can be tested against the alternativehypothesis that output levels and annual changesin those levels were determined primarily by ini-tial conditions and external economic shocks, asdescribed earlier.

The extent of economic policy reform has beenmeasured in a liberalization index developed bythe World Bank to quantify progress in the tran-sition to a market economy.4 This index measuresreforms needed to make markets the main mecha-nism for allocating resources, such as eliminatingcentral planning and mandated allocationsthrough government orders and creating condi-tions to allow private production. The index alsocovers reforms to ensure the efficient functioningof markets, such as stabilizing the macroeconomicenvironment, liberalizing the trade regime, andpursuing procompetition policies. The indexranges from zero to one, with zero representingan unreformed, centrally planned economy, andone denoting the standards of a market economy(figure 2.1). In 1998 the index was the highest forcountries in Central Europe and the Baltics andthe lowest for countries such as Belarus,Turkmenistan, and Uzbekistan, which had yet toembark on a course of reform.

A substantial literature uses cross-country sta-tistical analysis of the transition from 1990 to 1999to demonstrate that better policies are associated—and significantly so—with higher annual growthof GDP in Central and Eastern Europe and theCIS, even when controlling for the effects of ini-tial conditions and external economic shocks (seeannex 2.1 for a summary of the main findings ofthis literature). The analysis allows for consider-able variation in the nature of the relationship be-tween policies and growth. One hypothesis sug-gests that a minimum critical mass of reforms needsto be in place before economic reforms have thedesired effects on performance. Below this thresh-old, it is possible that additional reforms couldhave a negative impact on output—that is, a clas-sic J-shaped response of output to policy reform.

In this view, implementing a few limited re-forms could disrupt production in state enterprises

FIGURE 2.1.

Progress in Policy Reform, 1990s

Source: de Melo, Denizer, and Gelb (1996); EBRD (2000).

HungaryPoland

Czech RepublicEstonia

Slovak RepublicSlovenia

LatviaLithuania

CroatiaKyrgyz Republic

KazakhstanBulgaria

Macedonia, FYRGeorgia

MoldovaRomaniaArmeniaAlbaniaRussia

UkraineAzerbaijanUzbekistan

TajikistanBelarus

Turkmenistan

1.00.80.60.40.20

Liberalization index

1990 19981995


15

without generating an attractive climate for re-structuring and new investment, resulting in loweroutput. One empirical test of this hypothesis (deMelo and Gelb 1996) finds that when countrieshave made only limited progress in reform (de-fined as a score of less than 0.4 on the liberaliza-tion index) additional reforms actually have anegative impact on growth. The response turnspositive once a minimum threshold of economicreform has been reached. This suggests that thereare important complementarities among elementsof a market-oriented reform program that have adecisive impact on the relationship between re-forms and growth.

Alternatively, it might take a long time forpolicy reforms to exert their full impact on out-put, with the initially adverse consequences inthe short-run giving way to beneficial effects lateron. The empirical results reported here imply thatpolicy reforms affect output growth over an ex-tended period. Our analysis suggests that this“time lag” in the impact of reforms on perfor-mance can be estimated at approximately threeyears (the hypothesis that the effect is entirelycontemporaneous can be strongly rejected; seeSelowsky and Martin 1998).

There also are significant differences betweenthe CIS and the other transition economies in thenature of the response of output to policy. A par-ticular difference is on the immediate (contempo-raneous) impact of reforms, which is negative inthe CIS and positive in Central Europe. This meansthat liberalization has an up-front “investmentcost” in the CIS countries, consistent with greatereconomic distortions at the start of their transi-tions. A higher share of negative value activitiesrequired vast reallocation of resources after liber-alization in the CIS. These countries also facedgreater obstacles to achieving that reallocation,such as physical size, distance to external mar-kets, large and inefficient company-town enter-prises in isolated regions, and greater political andconstitutional turmoil inhibiting investment.

What Initial Conditions Matter and When DoThey Matter?

There are reasons to expect that the factorsexplaining the initial output collapse are not

entirely the same as those determining latereconomic performance. Some developments ac-companying the onset of transition, such as thebreakdown of payment systems, are likely to havehad a stronger impact on the initial period ofthe transition. Similarly, the effects of some un-favorable initial conditions, such as repressedinflation, are likely to dissipate over time.

In our statistical analysis, the unbundling ofinitial conditions into structure, distortions, andinstitutions provides a more nuanced answer to thequestion of the importance of initial conditionsversus policy reforms in explaining the recessionand recovery periods of the transition experience(annex 2.2). First, initial conditions are more im-portant factors in explaining the differences acrosscountries during the initial period of output de-cline (1990–94) than over the full 10 years of tran-sition. The three (aggregate) indicators of initialconditions defined earlier explain 51 percent of thevariation in the average rate of growth across coun-tries during 1990–94, but only 41 percent of thevariance in average growth during the decade.

Second, our results suggest that different typesof initial conditions were more significant in theearly and later stages of transition. Initial distor-tions in the economy—including factors such assevere repressed inflation or high black marketexchange rates and absence of pretransition policyreforms—are most closely associated with lowerperformance during the first years of transition.Initial institutions—including factors such as ab-sence of “market memory,” measured by the num-ber of years under socialism, and general devel-opment of national institutions, as determined bythe length of prior experience of nationhood—are more strongly associated with variations insubsequent performance.

Third, while initial conditions have a greaterimpact on the initial collapse of output than onthe subsequent recovery, the impact of policiesbecomes stronger as the transition progresses—although it is still significant in the early stages.Indeed, policy variables are statistically significantin both periods, implying that market-orientedpolicy reforms not only speed economic recoveryand promote growth in the medium-term, but alsomitigate the effects of the transition recession inthe short term.


16

What If Policies Themselves Are Endogenous?

Policies, in general, are endogenous. Thechoice might depend on initial conditions

that might influence the possibility of policy con-sensus behind reform (de Melo and others 1997).They might be influenced by the desire to reclaimnationhood after the dismantling of the SovietUnion. Alternatively, what is feasible at one stagemight be partly determined by previous policydecisions. For example, private monopolies thatemerged from privatization at an early stage ofreform might attempt to block entry or resist aregulatory framework. Or new, small entrepre-neurs may try to press for faster reforms in prop-erty rights and in the court system.

The literature reports some limited successin finding correlates to policy reform, such asthe initial and contemporaneous level of politi-cal freedom (Dethier, Ghanem, and Zoli 1999;de Melo and others 1997). Part 3 of this reportdescribes how choices about the structure of thepolitical system can shape the adoption of eco-nomic reforms. An important result from thecross-country evidence is that policy reforms re-main strongly significant in explaining outputperformance even when policies are taken asendogenous (that is, when output growth andpolicies are jointly determined in a simultaneousequation model; see de Melo and others 1997).

Does the Speed of Reform Matter?

The relationship between the speed of reformand economic growth has been the subject

of controversy. Some economists argued for ad-vancing reforms in all areas as fast as possible;others criticized such a strategy as imposing un-necessarily high cost. The most interesting partof the debate focuses on the sequencing of poli-cies—on the relative speed of different types ofreform. Advocates of moving fast in areas ame-nable to rapid reform argue that the synergiesamong different components—for example,privatization together with liberalization of pricesand trade—may generate enough gains and win-ners to maintain the reform momentum. Theneed to take advantage of windows of opportu-nity is also cited as important in that decision.

By contrast, advocates of slower reform pointout that going ahead with reforms that can beimplemented quickly—“stroke of the pen” re-forms—without waiting for those that take moretime, such as the creation of institutions that sup-port markets, significantly reduces the benefitsof these reforms. The loss could be so severe asto generate output losses and also lead to thecreation of interest groups opposing those re-forms requiring more time. Some of these issuesare reviewed on part 3.

The amalgamation of different types of policyreforms into a single aggregate indicator preventsstatistical cross-country analysis from directlyshedding light on the desirability of progressingrapidly along all dimensions of reform. Someindirect evidence is provided by the finding thatoutput in each year is significantly associatedwith the level of policy reforms achieved up tothat year—that is, with cumulative policy reform.So the quicker a high level of liberalization isreached and sustained the sooner the economycan attain higher growth.5

These results are best seen as a broad-brushcharacterization of the main contours of transi-tion. They do not provide the full story of thetransition. Although policies and initial condi-tions account for more than half the variabilityof output growth across countries and years, theystill leave substantial room for other factors in-fluencing growth.6 A full explanation of outputperformance would have to include more coun-try-specific factors—as well as shocks and otheromitted factors—and a detailed analysis of indi-vidual countries or smaller groups of countries.This is done in the rest of this report.

Annex 2.1. Summary of Cross-CountryEmpirical Literature on Growth in TransitionEconomies

A slund, Anders, Peter Boone, and SimonJohnson. 1996. “How to Stabilize: Lessons

from Post-Communist Countries.” BrookingsPapers on Economic Activity 26(1): 217–313.

This paper finds differences in the determinantsof output changes during 1989–95 (no significantassociation with policy reforms) and end-of-periodoutput level (liberalization and inflation significant).


17

Berg, Andrew, Eduardo Borensztein, RatnaSahay, and Jeromin Zettelmeyer. 1999. “TheEvolution of Output in Transition Economies:Explaining the Differences.” IMF Working Pa-per No. WP/99/73, International MonetaryFund, Washington, D.C.

An extensive exploration of the issues in modelspecification with annual data for 26 countriesfrom 1991 to 1996 (different period for a fewcountries). A very general initial model includes:

• Macroeconomic variables (fiscal balance, in-flation, and exchange rate regime)

• Structural reforms (Liberalization Index in itsthree separate components, plus interactionsterm defined by multiplying the Liberaliza-tion Index by the share of the private sectorin the economy)

• Initial conditions (from de Melo and others2001, initial GDP per capita and growth, ur-banization, natural resource endowment,index of initial repressed inflation or actualinitial fiscal imbalance and inflation, shareof agriculture, trade dependence, a measureof overindustrialization, and reforms beforecollapse of central planning)

• Other controls (average OECD growth, termsof trade, and dummies for war or conflict).

In addition to the large number of variables,they include first and second lag of macroeco-nomic variables and up to third lags of struc-tural reform indexes. The initial conditions arealso parameterized by time to allow for theirimpact to decline or dissipate after a period (butthe precise specification and statistical tests arenot published with the paper). Both the level ofGDP (in logs) and the annual rate of growth areused as endogenous variables.

The initial specification contains too manyvariables to get significant results about effectsof individual variables, but it can be used to testthe hypothesis that certain categories of variablesare irrelevant. Thus the model rejects stronglythe hypothesis that none of the macroeconomicpolicy variables matters and/or that none of thestructural variables matters for growth.

The initial broad specification is progres-sively simplified by eliminating variables withlow statistical significance to arrive at two final

specifications that include 13–15 policy vari-ables, including lags, plus the controls and ini-tial conditions (the precise number of para-meters estimated is not explicitly shown).

Some summary results: most of the variabil-ity in growth is associated with differences inpolicies rather than initial conditions (depend-ing on the time-declining impact of initial condi-tions); the difference between CIS and Centraland Eastern Europe can largely be explained bydifferences in policies.

Campos, Nauro F., and Fabrizio Coricelli. 2000.“Growth in Transition: What We Know, WhatWe Don’t, and What We Should.” Centre for Eco-nomic Policy Research, London, United Kingdom.

The authors use the contemporaneous Lib-eralization Index and its components on theirregressions; only one component is statisticallysignificant. Inflation is significant (with nega-tive effect on growth) and so is the presence ofan International Monetary Fund program. In-stitutional variables are significant and positive(rule of law and quality of bureaucracy). Initialconditions are measured as the principal com-ponents of a set of reasonable-sounding indica-tors (dependence on Council of Mutual Eco-nomic Assistance trade, repressed inflation, andoverindustrialization).

de Melo, Martha, Cevdet Denizer, and AlanGelb. 1996. “From Plan to Market: Patterns ofTransition.” World Bank, Washington, D.C.

de Melo, Martha, Cevdet Denizer, Alan Gelb,and Stoyan Tenev. 1997. “Circumstance andChoice: The Role of Initial Conditions and Poli-cies in Transition Economies.” World Bank,Washington, D.C.

These are two influential papers that intro-duced the measures of policy reform and initialconditions most widely used in the literature. Thefirst paper introduced the Liberalization Index,defined as a weighted average of policy reformsin three areas: internal markets, external mar-kets, and privatization and private sector entry.The paper’s empirical analysis emphasizes “pat-terns of transition” rather than models of policyresponse and statistical test of hypotheses.


18

The second paper focuses on the role of initialconditions, for which the authors calculated a setof 12 indicators widely used in subsequent work(initial income, urbanization, natural resourceendowment, location, pretransition reforms, ini-tial repressed inflation, overindustrialization withrespect to Chenery’s “norms”, shares of trade withCouncil of Mutual Economic Assistance, blackmarket rate for foreign exchange, new versus oldstates, and years under central planning). The largenumber of indicators is reduced to a more man-ageable set of two by the method of principal com-ponents. Thus the transformed variables are de-fined as linear combinations of the initial variablesin a way that preserves the maximum of theirvariability. This is useful for statistical power, butmakes it more difficult to determine what initialconditions really matter.

The authors estimate a model where growth(in 1992–95) is explained by initial conditions,policy reforms (measured by the aggregate Lib-eralization Index or its cumulative value), anda war dummy variable. The authors also esti-mate a simultaneous equations model forgrowth and the Liberalization Index (with lib-eralization policies determined by initial con-ditions plus a war dummy and an index of po-litical freedom), but find that there is negligiblesimultaneity bias on the growth equation. Bothinitial conditions and the Liberalization Indexare quite significant in the growth equation. Thepaper also presents the “growth patterns”model discussed below, augmented by the twovariables for the initial conditions.

de Melo, Martha, and Alan Gelb. 1996. “A Com-parative Analysis of Twenty Transition Econo-mies in Europe and Asia.” Post-Soviet Geogra-phy and Economics 37(5): 265–85.

The authors discuss “patterns of transition”for the 26 European and Central Asian coun-tries plus China and Vietnam. The exogenousvariables are the Cumulative Liberalization In-dex (CLI), plus dummy variables for regionaltension and for Central and Eastern Europe.However, the paper postulates a model in whichthere is discontinuity when CLI = 0.4. This isso because the authors define a “nonreform”

and a “reform” pattern of growth, and coun-tries switch from the former to the latter whenthe CLI reaches 0.4 (that is, changes in CLI atlower or higher levels do not have any effect;only those getting CLI over the threshold havean impact). On the other hand, the two “pat-terns” are estimated with high degree of gener-ality (with up to 13 parameters). The two re-sulting patterns are plausible (for example,“reform” has an initial drop in output but it isincreasingly positive after the second year), andthe hypothesis that they are equal (that is, thatpolicies do not matter) can be strongly rejected.

Fischer, Stanley, Ratna Sahay, and Carlos A.Veight. 1998. “How Far is Eastern Europe fromBrussels?” IMF Working Paper No. WP/98/53,International Monetary Fund, Washington, D.C.

The authors find that the Liberalization In-dex and stabilization variables (inflation and fis-cal deficit) are significantly associated withgrowth. They introduced “reform time” to ad-just for different starting points of the transitionprocess in different countries.

Havrylysyhyn, Oleh, and Ron van Rooden. 1999.“Institutions Matter in Transition, but so Do Poli-cies.” Paper presented at the Fifth DubrovnikConference on Transition, Dubrovnic, June 1999.

The emphasis is on variables defining insti-tutional development: “Index of Economic Free-dom” from the Heritage Foundation (1994–97);indicators for democracy, rule of law, and oth-ers from the “Nations in Transit” reports fromFreedom House; institutional environment fromthe survey for the World Bank’s World Develop-ment Report 1998/99: Knowledge for Develop-ment (1998); and country risk ratings from“Euromoney.” They follow de Melo and others(1997) on the treatment of initial conditions (thesame variables are aggregated into the same twoprincipal components). A similar aggregation ap-proach is used to summarize the eight institu-tional variables into one or two principal com-ponents. One innovation is to include a simpleform of time dependence for initial conditions(impact at t is b

IC /t) and instrumental variables

(impact on t is t*bIV

).


19

Hernandez-Cata, Ernesto, 1997. “Growth andLiberalization during the Transition from Planto Market.” IMF Staff Papers 44(4): 405–29.

The paper models the reallocation of capitalfrom the old to new sector (as in Berg and others1999). The paper confirms the significance of lib-eralization and stabilization, with similar impactson the CIS and Central and Eastern Europe.

Heybey, Berta, and Peter Murrell. 1999. “TheRelationship between Economic Growth and theSpeed of Liberalization during Transition.” Jour-nal of Policy Reform 3(2): 121–37.

The authors find that average growth over thefirst four years of transition does not depend onthe increase in the Liberalization Index over theperiod, after accounting for possible endogeneityof liberalization policies (the instruments used arethe initial level of liberalization, share of indus-try, and an index of political freedom). The au-thors conclude that initial conditions are “muchmore important than policy variables.”

Popov, Vladimir. 1998. “Will Russia Achieve FastEconomic Growth?” Communist Economies andEconomic Transformation 10(4): 421–35.

Popov, Vladimir. 1999. “Shock Therapy versusGradualism: The End of the Debate.” CarletonUniversity, Ottawa.

Castanheira, Micael, and Vladimir Popov. 1999.“Framework Paper on the Political Economy ofGrowth in Russia and Central and Eastern Eu-ropean Countries.” Global Development Net-work, World Bank, Washington, D.C. Processed.

The authors use only the cross-section vari-ability in the data as they take average rate ofgrowth (over the whole sample or separatesubperiods) as a dependent variable. The Liber-alization Index is found significant for growthrecovery (1994–98), but not for the overall pe-riod (1989–98) or the subperiods (1989–96 and1996–98). Other control variables: war dummyand average inflation. Initial conditions are ac-counted for by including an Index of Initial Dis-tortions (defense expenditures/GDP –0.03 + de-viations in industrial structure and trade

openness from normal level + share of trade withFSU + share of trade with Central and EasternEurope/3), the initial level of GDP per capita,the decline in government revenues/GDP, andshadow economy/GDP.

Selowsky, Marcelo, and Ricardo Martin. 1998.“Policy Performance and Output Growth in theTransition Economies.” American EconomicReview—Papers and Proceedings 87(2): 350–53.

The paper uses combined cross-section andtime series data to test a model, allowing for adelayed impact of reforms on growth, with crudecontrol of initial conditions (Central and EasternEurope versus the CIS) and other factors (dummyvariables for period of war or internal conflict).The Liberalization Index is found to be highly sig-nificant and to exhibit significant differences be-tween its initial and long-term impact. There arealso differences in the dynamic impact of reformin Central and Eastern Europe and the CIS.

Annex 2.2. Additional Empirical Analysis

As mentioned in the text, the authors undertook additional empirical analysis in prepa-

ration for this paper to elucidate the types of ini-tial conditions that made a difference for coun-try performance, as well as possible differencesin the role of initial conditions and policies atdifferent stages of the transition.

The analysis of initial conditions was basedon the indicators developed by de Melo and oth-ers (1997), plus the terms of trade loss estimatedby Tarr (1994). The 13 indicators were dividedin three groups and then aggregated into three“synthetic” indices—initial structure, initial dis-tortions, and initial institutions. Table A2.1shows that the three indices “explain” (as indi-cated by the R-squared of the regressions) 51percent of the variance on average growth acrosscountries during the initial transition recession,and 44 percent during 1995–99.

Table A2.2 presents the results of the regres-sion of the annual rate of growth for all countrieson the Liberalization Index and initial conditions,controlling for conflicts (war dummy), allowingfor a dynamic impact of policies—as in Selowsky


20

TABLE A2.1.

Regression of Average Growth on Initial Conditions,1990–99

Average growth, Average growth,Initial conditions 1990–94 1995–99

Structure –0.30 0.18Distortions 1.03a 0.24Institutions 0.25 0.82a

R-squared 0.51 0.44

a. Significant at 95 percent.Source: Martin (2000).

and Martin (1998)—and for different effects dur-ing the early and late 1990s. The table shows thatthere are differences in the impact of policies andinitial conditions in the two subperiods: liberal-ization policies have a stronger positive impactduring 1995–99, while initial conditions have astronger impact on the earlier period. In both casesthe difference is statistically significant, as thehypothesis of equal effect in both periods is re-jected at a high level of confidence.

Notes

1. The three aggregate indicators (structure, distor-tions, and institutions) are a linear combination ofthe individual components, with weights equal to theirrespective coefficients in a regression on averagegrowth.

2. The Council of Mutual Economic Assistance isthe free-trade area encompassing the Soviet Union andEastern Europe.

3. Easterly (2000) shows that growth in most devel-oping countries slowed in the past two decades. Me-dian per capita income growth was 0 percent, com-pared with 2.5 percent in 1960–79. The slowdown isattributed to the deterioration in the external envi-ronment, specifically lower growth in OECD coun-tries, particularly Europe and Japan, and higher realinterest rates.

4. These indicators, developed by de Melo,Denizer, and Gelb (1996), stop in 1997. More re-cent work uses transition indicators of the Euro-pean Bank for Reconstruction and Development,which include 1998 and 1999.

5. The usual caveats about attributing causality toregression coefficients apply. However, the coefficientsstill provide the best inference available from the com-bined experience of all transition economies.

6. For example, these two factors would allow“predicting” the rate of GDP growth only with anuncertainty band roughly 11 points wide. More tech-nically, the estimated standard error of the regres-sion with the percentage rate of growth as depen-dent variable is about 6.89, so that a two-way 90percent confidence band would have a width of11.37 points (1.65 times the standard error). This isan extraordinarily large number compared withvariations in growth in the OECD countries.

TABLE A2.2.

Regression of Annual Output Growth on Policies andInitial Conditions Allowing for Differential Effects Earlyin Transition, 1990–99

Endogenous variable:Annual rate of growth

1990–99

Exogenous variable Coefficient t-value

Liberalization (t) –11.01a –4.08Liberalization (t – 1) 8.92a 2.60Liberalization (t – 2) 3.32 1.60Liberalization (t), 1995–99b 8.26a 2.02Liberalization (t – 1), 1995–99b –4.03 –0.71Liberalization (t – 2), 1995–99b –2.18 –0.60War –11.52a –5.45Initial conditions 0.09 0.38Initial conditions, 1990–94c 1.82a 4.50R-squared (number of observations) 0.60 (200)Standard error of estimate 6.89

a. Significant at 95 percent.b. The coefficient for these variables measured the additional ef-

fect of liberalization during 1995–99.c. The coefficient for this variable measured the additional effect

of initial conditions during 1990–94.Source: Martin (2000).

Part 2

Policy and InstitutionalChallenges Ahead

23

3The Quest for Growth

Countries started the transition facing common challenges. The production system was de-signed for the exigencies of a command economy. Much industrialization was based on cheapenergy and subsidized transport. In addition, the coordination and monitoring of central

planning meant few links among enterprises, with large state enterprises forming production anddelivery chains of vertically integrated, bilateral monopolies or monopsonies under the aegis ofbranch ministries.

Opening to the world revealed productivity differences across sectors and enterprises, given theenergy-intensive structure of production. In April 1992 the Russian Federation’s domestic oil pricewas still only 3 percent of the world price. Shutting down any links in those extensive chains ofproduction 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 ofliberalization, two further policy responses were required. The first was to impose market disciplineon inherited enterprises—so that they would face the incentive to restructure and, in so doing, becomemore productive and able to compete at the new prices. Failure to do so would lead to closure. Thesecond was to encourage the creation of new enterprises, which became possible after liberalizationcreated the legal opportunity for private investment.

It is reasonable to assume that investment in new enterprises would be undertaken only with anexpected rate of return at least equal to what could have been realized by investments in an enterpriseundergoing restructuring. Since restructured enterprises are, by definition, more productive than oldenterprises, this yields a productivity ranking (figure 3.1). The yardstick used here and throughout thereport is to rank enterprises by labor productivity, not total factor productivity. (The circumstancesunder which this is broadly permissible are explored in annex 4.2.) Empirical evidence collected froma 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).1


24

Aggregate growth in the economy reflectsthe interplay between old enterprises in need ofstate support that, by absorbing more resourcesthan they produce, reduce growth, and restruc-tured and new enterprises, which increase it (see

FIGURE 3.1.

Productivity Distribution of Old, Restructured, and New Enterprises

Note: The figure allows for outliers in both directions, as there is no reason why single old enterprises, everything elsebeing equal, might not occasionally produce higher value added per employee than new enterprises, or why new enterprisesmight not occasionally have disappointing results.


Productivity of restructured enterprises

Productivity of new enterprises

Range ofnew enterprises

Productivity of old enterprises

Range ofold enterprises0

Range ofrestructured enterprises0

0

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 transition economies inthe early summer of 1999, adding five more transition economies later that year. The survey used face-to-face interviews with enterprise owners orsenior 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 andaffected by the quality of the business environment and the relationship between enterprises and the state. The survey posed detailed questionsabout the enterprises’ business and competitive environment and about the different restructuring actions they had taken in the recent past. (SeeHellman 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 andlarge enterprises. Box table A provides some basic information on the distribution by region, size, origin, sector, and location of the enterprisesample. The survey included some 125 enterprises from each of the 25 countries, with larger samples in Poland and Ukraine (almost 250enterprises) and in Russia (more than 550 enterprises). The full sample comprises 3,954 enterprises, more than half from the Central andEastern European region (including the Baltics) and the rest from the countries of the CIS. The sample is dominated by small and medium-sizeenterprises (SMEs); half of them employed fewer than 50 people, and only 8 percent employed more than 500. More than half the enterpriseswere newly established private enterprises, 27 percent were privatized, and the remaining 16 percent were state owned. The enterprises aredivided 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)

Ruehl and Vinogradov 2001). The pattern ofgrowth is initially dominated by the negativecontribution of old enterprises, which causesoutput to decline. Conditions at the beginningof transition—such as repressed inflation,

The Quest for Growth

25

multiple exchange rates, and the terms of tradelosses associated with trade liberalization andthe breakup of the Council of Mutual EconomicAssistance trading area, which would in turnhave been reflected in the inherited capitalstock—were important determinants of this de-cline. With time, restructured and new enter-prises acquire the critical mass needed to over-come the negative effects of old enterprises togenerate economywide growth. The speed of

BOX 3.1 CONTINUED

TABLE A.

Characteristics of the Business Environment and Enterprise Performance Survey Sample, 1999

Number of enterprises Percent

Region CIS 1,866 47.2Central and Eastern Europe and the Baltics 2,088 52.8

Size Small (fewer than 50 workers) 1,944 49.2Medium (50–500 workers) 1,690 42.8Large (more than 500 workers) 318 8.0

Origin New 2,176 56.5Privatized 1,050 27.2State-owned 627 16.3

Sector Industry Farming 453 11.5Mining 33 0.8Manufacture 1,191 30.1Construction 343 8.7Power generation 16 0.4Total 2,036 51.5

Services Trading 541 13.7Retail 571 14.5Transport 232 5.9Finance 67 1.7Personal services 214 5.4Business services 245 6.2Communications 15 0.4Other 30 0.8Total 1,915 48.5

Location Capital city 1,220 30.9Large city 704 17.8Town 1,694 42.8Rural 336 8.5

Source: Hellman and others (2000).

Given the sample size and specific quotas, the survey cannot be used to measure the number and share of new enterprises in each country.However, it can provide valuable evidence about how the performance and behavior of new enterprises differ from state-owned and privatizedenterprises. It also provides an opportunity to test whether perceptions of the business environment differ systematically across differentcategories of enterprises. (For a full analysis of the data on small enterprises surveyed, see EBRD 1999).

this depends on policy choices. Differences inconditions at the end of communism—and inexogenous shocks and policy choices in the1990s—put countries in vastly different circum-stances today.

For many countries in the CIS—and forthose in Southeastern Europe, such as Bulgaria,FYR Macedonia, Romania, and the Federal Re-public of Yugoslavia, which have seen steep de-clines in incomes since the onset of transition—


26

restoring sustained growth and rebuilding thestate are key priorities.

Without growth it will be impossible to gen-erate income-earning opportunities for house-holds or to generate the resources to provide suchbasic public goods as a legal and judicial system,secure property rights, and basic infrastructure.Nor will it be possible to maintain investmentsin education and health that have been adverselyaffected since the onset of transition, or to putin place a social safety net targeted toward themost vulnerable. Without a functioning statethere will be no capacity to provide public goodseven if the resources were available.

Growth is also important for the more re-form-oriented countries in Central Europe andthe Baltics. Per capita incomes in the threewealthiest countries seeking accession to theEuropean Union as a fraction of the EuropeanUnion average are still only 68 percent forSlovenia, 59 percent for the Czech Republic,and 49 percent for Hungary. However, an ex-clusive focus on growth, while providing basic

public goods and protecting the most vulner-able, is not enough for these countries. Theadvanced reformers in Central Europe and theBaltics need to consolidate the gains of the firstdecade of transition and address what could becalled “second-generation” issues in the transi-tion toward an effectively functioning marketeconomy. They have to secure control overquasi-fiscal and contingent liabilities. They haveto undertake reform in labor and financial mar-kets to allow the benefits of growth to be morewidely shared across the population. They alsohave to restructure social sector expendituresto make them fiscally more affordable withoutimpairing the social safety net. Many of thesereforms overlap with those required to join theEuropean Union.

A Tale of Two Approaches

Looking at the assets of old, restructured, andnew enterprises reveals that countries restor-

ing sustained growth need to create a policy en-vironment that simultaneously disciplines old en-terprises (state enterprises, privatized butunrestructured enterprises, and agricultural col-lectives) and encourages “new” enterprises (in-cluding both greenfield investments and restruc-tured spinoffs from newly privatized enterprises).

Discipline entails hardening budget con-straints (a concept owed to Kornai 1986), intro-ducing competition to product markets, moni-toring managerial behavior to generate incentivesfor efficient resource use and prevent such abuseas asset stripping and tunneling (box 3.2), andproviding viable exit mechanisms for inefficiententerprises. It thus forces old enterprises to re-lease assets and labor, which then become avail-able for more efficient reallocation to restruc-tured and new enterprises. Old enterprises alsodivest themselves of social assets—such as hous-ing, health clinics, and kindergartens—which re-quires shifting the locus of social protection fromenterprises to local governments.

Encouragement entails reducing excessivemarginal tax rates, simplifying regulatory pro-cedures, establishing secure property rights, andproviding basic infrastructure. That allows re-structured and new enterprises to absorb labor

FIGURE 3.2.

Performance of Old and New Enterprises, 1996–99(percentage change)

Source: Hellman, Jones, and Kaufmann (2000).

–5

0

5

10

15Sales

Investment

EmploymentExports

Debt

New enterprises Old enterprises


27

BOX 3.2.

The Problem of Tunneling

Tunneling is the legal expropriation of income and assets belonging to minority shareholders (Johnson, La Porta, Lopez-de-Silanes, and Shleifer2000). Diverting cash flow and asset stripping are forms of tunneling. Managers usually do the tunneling and in purely private enterprises areusually acting at the behest of large or controlling shareholders. In partly state-owned enterprises, managers are usually acting by themselves orat the behest of private shareholders.

Tunneling should be distinguished from theft, which is illegal. However, if expropriation by management occurs in a fully state-ownedenterprise, this is almost always illegal. Tunneling should also be distinguished from rent seeking. Rent seeking refers to the efforts of enterprisesto 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 thestate, 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 tunnelingwere 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 Tunneling Legal?

Tunneling is legal for two reasons. First, there may be important loopholes in the legal protection of investor rights. Laws governing “relatedparty” transactions are often vague or inadequate. Consider the case where company A is controlled by the managers of company B and the twocompanies enter into a transaction. If company A supplies inputs at above-market prices (or sells 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 thatrelated party transactions require full disclosure and supervision by independent parties (as in the United States), tunneling through transferpricing is easy and legal.

Second, the courts may be unwilling or unable to apply even precise statutes. There are cases in Western Europe where courts have lookedat instances of tunneling and pronounced it legal. For example, in civil law countries, if the courts cannot determine whether someone washarmed, they will often not punish an action that clearly 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 lawas 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 toconsistently transfer small amounts of shareholder value out of companies. Such “seepage” of shareholder value is often considered an ordinarycost of doing business by investors in countries with a weak rule of law. But the evidence indicates that in environments with poor protection ofshareholder rights corporate dividends are smaller, the valuation of companies is lower, and financial development is slower than in countrieswith stronger protection of shareholder rights (Johnson, La Porta, Lopez-de-Silanes, and Shleifer 2000). There is also some evidence that suchseepage reduces the growth of and limits investment in capital-intensive sectors.

In periods of economic crisis in a context of weak investor protection, tunneling can increase substantially. If managers foresee their enterprise’sdemise, they have strong incentives to tunnel enterprise assets. The Russian economic crisis of 1998 reportedly triggered widespread and extremetunneling (Johnson, Boone, Breach, and Friedman 2000). The costs of such tunneling can be enormous. If investors expect that tunneling hasbecome 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.

Tunneling and Capital Flight

Theoretically, tunneling can remain purely domestic; that is, wealth can be transferred from minority shareholders to managers while remaininginside the country. Empirically, however, evidence suggests that episodes of large-scale tunneling tend to coincide with high levels of capitalflight (Johnson, Boone, Breach, and Friedman 2000). There are several reasons for this correlation. The risk that existing laws permittingtunneling 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



28

domestic banks that suffer from the same weaknesses of the broader environment. If managers can tunnel assets freely from their ownenterprises, 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 enterprises during temporary periodsof poor performance. In such cases tunneling becomes a form of cross-subsidization within a related group of companies. In environments whereenterprises might have limited access to short-term credit because, for example, the banking system is weak, “propping” temporarily troubledenterprises through transfers from related companies could ease access to capital in the future. So, despite the negative consequences oftunneling, the function of propping as a backstop in insufficiently developed financial markets could explain why investors are still willing toinvest in such an environment (Shleifer and Wolfenson 2000).

Increasing Vulnerability to Economic Crisis

Institutions protecting entrepreneurs against government expropriation are crucial to transition and economic development generally. A lack ofprotection 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 notcapital intensive. Nevertheless, weak investor protection 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.

BOX 3.2 CONTINUED

and assets made inexpensive by the downsizing.By creating a stable and predictable business en-vironment, the policies of encouragement gener-ate incentives for enterprises to invest. As moreand more new enterprises enter the market, anincreasingly competitive environment develops.At the same time the policy environment mustrestrain predatory behavior by new enterprisesthat seek to extract special preferences from thestate and, in so doing, erect barriers to competi-tion and further entry.

The more advanced reformers in the CSB,now facing second generation issues, are muchfarther along in implementing an environmentof discipline and encouragement. Discipline hasbeen established, but it needs (as in Poland) tobe maintained, because loss-making state en-terprises in coal mining, steel, and railways stillimpose a costly burden on the budget. In thesecountries new and restructured enterprises stilllead growth, but unemployment remains stub-bornly high. In addition to maintaining the dis-cipline established during the first decade of re-form, the focus of encouragement for jobcreation in new enterprises needs to be remov-ing bottlenecks in infrastructure and reforminglabor and financial markets and the social pro-tection system.

Discipline and Encouragement

As the case of Poland suggests, an important in-gredient of market discipline is the hard budgetconstraint on state enterprises, together with suf-ficient standards of corporate governance to pre-vent large-scale asset stripping before privatization.It was understood that no subsidies would go tostate enterprises after the beginning of 1990, butit took 18–24 months for the government’s com-mitment to hard budget constraints to be seen asfully credible by enterprise managers. Poland’seconomic growth—which resumed in 1992, theearliest among the transition economies—was firstdue largely to better use of existing assets by en-terprises spun off from state enterprises. Not until1995 was there a big boom in domestic invest-ment—and not until 1996, with Poland in its fifthyear of growth, was there a take off in foreigndirect investment.

At the same time successive governments un-dertook structural reforms to generate an invest-ment climate favorable for the entry of new en-terprises, in particular small and medium-sizeenterprises (SMEs). Like many other countrieswith a socialist past, Poland had a high indus-trial concentration, with the leading enterprisehaving more than 30 percent in more than 60


29

percent of the markets (at a three-digit level).The new enterprises signaled that product mar-ket competition would press state enterprises tobecome efficient. During 1990–98 the numberof individual- and family-owned enterprises rosefrom 1.2 million to nearly 2.8 million. Similarfigures held for Hungary. In both countries theshare of employment and value added of newenterprises, the engines of growth, rose to 50percent or more.

Protection and Discouragement

The disposition of assets among old, restructured,and new enterprises also provides a useful per-spective on such countries as Romania and Rus-sia. They have protected rather than disciplinedold enterprises through subsidies—grantedthrough the budget, energy consumption, andthe banking sector. In addition, their institutionsof public and corporate governance are notstrong enough to prevent asset stripping. Theydiscourage or at best only selectively encouragethe entry of new enterprises because the state’scapacity to provide key public goods is weak andthe investment climate poor. Tax rates are high.Licensing and registration procedures are opento abuse. Furthermore, the legal and judicial sys-tem is unable to enforce property rights.

Adding to this discouragement, subnationalgovernments engage in anticompetitive practicesto protect established enterprises at the expenseof new enterprises in their jurisdictions. Examplesof protection abound. Transfers to selected enter-prises and conglomerates in Russiain 1992 through credit at highly subsidized ratesare estimated at 33 percent of GDP, financedthrough a massive inflation tax on households andenterprises without political connections. Subsi-dies implicit in soft budgets amounted to another5 percent of GDP in 1996 and 3.5 percent in 1997.

Romania reverted in 1994 to directed credit,price controls, and budgetary and extrabudgetarytransfers—and reopened enterprises earlier de-clared closed. These attempts to use the economy’sold capital stock precipitated a macroeconomiccrisis in 1996. Again, such transfers were financedthrough implicit or explicit taxes on households,new enterprises, and enterprises that restructuredenough to survive the market test.

The result was to prevent or postpone closureof the least productive old enterprises and the re-structuring of enterprises with good prospects. Inaddition, the entry of new enterprises that wouldbe viable without state support was restrained.The protect-and-discourage strategy thus createsan environment where resource transfers flow ina direction opposite to that in a discipline-and-encourage environment.

The Associated Fiscal Adjustment…

The two contrasting approaches of disciplineand encouragement and protection and

discouragement are also broadly mirrored by theassociated fiscal adjustment that has taken place.Enterprises used to be a captive source of revenuein transition economies. The state’s loss of con-trol over them also meant the loss of fiscal con-trol and the need for political acceptance of a re-duced public sector. Government revenues as ashare of GDP fell from around 38 percent in 1992to 31 percent in 1998 in the CIS and from 44percent to 39 percent in the CSB. Stabilizing in-flation, in practice often the first order of busi-ness, required that expenditures be reduced too.They fell precipitously from 57 percent of a rap-idly declining GDP in 1992 to 37 percent in 1998in the CIS—and, less sharply, from 45 percent in1990 of a more modestly reduced GDP to 41 per-cent in 1998 in the CSB. This situation gave riseto two kinds of adjustment.

First, a substantial amount of spending wasmoved out of the budgetary arena. Explicit bud-getary subsidies to enterprises generally fell acrossmost of the region, but enterprises were still sup-ported in a variety of ways. Implicit subsidieswere channeled largely through the energy sec-tor, which would then pass the costs back in theform of arrears to the budget. In addition, bankloans were rolled over and the enforcement oftax and other rules was lax. In Russia, for ex-ample, explicit budget subsidies to the enterprisesector declined from 10.2 percent of GDP in 1994to 5.9 percent in 1998. But total budget subsi-dies to the enterprise sector, which also includethe net increase in tax arrears as well as inflatedprices of goods procured by the government andpaid for by offsetting tax arrears, rose as a sharefrom 10.9 percent of GDP to 16.3 percent.


30

Moreover, implicit subsidies extended throughbank loans and public utilities were siphoned offalong the way, coexisting with wage payment ar-rears by enterprises, thus not helping workers whowould have benefited if soft budget constraintshad indeed been an instrument of social protec-tion. Faced with a shrinking tax base and moreand more claimants being assisted outside thebudget, governments attempted to collect morerevenue by raising taxes. This discouraged poten-tial entrants and, with abuses of discretion by gov-ernment officials, drove them underground.2 In ad-dition, this pattern of fiscal adjustment, bysignaling business as usual, helped protect enter-prise managers from restructuring or closure.

A second mode of fiscal adjustment reallocatedpublic expenditures to the social sectors to cush-ion the impact of transition on the vulnerable.That made it politically possible to discipline oldenterprises into shedding labor. In Poland, forexample, pensions were critical in preventing theelderly from falling into poverty. Indeed, high so-cial spending was more affordable in the CSBbecause government revenue was significantlyhigher than in the CIS countries. That also helpedto create a constituency for the discipline neces-sary to ensure a return to growth.

However, this alone would have been un-sustainable without the rapid growth of newenterprises. Their high labor productivity gavethem the potential to offer displaced workers aviable outside option rather than a return tosubsistence activities, and that created a con-stituency for encouragement. The CSB still re-sorted to off-budget activity and created con-tingent liabilities. But these countries show theoverarching role of fiscal policy and the insti-tutions of budget management in supporting agrowth-oriented adjustment by redistributingpart of the “reform dividend” to those whowould otherwise bear its costs.

…And the Role of Labor Markets

As with fiscal policy, two broad patterns oflabor market adjustment can be identified.

The first, largely associated with the CIS and thecountries of Southeastern Europe, involved a

decline in employment significantly smaller thanthe massive collapse of output and labor demand.The adjustment took the form of lower realwages as well as the emergence of arrears andnonpayment of wages. Without vibrant new en-terprises, labor moved to low-productivity ser-vices and subsistence agriculture. Together withlabor hoarding by enterprises, these sectorsserved as shock absorbers in view of the lack ofa functioning social policy.

The second pattern, broadly prevalent in theCSB, saw employment decline with output. Jobdestruction was concentrated in existing enter-prises, while job creation was to be found almostexclusively in new enterprises. Here too there iscross-country variation. For Poland, which exem-plifies the discipline-and-encourage approach totransition, enterprise restructuring in key sectorsof the economy and concomitant increases in la-bor productivity meant that output, growing rap-idly since 1992, has outpaced job creation to apoint where unemployment stood at 17 percentin August 2001. Budget constraints were softerand enterprise restructuring more limited in theSlovak Republic, which grew at an annual aver-age rate of 4.7 percent between 1994 and 2000.The stop-and-go nature of the reform effort re-sulted in an insufficient creation of jobs in newfirms. Employment remained largely unchangedand unemployment reached nearly 19 percent inthe second quarter of 2000—the highest rate ofunemployment in Central Europe.

In both countries, the aggressive use of socialsector expenditures in the form of generous so-cial assistance and unemployment insurance pro-grams cushioned the risk of poverty in the face ofhigh unemployment. But as in many countries inCentral Europe and the Baltics, these programsdiminished the incentives for workers to look forjobs. This, together with high payroll taxes, hasconstrained the growth of employment.

Notes

1. “Old” enterprises are defined as enterprises es-tablished before 1989 in Central Europe and theBaltics and before 1991 in the CIS. Increments ingrowth rates are 5 percent, with –5 percent as the


31

origin and zero the first increment, to capture thenegative export growth rate reported by the old en-terprises in the sample for 1996–98, a time whenoverall growth across the region had picked up.

2. Estimates of the unofficial economy as a share ofGDP in 1995 are 42 percent in Russia, 49 percent inUkraine, and more than 60 percent in Azerbaijan andGeorgia (Johnson, Kaufmann, and Shleifer 1997).

33

What policies and institutions are required to bring about a discipline-and-encouragementenvironment? While no one policy can be assigned to a single outcome in an interrelatedsystem, 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 andbanks, 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 privateand 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 goalof 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, andUzbekistan, 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 heldover from the Soviet-style command economy.

Encouragement starts with liberalizing prices and trade to enable the entry of new enterprises. Butit 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 toinvest. These can be roughly divided into two categories: the quality of public goods and the extent ofnonmarket obstacles to competition.

The key public goods that most directly affect the quality of the investment climate are a legal andjudicial 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 overtime, 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.

4Discipline and Encouragement


34

Tearing down the obstacles that discourageinvestment and competition is just as important.The inheritance of the command economy in-cluded stifling administrative barriers to entry,over-regulated labor markets, and tremendousdiscretion for bureaucrats. New barriers haveemerged in the transition. The most prominentof these has been an overly complex anddistortionary tax regime that pushes new entre-preneurs into the informal economy. These ob-stacles create an environment in which corrup-tion and uncertainty undermine investment.

Juxtaposing discipline and encouragementwith protection and discouragement highlightstwo contrasting modes of adjustment. In reality,country outcomes span a range of intermediatepossibilities depending on whether liberalization,hard budget constraints, and an enabling busi-ness environment were pursued—and in whatorder and how vigorously. Though it is not pos-sible to categorize transition economies into aspectrum of different modes of adjustment, coun-try examples can be used to exemplify alterna-tive transition paths:

• The policies of discipline and encouragementwere pursued most consistently in Estonia,Hungary, and Poland.

• Within the broad category of discipline andencouragement, softer budget constraints—and hence less discipline—prevailed for a longtime in the Czech Republic, Lithuania, andthe Slovak Republic. Indeed, the resultinglack of industrial restructuring is widely heldto have precipitated the Czech crisis in 1996–98. Harder budget constraints and faster re-structuring can help reorient these economiestoward the path followed by the first group.

• Bulgaria, the Kyrgyz Republic, Moldova, Ro-mania, the Russian Federation, and Ukraineliberalized their economies, but for a long timefailed to maintain discipline through hardbudget constraints—and they could not con-tain tunneling and theft through either law oradministrative control. The competition forresources between old and new enterprisesmakes it difficult to provide encouragement ifthe discipline for old enterprises is relaxed.Russia and Ukraine encouraged new entry

early in the transition. However, the captureof the state by a narrow set of vested inter-ests—old enterprises and well-connected earlyentrants—discouraged further entry at laterstages of transition and created a poor invest-ment climate. This resulted in a pattern of pro-tection and selective encouragement.

• Belarus, Turkmenistan, and Uzbekistan,which neither liberalized nor hardened bud-get constraints, strongly discouraged newentry. Access to foreign exchange and crediton special terms softened the budget for stateenterprises. Reliance on mechanisms inher-ited from the command economy continued.The survival of a highly centralized structureof political power did limit the extent of as-set stripping and other forms of theft thatproved so damaging to growth in the previ-ous groups of countries. Yet this element ofdiscipline came at the cost of a highly pro-tective stance that discouraged entry of newenterprises and SMEs.

• For an example of partial liberalization andweak discipline of state enterprises (thoughwith strong restrictions on asset stripping)coupled with an environment strongly support-ive of new entry, one must look outside East-ern Europe and the CIS to China (box 4.1).

Why particular countries or groups of coun-tries get on paths of either discipline and encour-agement or protection and discouragement canbe traced to economic and political conditionsthey faced at the onset of transition; the struc-ture of political institutions that determined therelative power of winners and losers from re-forms; and initial choices regarding the pace,comprehensiveness, and sequencing of reforms.Understanding the forces that shape the environ-ment helps define the key challenge in the politi-cal economy of discipline and encouragement;fostering new coalitions of winners and losers iscrucial for shifting the incentives of old enter-prises and promoting the development of newenterprises and thus for improving the probabili-ties of reform.

Meeting the challenges of discipline and en-couragement is important if growth is to be re-stored, its quality enhanced, and its benefits


35

BOX 4.1.

Can the CIS Learn from China’s Reform Experience?

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 thisexperience offer lessons for the countries of the CIS—particularly for such countries as Uzbekistan and Belarus, which have yet to embarksubstantially on economic reform? Or are conditions so different between China and these countries that China’s experience does not offerlessons for these countries?

Since the onset of 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 sectorthat employed 71 percent of the labor force and started a virtuous cycle, not only in agricultural development but also in rural industry. Thesubstantial 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 productquality 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, ledChinese 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. Thisallowed the banking system to channel net flows of 5 percent of GDP to borrowers. Some of the flows covered state enterprise losses and helpedfinance high-productivity investment in new and nonstate enterprises. Seigniorage accruing to the government in the 1990s allowed the budgetdeficit and substantial needs of loss-making enterprises to be financed through money creation with low inflation, leaving financial workouts andloan 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 ofmarkets was to lead to a massive inflow of foreign direct investment. The explosive growth of enterprises operating under the banner of townshipsor villages in rural areas was a source of internal competition, while external sector liberalization in coastal areas was a source of externalcompetition 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 percentof the labor force was engaged in agriculture in 1990, compared with 71 percent in China. In principle, the energy sector in Russia, which gainedfrom partial price liberalization on the order of 11 percent of GDP, could have been used to compensate the losers from reform. However, thestate’s loss of control rights over the sector, which had already occurred during the years of reform socialism, implied that much of the gainswent abroad through capital flight, reflecting, among other things, a poor investment climate in the country. This is reflected in the difference inthe ratio of fixed investment to GDP, which was 22 percent in Russia compared with around 34 percent in China, a large part of which—perhaps10 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 breakupof the former Soviet Union and before price liberalization, had led to substantial involuntary savings, resulting in a “monetary overhang”—mostlyhousehold claims on the state banking system, estimated at roughly a third of household wealth for the former Soviet Union in 1990. Attemptsto sterilize this overhang were unsuccessful, and the savings were extinguished by the burst of inflation following price liberalization at the onsetof transition. Not surprising, financial savings were slow to recover from this episode that, together with currency confiscation, led to substitutionaway from the domestic currency to commodities and foreign exchange. In marked contrast to China, the ratio of M2 to GDP in Russia fell from68 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 theChinese labor force worked in the state sector (and hence were entitled to a range of social benefits), compared with 90 percent 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 forcewas employed in state and municipal enterprises and organizations, including state farms, and a further 6 percent worked in collective farmsand consumer cooperatives, leaving only 1 percent in individual and private enterprises. Because of the collapse of demand for industriesproducing capital goods and defense-related outputs and the difficulty of switching away from trade with the Council of Mutual EconomicAssistance to Western markets, a much larger share of the Russian economy was not viable following price and trade liberalization. The costs ofpropping up rather than restructuring the state sector would have been prohibitive. Thus, a strategy of protecting the state sector through fiscaland financial transfers from the rest of the economy, unlike in China, would not have allowed adequate economic space for new enterprises toemerge as sources of growth.



36

BOX 4.1 CONTINUED

Finally, the government’s ability to manage the process was critical. China contained the abuses from partial reform to some extent byexercising tight political control over asset stripping, arbitraging between controlled and market prices for private gain, and corruption. Incontrast, the exit from communism, particularly in the countries of the CIS, led to a collapse of state institutions and, in the absence of aframework 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 ofCentral Europe and the Baltics. These include, for example, the strong role played by new and nonstate enterprises in generating growth and thepositive role of increasing trade with market economies and foreign direct investment. Yet the large differences in initial conditions betweenChina 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 agriculture and nonstate industry). Part of these gains was available for transfer to the losers from thesecond transition from a command to a market economy. Equally important, those losers—those from the unviable sectors—did not account fora big part of the economy. The state’s capacity to manage public assets enabled a slower move to market conditions for loss-making stateenterprises, before they were subjected to full market discipline in a growing economy. If a country is able to follow such a phased strategy, thereis 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 enterprisesaccounted for a higher share of the economy and needed to be subjected to hard budget constraints for resources to be liberated and used bythe winners (new private enterprises). This led first to a cut in output and activity, then a recovery. In addition to these differences in political andeconomic conditions, central planning was far more entrenched in Russia than in China. In the 1970s central government agencies in the formerSoviet Union physically allocated about 60,000 different commodities throughout the plan. In China the number was about 600, unchangedfrom 1965. Nor, unlike the former Soviet Union, did China have the problems posed by “giantism” and enormous regional specializationunderpinned 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 bankingsystem, which was used to support loss-making state enterprises, is saddled with nonperforming loans of 30–40 percent of annual GDP.Restoring the health of the state banking system on account of those loans could lead the stock of domestic government debt to rise from about20 percent to 75 percent of GDP. Servicing this debt will pose a major fiscal challenge for the government. This mirrors the experience oftransition economies such as Bulgaria and Romania, where support for state enterprises resulted in a high proportion of nonperforming loansin the banking sector. Moreover, as in several transition economies, surveys report that asset stripping and excessive wage competition inChina’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, GDPas a proportion of its 1990 level was 85 percent in Belarus and 94 percent in Uzbekistan, significantly higher than the corresponding numbersin other CIS countries. The initial fall in GDP in Uzbekistan during the transitional recession was more like the shallow dip in Central and EasternEurope, which may be attributed to low rates of industrialization and urbanization that, together with self-sufficiency in energy, made thesecountries 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 enoughpolitical control over state enterprises to limit the risk of spontaneous privatization and excessive asset stripping. Such control has not alwaysbeen effective; thus, Uzbekistan has seen sizable capital flight amounting to nearly 20 percent of merchandise exports. Second, governmentswere able to channel an infusion of resources to priority state enterprises and sectors, such as large-scale chemicals and automobiles inUzbekistan 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 thatturned the country into a net oil exporter in 1996. The effect of policies that discriminated against agriculture was a transfer of an estimated 4–5percent of GDP out of the sector during 1996–98. In addition, state banks, used to transfer resources to favored sectors and enterprises, facethe prospect of large losses. In Belarus the infusion of resources came from goods-for-energy barter deals with Russia on terms that were highlyfavorable, 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 investmentclimate conducive to entry of new enterprises. In Belarus, for example, the share of employment accounted for by small enterprises (employing50 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’sexperience and strongly encourage the growth of new enterprises as a basis for wealth creation and economic growth.


37

widely shared. The challenges are also interre-lated. Entrepreneurs have no incentive to bringin new management, develop new products, orseek new markets needed for growth-orientedrestructuring if explicit subsidies or implicit sup-port extended through banks and public utili-ties and protection from competitive pressureskeep them afloat. Evidence from business sur-veys confirms that enterprises facing hard bud-get constraints and a degree of competition aremore likely to undergo managerial turnover anddevelop new products (Carlin and others 2001).

Industrial enterprises also face major ob-stacles to downsizing if arrangements to helpthem divest social assets, such as housing, sana-toriums, and kindergartens, are unavailable. Thisproblem is further compounded by the lack ofpublic resources to finance an adequate systemof social protection to replace these divested as-sets for well-targeted groups. Startups andspinoffs that could fuel innovation and growthdo not enter the fray if the tax and regulatorysystem does not level the playing field betweenthem and incumbent state-owned and privatizedbut unrestructured enterprises. If the state can-not administer the rule of law, particularly con-tract enforcement and secure property rights, rent

seeking and widespread theft of erstwhile stateassets dominate economic activity. The state can-not afford the social safety net required for re-structuring or investments in human capital if,together with enterprises, it is caught in a perva-sive web of arrears and unpaid taxes.

The reform agenda required to enable growthto resume calls for effective public and privateinstitutions. Political institutions are critical inmediating the interactions between winners andlosers from various reforms. The ability of the stateto secure property rights depends on the effec-tiveness of the legal system. Mechanisms of cor-porate governance dictate what kind of owner-ship structure is most likely to bring aboutefficiency-enhancing enterprise restructuring. Im-proved budget management is necessary to en-sure that the state maintains a responsible fiscalpolicy. Effective administration of programs ofsocial assistance and insurance are required so thatthe “reform dividend” generated by growth is usedto assist those adversely affected by transition.

But policy choices are critically important aswell; the example of East Germany demonstrateshow poor policy choices at the start of transi-tion can undermine even the most favorable en-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 transitionin East Germany has been unique among the former socialist countries. It appeared to start from a privileged position in at least three respects.

First, all economic, political, and social institutions and arrangements of West Germany, as well as its entire legal framework, were adoptedwithout change and without delay. East Germany did not face the difficult challenge of building a new institutional framework from scratch. Giventhat it is now conventional wisdom to argue that the key deficiency of the transition in many countries has been insufficient attention to buildingan 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 transitioneconomies, East Germany received huge transfer payments financed partially by a new “solidarity tax” leveled on West German incomes. Annualtransfers 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) for the 1991–97 period. Current transfers still average DM 140 billion, or more than 4 percent of West Germany’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 massiveretraining and public works programs.



38

BOX 4.2 CONTINUED

Third, unification gave East Germany automatic membership in the European Union (and the World Trade Organization). While the CISsuffered from European antidumping rules and Central Europe’s Central Europe Free Trade Agreement members complained of being blockedfrom the Common Market (while being subjected to cheap imports, especially in the agricultural sector), East Germany benefited from speedyand 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. EastGermany’s growth performance has lagged behind other transition economies with far fewer advantages (see box figure A). Its initial economicdecline was deeper than in its transition economy neighbors. Though East Germany’s growth rates were higher than those in other countries inthe region for a few years, its subsequent growth has been among the slowest in Europe. The hope that East Germany would quickly catch upwith West Germany has not materialized.

FIGURE A.

Comparison of Real GDP, 1989–2000

Note: The average for the CSB is weighted by population.


Index of real GDP (1988 = 100)150

125

100

75

501989 20001990 1991 1992 1993 1994 1995 1996 1997 1998 1999

Germany

East Germany

Poland

Average for Central andSoutheastern Europeand the Baltics

How can such poor performance be explained despite the enormous advantages that East Germany enjoyed in comparison with othertransition economies? A simple answer is that even a good institutional framework backed by massive subsidies cannot overcome the negativeconsequences of inappropriate policies. Two early policy choices had particularly damaging consequences. First, the conversion rate of West toEast German marks reflected political pressures and not economic realities, vastly overvaluing the East German currency. Second, the attemptto bring East German wages in line with West German wages, while average labor productivity in the East remained at approximately a third ofthe West, severely damaged East Germany’s competitiveness. As a result, a larger segment of the inherited capital stock was scrapped asunproductive 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 transferpayments. Such payments continue to constitute between 55 percent and 65 percent of gross national product. Clearly, the current level ofsocial 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 slowdownin the rate of new enterprise growth, as shown in box figure B. The net creation of new enterprises, which had outperformed West Germany in1991, 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—anemicgrowth of new enterprises that have been the main drivers of growth elsewhere in the transition. Indeed, investment data suggests that evenGerman investors are “jumping over” East Germany to invest in the Czech Republic, Hungary, and Poland, where unit wage costs areconsiderably lower.



39

New Enterprises Drive the Transition

The success of a discipline-and-encouragestrategy is predicated on the ability of new

and restructured enterprises to emerge as enginesof economic growth. Moving assets from thepublic to the private sector is thus an importantelement of transition. In 1999 the share of theprivate sector in CIS GDP was 20 percent inBelarus; 55 percent in Kazakhstan and Ukraine;60 percent in Armenia, Georgia, and the KyrgyzRepublic; and 70 percent in the Russian Federa-tion (figure 4.1). These figures compare favor-ably with Central and Eastern Europe: 55 per-cent in Macedonia and Slovenia, 60 percent inBulgaria and Romania, 65 percent in Latvia andPoland, and 80 percent in the Czech Republicand Hungary.

The picture is quite different, however, for newenterprises, which typically need encouragementin the form of a favorable business environment(figures 4.2 and 4.3). Using small enterprises em-ploying fewer than 50 workers as a proxy for newenterprises, their contribution to value added in

1998 was around 55–65 percent of GDP in theCzech Republic, Hungary, and Lithuania, com-pared with 10–20 percent in Belarus, Kazakhstan,Russia, and Ukraine.1 Data on small enterprisesas providers of employment divide countries intotwo groups: leading reformers and countries fur-ther behind. Small enterprises’ share of employ-ment in 1998 was about 50 percent for leadingreformers such as the Czech Republic, Hungary,Latvia, Lithuania, and Poland, roughly the sameas the European Union. For countries less far alongthe path to a market economy, such as Belarus,Kazakhstan, Russia, and Ukraine, the share wasbetween 10 and 20 percent.

The share of small enterprises in employmentand value added differs widely across the region(see figures 4.2 and 4.3). In many parts of theCIS and Southeastern Europe—where growthis low and income per capita is substantially be-low pretransition levels—the share of employ-ment in small enterprises hovers around 20 per-cent. In contrast, the share of small enterprises

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 casedemonstrates how poor policy choices at the start of transition can undermine even the most favorable environment.

FIGURE B.

Net New Enterprise Registration in East and West Germany

Source: Ruehl and Vinogradov (2001).

ln 1000300

250

150

100

50

020001991 1992 1993 1994 1995 1996 1997 1998 1999

West Germany

East Germany


40

in total employment and value added is morethan 50 percent in the CSB, characterized byhigh and sustained economic growth and eitherapproaching or surpassing pretransition percapita incomes.

The rapid growth of small enterprises in help-ing bring about a higher contribution to valueadded and employment among the leading re-formers shows that economic policy has directedthe process of factor reallocation. That promptsthree questions (table 4.1).

First, is labor productivity in small enter-prises, measured as value added per employee,higher than in large enterprises in transitioneconomies? In the aggregate, small enterprises

consistently account for a larger fraction of to-tal value added than of total employment. La-bor productivity is indeed higher in small enter-prises compared with large enterprises.Moreover, this is true independent of progresstoward a market economy. New companies aremore productive than inherited enterprises incountries as diverse as Hungary and Ukraine.

Second, is aggregate labor productivity,measured as aggregate value added per em-ployee, higher in an economy where small en-terprises account for a higher proportion ofvalue added and employment than in anotherwhere those shares are lower? Small enterpriseshave had high and growing shares of employ-ment and value added in the leading reformersin the CSB (figures 4.4 and 4.5). However, thereappears to be a threshold—of around 40 per-cent for the shares of small enterprises in em-ployment and value added—below whicheconomies do not take off in terms of growth.This echoes the finding that there must be aminimum critical mass of reforms, below whichthe economy does not respond to policies. Theshares of new enterprises in employment andvalue added are high and above the thresholdin the leading reformers in the CSB. But bothshares remain low and well below the thresh-old in the slow-growing economies of the CIS.The notion of a threshold is important in theinteraction between the old and new sectors ofthe economy.

Third, can differences in labor productivitybetween small and large enterprises constitute asource of growth? The higher productivity ofsmall enterprises needs to be complemented byan incentive for labor and capital to move to thatsector—for small enterprises to increase aggre-gate growth. Higher labor productivity in smallenterprises implies lower labor intensity per unitof output. This observation and the assumptionthat small enterprises are less capital-intensivethan large enterprises imply that labor and capi-tal have a higher marginal product in small en-terprises (annex 4.2). Movements from large tosmall enterprises are adding value and thus are asource of growth. The large gap in labor pro-ductivity in countries such as Kazakhstan andUkraine shows an unrealized potential forgrowth in the new sector.

FIGURE 4.1.

Private Sector Share in GDP, 1999


Hungary

Poland

Czech Republic

EstoniaSlovak Republic

Slovenia

Latvia

Lithuania

Croatia

Kyrgyz Republic

Kazakstan

Bulgaria

Macedonia, FYR

Georgia

Moldova

Romania

Armenia

AlbaniaRussian Federation

Ukraine

Azerbaijan

Uzbekistan

Tajikistan

BelarusTurkmenistan

100806040200

Percent


41

The difference between the two sets of enter-prises—old and new—begins to erode over timefor two reasons. Old enterprises either close orrestructure, raising labor productivity, and em-ployment growth in new enterprises at some pointreduces labor productivity in those enterprises.There is a marked productivity difference betweenthe two sets of enterprises for Russia and Ukraine,but a noticeable narrowing of that difference forthe Czech Republic, Hungary, and Lithuania far-ther along the path to a market economy.

A Small Number of High-Productivity SmallEnterprises Is Not Enough

As the transition proceeds, labor shed bydownsizing old enterprises either finds its

way into new and more productive employmentor migrates to unemployment or subsistence ac-tivities. Labor hoarding in the old sector maypersist for a long time, with new enterprises act-ing merely as passive receptacles for such trans-fers. Alternatively, when the investment climate

FIGURE 4.2.

Share of Employment in Small Enterprises, 1989–98

Source: World Bank database on SMEs.

Percent60

40

20

01989 1990 1991 1992 1993 1994 1995 1996 1997 1998

Belarus

Czech Republic

Hungary

Kazakhstan

Latvia

Lithuania

Poland

Russia

Ukraine

FIGURE 4.3.

Share of Value Added in Small Enterprises, 1989–98


Percent80

60

40

20

01989 1990 1991 1992 1993 1994 1995 1996 1997 1998

Ukraine

Russia

Poland

Lithuania

Latvia

Kazakhstan

Hungary

Georgia

Czech Republic


42

is conducive to entry, new enterprises competewith the old sector, rapidly increasing their sharein employment and typically attracting the mostqualified individuals.

As noted, a threshold of about 40 percentfor the shares of small enterprises in employmentand value added needs to be crossed for newenterprises to absorb the resources released by

the old sector and contribute to sustainablegrowth. Simply having a small number of highlyproductive small enterprises is not enough. Un-less it is combined with rapid growth in the shareof employment, the small sector will not developthe critical mass to lead aggregate economicgrowth (see figures 4.4 and 4.5).

Russia illustrates the inability of the new sec-tor to contribute to sustained growth. The shareof value added for small enterprises more thandoubled, from 10 percent to 23 percent between1991 and 1998, in part reflecting a steadily de-clining GDP. But their share in employment hasremained low and largely stagnant. They do notseem to have incentives to increase their employ-ment or to multiply—so they continue to fall wellshort of the threshold for growth to take off. Alarge chunk of the labor force remains mired inunrestructured state and private enterprises.

In countries where aggregate employmentpicked up, it did so after the recovery of aggre-gate output (figure 4.6). This empirical investi-gation of new enterprises and their interactionwith old enterprises suggests the following:

• A sharp and early decline in aggregate em-ployment precedes the rapid growth of new

TABLE 4.1.

SMEs Have Higher Labor Productivity, 1998(percent)

Share of total Value added Value addedCountries SMEs employment (total) (per employee)

Belarus 37.7 15.9 — —Czech Republic 97.0 48.7 53.5 109.9Georgia 88.6 39.6 39.3 99.2Hungary 96.1 54.9 63.6 115.8Kazakhstan 88.6 15.6 22.4 143.6Latvia 91.2 45.5 50.4 110.9Lithuania 97.4 55.1 55.3 100.4Poland 92.1 45.7 54.4 118.9Russia 56.3 18.6 23.0 123.7Ukraine 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.

FIGURE 4.4.

Value Added per Employee in Small Enterprises,1998

Source: Eurostat (1998).

Hungary

Czech Republic

European Union

Kazakstan

Georgia

RussiaUkraine

403020100

Thousand euros


43

Figure 4.5.

Index of GDP and Shares of Value Added and Employment Accounted for by Small Enterprises,1989–98


Czech Republic

Index Percent120

100

80

60

40

20

0

60

50

40

30

20

10

0

GDP index

Value added by small enterprise as percentage of total value added

Percentage of employees in small firms

1989 1990 1991 1992 1993 1994 1995 1996 1997 1998

Russia

Index Percent120

100

80

60

40

20

0

60

50

40

30

20

10

01989 1990 1991 1992 1993 1994 1995 1996 1997 1998

Lithuania

Index Percent120

100

80

60

40

20

0

60

50

40

30

20

10

01989 1990 1991 1992 1993 1994 1995 1996 1997 1998

Ukraine

Index Percent120

100

80

60

40

20

0

60

50

40

30

20

10

01989 1990 1991 1992 1993 1994 1995 1996 1997 1998

Hungary

Index Percent120

100

80

60

40

20

0

70

60

50

40

30

20

10

01989 1990 1991 1992 1993 1994 1995 1996 1997 1998

Kazakhstan

Index Percent120

100

80

60

40

20

0

60

50

40

30

20

10

01989 1990 1991 1992 1993 1994 1995 1996 1997 1998


44

GDP index Employment

FIGURE 4.6.

Employment and GDP, 1990–98(index 1990 = 100)

Hungary

Index120

100

80

60

401990 1991 1992 1993 1994 1995 1996 1997 1998

Czech Republic

Index120

100

80

60

401990 1991 1992 1993 1994 1995 1996 1997 1998

Poland

Index140

120

100

80

60

40

Ukraine

Index120

100

80

60

40

20

Russian Federation

Index120

100

80

60

40

Kazakhstan

Index120

100

80

60

40

1990 1991 1992 1993 1994 1995 1996 1997 1998 1990 1991 1992 1993 1994 1995 1996 1997 1998

1990 1991 1992 1993 1994 1995 1996 1997 19981990 1991 1992 1993 1994 1995 1996 1997 1998


Bulgaria

Index120

100

80

60

401990 1991 1992 1993 1994 1995 1996 1997 1998

Romania

Index120

100

80

60

401990 1991 1992 1993 1994 1995 1996 1997 1998


45

enterprises. With the former as an indica-tor of the fast dismantling of old industries,the rapid demise of the old sector seems nec-essary but not sufficient for the growth ofnew enterprises. A plausible reason wouldbe that the rapid dismantling of the old sec-tor lowers the price of its assets. Cheap re-sources become easily available to new en-terprises, useful when financing is notavailable and investment not forthcoming.That is why discipline is a crucial elementof growth. But it is not enough, for encour-agement is also required.

• Countries that reached the trough of the tran-sition recession sooner had faster growth inthe new sector. Where sustainable growthreturned, the share of small enterprises hadreached a critical mass. Where it did not,people remained “unemployed on the job,”as in the CIS and countries in SoutheasternEurope. Aggregate employment started to fallonly late in the process. These observationssuggest a sequence where hard budget con-straints are imposed and the old sector de-clines before the new sector can grow.

• While disciplining the old and encouragingthe new appear to be complementary, the oldsector has generally, though not invariably,proved unable to survive even where budgetconstraints have been soft and the new sec-tor has not emerged. In such cases, mainly inthe CIS, agriculture and low productivity ser-vices have served as “shock absorbers” forthose forced to leave old industries.

Can We Have It Both Ways, That Is,Protecting Old Inherited Enterprises andEncouraging New Enterprises?

The analysis makes clear that policies encour-aging the new private sector while subject-

ing them to market discipline must get thehighest priority. But must discipline and encour-agement go together? Or is it possible to encour-age the new sector while protecting the state en-terprises and privatized enterprises that continueto behave as if they were public? The economicargument seeing discipline and encouragementas complementary goes as follows. Continued

resource transfers for nonviable state-owned en-terprises, other things being equal, either requirethe consolidated government to loosen its fiscalstance, raise taxes elsewhere in the economy, orengage in off-budget activity.

• A looser fiscal stance can weaken the cred-ibility of the government’s stabilization pro-gram and, by increasing market perceptionof risk, raise domestic interest rates, thuscrowding out the new private sector.2

• Intensifying taxation of the potentially moreefficient emerging private sector can start avicious cycle that pushes enterprises into theinformal sector, thus lowering tax revenueand further increasing tax rates.

• Off-budget activities include issuing loanguarantees, establishing insurance schemes,and providing explicit or implicit cover gen-erally for politically motivated activities thattend to benefit incumbents.

These outcomes end up protecting state en-terprises and collective farms and discouragingnew enterprises. However, the relationship doesnot run simply from the costs of propping upunviable enterprises (protection) to discourag-ing the new private sector. The lack of a vibrantprivate sector limits the outside options thatmight attract workers and potential entrepre-neurs from old enterprises. It also limits prod-uct competition, an essential element of a disci-plining environment. Both weaken the incentivesfor closure and restructuring of nonviable en-terprises. These considerations explain why im-pediments to the efficient exit of enterprises dis-courage the growth of new enterprises, and whythe emergence of a new private sector wouldaccelerate the closure and restructuring of state-owned enterprises.

Further evidence on how protection of theold sector discourages the new sector comesfrom the following examples. The first is pro-tection through the banking sector. Small en-terprises have grown less in such CSB countriesas Bulgaria and Romania, where the bankingsector was a major conduit for loans to the oldstate-owned enterprises and farm collectives. Inthose countries nonperforming loans increased


46

as a share of total banking sector loans duringmuch of the 1990s. Contrast that with the vi-brant growth of new enterprises in the Baltics,Hungary, and Poland, where these loans weresharply reduced over time. In 1998 nonper-forming loans still represented 34 percent oftotal loans in Romania, while they were 4 per-cent in Estonia, 6 percent in Hungary, 11 per-cent in Poland, and 13 percent in Lithuania.These loans to old enterprises prevented the ex-pansion of bank credit to new, small, and lesspolitically connected enterprises. In addition,they became a major factor in triggering a bank-ing and macroeconomic crisis that called for asharp stabilization and credit tightening—allreducing the growth of new enterprises.

The second example is protection through spe-cial allocation of inputs. In Belarus and Uzbekistanthe old industrial sector has remained protected

through favorable foreign exchange regimes,directed credit, and high trade protection.Specific large foreign investments have also beenfavored. As a result new smaller enterprises facethe residual of these allocations in the market.What remains in credit and foreign exchange mustbe purchased at costs several times higher thanwould have been paid in unified markets. InUzbekistan small enterprises have to pay threetimes more for foreign exchange than do largestate enterprises to finance their imports. As a re-sult new enterprises have been squelched. InBelarus, the share of small enterprises in the totalnumber of enterprises is a mere 26 percent, sub-stantially lower than other CIS countries. Whilethese special regimes may have helped avoid asharp output decline in the old industrial sectorand aggregate output, they cannot provide thebasis 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 theCIS. 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 favorablegrowth performance stands in stark contrast to its extremely poor record on macroeconomic stabilization and structural reform. Apart from abrief spell of relative stability in 1996 and 1997, inflation ran at triple digits throughout the 1990s, reflecting money-based financing of publicsector deficits. The European Bank for Reconstruction and Development ranks Belarus 25th out of 26 transition economies in overall progressin economic reform.

Although Belarus has largely abolished the Soviet command system, a variety of controls over the economy remain. The state has continuedto 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 throughadministrative means. A multiple exchange rate system remained in place until September 2000, operating as a heavy tax on exporters and asubsidy for certain importers. While international trade has been liberalized, some restrictions remain in conjunction with domestic pricecontrols. 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 containedoutput decline? There are four important considerations. First, Belarus is successful only in relation to the other CIS countries. Compared withtransition 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, Belarusalso tended to accumulate substantial energy arrears to Russia, which were later settled through barter transactions that artificially maintaineddemand 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 andprevented 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 maintainindustrial production have been drained from other sectors of the economy. In particular, the Belarusian system has stifled the growth of new



47

enterprises. The multiple exchange rate system and pervasive state controls have shifted private initiative into arbitrage activities and theshadow 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 Transition Economies

Real GDP in 2000 Average inflation,a EBRD transition Government revenueCountry (1990 = 100) 1991–2000 indicator rating, 1999 b to GDP, 1999c

CSB average 106 41 3.1 39Czech Republic 99 13 3.5 41Hungary 109 20 3.7 42Lithuania 68 88 3.2 32Poland 147 26 3.5 43Romania 83 102 2.8 32

CIS average 62 185 2.3 24Belarus 89 344 1.6 42Georgia 30 257 2.9 16Kazakhstan 65 163 2.7 19Russian Federation 66 163 2.5 34Ukraine 43 244 2.5 35Uzbekistan 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 implementedreforms in many CIS countries, Belarus’s inaction enabled it to avoid some key mistakes in the early stages of transition. First, while continued stateownership did little to promote more efficient operational or strategic decisionmaking at the enterprise level, it did deter the large-scale assetstripping, tunneling, and tax evasion that has damaged growth in the early stages of transition in some other CIS countries. Nevertheless, the rapiddecentralization 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 thesetaxes in smoothing the initial output decline. The strength of fiscal revenues has kept the scope for supporting declining sectors and sociallyoriented expenditures. Belarus will need to maintain its ability to collect taxes if it chooses to embark on economic reform. Though it may betempting to explain this fiscal performance as a result of the authoritarian state, such a political regime has not guaranteed similar outcomes inTurkmenistan or Uzbekistan.

The third lesson is that current account revenues matter, especially in transition. Other CIS countries saw their exports to Russia collapsewhile their energy imports became far more expensive, leading to a massive terms-of-trade shock. While Central European transition economieswere redirecting their exports to the European Union, CIS countries had nowhere to go but Russia. Belarus continued to earn export and servicerevenues 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 economyfrom the rapid decline felt elsewhere in the region. Yet such an outcome was hardly guaranteed by the lack of economic reform, but rather bydependence on Russia, whose willingness to subsidize, directly and indirectly, Belarus’s inaction was a function of the geostrategic importanceof 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 stateinstitutions can be simply shifted from maintaining the status quo to implementing forward-looking economic policies in an environment

BOX 4.3 CONTINUED



48

The third example is protection through taxand utility arrears. Nonpayments by enterprisesto the main utilities in the energy sector remain amajor source of subsidization, particularly in Rus-sia and Ukraine, but also in Georgia, the KyrgyzRepublic, and Moldova. This has delayed restruc-turing of energy-intensive industries and sheddingof assets that smaller new enterprises could use.The discretion of negotiating and settlingnonpayments complements the natural behaviorof these utilities to act as discriminating monopo-lies. New, more energy-efficient enterprises arecharged more to compensate for the revenue lossesfrom the old, less energy-efficient enterprises.

Tax exemptions for large enterprises and ag-riculture collectives—popular until recently inUkraine, where these initiatives could be unilat-erally introduced by the legislature—generate ahighly discretionary tax regime and are an im-portant source of bribes. The same is true for theuse of negotiated offsets as a way to pay taxes inRussia, and tax avoidance by large enterprises inGeorgia. New small enterprises are usually lesspowerful in this environment and end up payinghigher bribes as a fraction of total profits.

In sum: the softer the budget constraint andthus the stronger the barriers to exit, the lowerthe contribution of small enterprises to employ-ment (figure 4.7).

Annex 4.1. Assumptions for Small and NewEnterprises

This discussion is based on the hypothesis thatsmall enterprises can be taken as approxi-

mating new enterprises. For this hypothesis, themargin of error for Central and Southeastern Eu-rope can be deduced from table A4.1. Although itseems safe to assume that most new enterprisesare small enterprises, the opposite need not be true.3

Table A4.1 lists the share of new and small enter-prises over total active enterprises in Central Eu-ropean countries for 1995. For example, of allactive enterprises in Albania in 1995, 68 percent(from column 2) were created as greenfield. Forthe same year our assumption would estimate theproportion of small enterprises to be 98.7 percent(from column 3). The difference of 30.3 percent(column 4) is the share of small enterprises thatwe consider to be new ones but actually are old.

BOX 4.3 CONTINUED

characterized by the rule of law is uncertain. Surely, if Belarus were to change course now and embark on economic reform, its initial conditionswould be far more favorable than those of the other CIS countries in the early 1990s, when demand was collapsing across the old Council ofMutual Economic Assistance. Most of Belarus’s trading partners are now growing and this should soften the negative shock of any reform effort.

While these considerations suggest that Belarus may still have the opportunity to develop into a strong economic reformer—once thepolitical will is there—an important question remains. Will economic reforms lead to the same erosion of public institutions observed in othertransition economies? Once opportunities for private wealth creation in the private sector arise, skillful civil servants may have greater incentivesto extract rents from the private sector or may even switch careers. The key challenge for Belarus, once it starts reforming, will ultimately be thesame as for all other transition economies: to establish good governance and strong market-supporting institutions.

FIGURE 4.7.

Soft Budget Constraints and Employment inSmall Enterprises, 2000

Source: EBRD (2000); World Bank database on SMEs.

Share of employment in small enterprises (percent)60

40

20

02 6

Soft budget constraints index

4

Hungary

Ukraine

Russia

Romania

Kazakhstan

Georgia

PolandEstonia

Czech Republic


49

Column 5 shows the ratio of small old enter-prises over small enterprises, that is, the probabil-ity of picking up a wrong element when samplingfrom the set of small enterprises. It is therefore ameasure of the error implied in assuming that smallenterprises equal new enterprises. The size of theerror for Albania is large: 30.3 percent means that,within the set of enterprises we assume to be newones, one out of three is not. But for Bulgaria theerror is 2 percent. Although Albania is extreme,many countries show high errors.

Figures for the Czech Republic, Hungary, andLithuania are a source of concern because thisstudy relies heavily on evidence from the threecountries. However, this crude estimate is anupper bound for the actual error. In fact, neithersmall-scale spinoffs nor new enterprises createdas private cooperatives are counted in column 2,although both can be considered new enterprises.The bias is particularly relevant for the coun-tries in the Federal Republic of Yugoslavia, suchas Slovenia, where many new enterprises werecreated as cooperatives.

The error tends to decrease over time. Be-cause the stock of companies in the hands of thestate at the beginning of the transition is eitherliquidated or privatized, the source of the erroris extinguished. The last column of table A4.1shows a projection for 1998 once a mortalityrate of 20 percent is assumed for both new andold enterprises. (Most of the data in this reportrefer to 1998.) The probability of error is esti-mated to have shrunk considerably for all thecountries this paper discusses.

Further evidence can be gathered through theanswers to the Business Enterprise Environmentand Performance Survey (see box 3.1). The sur-vey results show that 52 percent of new enter-prises are controlled by an individual owner (ver-sus 24 percent of old enterprises), and that intwo-thirds of the new enterprises the majorityownership is held by no more than three share-holders (true for less than half of old enterprises).Moreover, new enterprises are the ones that relymost on internal funds and family financing andthat have more trouble getting bank credit. The

TABLE A4.1.

Differences between New Enterprises and Small Enterprises, 1995 and 1998(percent)

Enterprises active January 1995 Projections for 1998

Small old Small oldenterprises as enterprises as

New Small Small old percentage of percentage ofCountry enterprises enterprises enterprises small enterprises small enterprises

Albania 68.4 98.7 30.3 30.7 18.8Bulgaria 96.0 98.1 2.1 2.1 1.1Czech Republic 86.1 98.6 12.5 12.7 7.6Estonia 81.9 96.4 14.5 15.0 7.8Hungary 84.5 99.0 14.5 14.6 9.6Latvia 83.2 95.2 12.0 12.6 6.2Lithuania 75.4 95.6 20.2 21.1 10.9Poland 88.6 98.7 10.1 10.2 4.9Romania 95.8 99.1 3.3 3.3 2.0Slovak Republic 92.6 98.4 5.8 5.9 3.4Slovenia 75.7 97.6 21.9 22.4 14.5Central European countries 88.3 98.6 10.3 10.4 5.7

Source: Eurostat (1998); authors’ calculations.


50

higher concentration in ownership and the ex-clusion from conventional credit channels is im-portant indirect evidence that new enterprisesare small.

Annex 4.2. Implications of the HigherProductivity of SMEs

In many transition economies SMEs havehigher average value added per employee than

the rest of the economy. As a result:

• Factors of production are likely to have anincentive to move to the SME sector, thus,increasing its size as a fraction of theeconomy.

• An increase in the size of the SME sector islikely to increase the growth rate of theeconomy.

Because SMEs represent the main source ofnew enterprises, we refer to the SME sector as“new.” The basic argument is that since SMEsare likely to be less capital intensive than largeror old enterprises, a larger value added per

employee means that primary factors of produc-tion earn more in that sector. So, by moving intothe sector, workers can get better wages, or en-trepreneurs can expect a better return for theircapital, or both. That is, primary factors in SMEshave more income to distribute among them.

Figure A4.1, based on the aggregate num-bers for Russia, illustrates the argument. Thetwo lines show the combination of wages (ormarginal product of labor, MPL) and profitrates (or marginal product of capital, MPK)compatible with the observed average productof labor in each sector (normalized so that av-erage value added per worker in the economyand total employment are both equal to 1). Theslope of each line reflects the average capitalintensity of the sector.4 The intersection withthe vertical axis is the average value added peremployee in each sector (as a ratio to theeconomywide value added per employee).

Assume that the “large enterprise” sector ischaracterized by point A, where the wage rate isabout 60 percent of the average value added perworker in the economy. The production possi-bilities indicated by the SME line would meanthat the new sector can pay wages 50 percenthigher at the same profit rate (indicated by apoint directly above A), or a much larger profitrate if paying the same wages (points directly atthe right of A, on the new line). So there wouldbe strong incentives for both labor and capitalto move to the SME sector.

Labor mobility in Russia has probably notbeen strong enough to equalize wages acrosssectors. So, on average, MPL for the SME sec-tor is likely to exceed MPL for the large enter-prise sector at the current position—particularlyconsidering the effective wage resulting fromold enterprises not paying their workers on timeor paying in kind, rather than the official wagerate. Because the existing capital stock is likelyto be more sector-specific than labor, its returnscould be even further from being equalized.Thus, a point like B appears as a reasonablecomparison point on the SME sector for a “typi-cal” bundle of labor and capital.

What are the gains to the economy when fac-tors move, say, from A to B? A critical element isthe extent to which factors, particularly capital,

Rate of profits (MPK, percent)

100 20 30

BA

FIGURE A4.1.

Factor Price Frontier: SMEs and the Rest ofthe Economy

Source: Authors.

Wage rate (MPL)1.5

1.0

0.5

0.0

SMEs Others


51

can be productively reallocated to new uses. Forthat, we define a “coefficient of shiftability to theSME sector” equal to the ratio of productivity ofcapital from the old sector to the productivity ofnew capital.5 Some types of capital, such as spe-cialized machinery and tools, may not be usableat all (a zero shiftability coefficient). Others, suchas liquid assets used as working capital, officeequipment, and some real estate, are likely to befully usable in the emerging sector. In many oldenterprises the main obstacle to productive real-location of capital may be its technological ob-solescence—though this also means that it is oflow productivity even for the old enterprises.

The figure shows the gain in GDP from real-locating to the SME sector 10 percent of the totallabor and capital of the economy, as a function ofthe “shiftability coefficient” defined above. Thepolar case of perfect capital shiftability—whencapital can move without any loss in productiv-ity—is also of interest because it represents thecase of new investment. Figure A4.2 shows thelarge efficiency gain in allocating those new re-sources to the emerging sector instead of usingthem on old enterprises; if 10 percent of laborand capital move to the SME sector, aggregatevalued added (GDP) would increase by almost 4percent (assuming constant returns to scale, asstandard on this type of model). The extra valueadded is the ex post validation of the ex ante po-tential earning difference, which provides the en-couragement for the creation of new enterprises.The large productivity gap in such countries asKazakhstan and Ukraine may thus be taken as anindicator of how much unrealized growth poten-tial exists in the new sector in these countries.

The other polar case is when only labor canmove to the new sector.6 In this case (showed bythe horizontal line in the figure) the gain is sim-ply the difference in marginal productive of la-bors across the sectors—about 1.2 percent in the“base” case marked by points A and B on thefactor price frontier.

Notes

1. The analysis is based on the hypotheses that smallenterprises can be taken as a proxy for new enter-prises, that spinoffs from the state sector do not mat-

ter because they reflect restructuring, and that theevolution of labor productivity in the manufactur-ing sector can be taken to approximate the extent ofrestructuring (see annex 4.1). The correspondingnumber for Georgia—40 percent—reflects growth ofnew enterprises plus the destruction of old produc-tive capacity due to armed conflict.

2. In Russia subsidies implicit in soft budgets ac-counted for two-thirds of net borrowing by the gov-ernment. As a share of GDP this was 7.5 percent in1996 and 5.6 percent in 1997. With the government’sdisinflation policy, this kept real interest rates in thetriple digits in 1995 and 1996, coming down to singledigits only in 1997.

3. Caves (1998) finds that most newly born enter-prises in market economies belong to the first twodeciles of the size distribution. A similar result can beexpected in transition economies, where starting largeenterprises from scratch is more difficult.

4. Thus the SME line’s lower slope reflects the as-sumption that SMEs are on average less capital inten-sive than large enterprises. For the average capital in-tensity of the economy we use a capital-output ratioof 6 (EBRD 2000). The SME line is constructed as-suming that they used about 8 percent less capital per

Shiftability coefficient (0 = null, 1 = complete)

0.50 1.0

Labor and capital move

Only labor moves

FIGURE A4.2.

Efficiency Gain from 10 Percent FactorReallocation to the SME Sector

Source: Authors.

Increase in GDP (percent)4

3

2

1

0


52

worker than the economy average (a capital-outputratio of 5, versus 6.23 for the rest of the economy). Ifboth sectors had lower overall capital intensity, thetwo lines would rotate out around the w-axis (for eachwage they would show a higher profit rate).

5. Thus (1-shiftability coefficient) is the share ofproductivity “lost” when moving from the old to thenew sector. The values on the figure are calculatedas follows. Value added per worker in each sector,v

i, can be written as v

i = r

i*k

i + w

i where k

i = capital-

labor ratio in sector i, wi = wage (or marginal prod-

uct of labor) in sector i, ri = profit rate (MPK) in

sector i, and i = 0 (old/large enterprises) or i = 1 (newenterprises/SMEs). If α is the “shiftability coefficient”

(0 for fixed factors; 1 for perfect mobility and sub-stitutability across sectors), the percentage gain inaggregate value added from reallocating 1 percentof labor and capital to the new sector is dV/V =w1 – w0 + k1 ( α r1 – r0). Of course, if the shiftabilitycoefficient, α, is too low—specifically lower thanα* = r1/r0—it is not worthwhile to reallocate capital,as the loss of output in the old sector would exceedproductivity in the new.

6. This is different from the case of α = 0 and capitaland labor reallocation. Here it is assumed that oldcapital is left in use in the old sector instead of beingreallocated, despite its low (or null) productivity inthe new sector.

53

Enterprises that subtract value at the prices prevailing after liberalization contribute negativelyto growth. They must be subjected to market discipline through such instruments as hardbudget constraints and competition in the markets for their products. Managers’ behavior

should be monitored using workers’ councils or banks while the institutions of corporate governanceare strengthened. At the same time, governments need to finance a targeted social safety net for thosehurt by the imposition of discipline on inherited enterprises. Both governments and enterprises need tomeet 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 andcontingent 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 PerformanceSurvey (see box 3.1) reporting arrears to the tax authorities and to state-owned energy producers, whichare two of the significant sources of implicit subsidy—is highest in the Caucasus and Moldova, followedby 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 theirlosses 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 timesas large. The main reason was the softness of the budget constraint of the enterprise sector—notonly 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 practiceof public utilities to borrow abroad with government guarantees. In Bulgaria and Romania an

5Imposing Discipline


54

excessive use of off-balance-sheet itemserupted into full-blown macroeconomic cri-ses during 1996–98 when contingent liabili-ties arising from the growth of bad debts inthe banking sector were converted into a fis-cal burden requiring explicit financing.

• As a result, all these countries needed “sec-ond generation” stabilization efforts in thesecond half of the 1990s.

• In Russia the failure to impose hard budgetconstraints through widespread use of arrearsand noncash payments led to governmentborrowing on an unsustainable scale, particu-larly in light of the global financial crisis in1998. That eventually led to default on a largepart of sovereign debt, an event that had se-rious repercussions throughout the CIS.

In the poorer CIS countries—Georgia, theKyrgyz Republic, Moldova, and Tajikistan—none of which entered the transition with exter-nal debt, soft budget constraints on public en-terprises, the difficulties associated with raisingenergy prices and disconnecting nonpayers toreduce consumption, and low tax collectionshave contributed to a rapid buildup of externaldebt. External shocks, such as the Russia crisisof 1998, have further exacerbated the problem.All four have a long-term solvency constraint andface tight liquidity in the next few years. Pre-liminary analysis suggests that new financing onhighly concessional terms and, in some cases,generous debt relief may be needed to restorethem to sustainable debt. The case for interna-tional action will be strengthened if the debtorcountries implement strong up-front adjustmentmeasures, eliminating tax exemptions and con-fronting powerful special interests (box 5.1).

Nonpayments Weaken the Incentives forEfficiency and Restructuring

Many governments have chosen to keepunviable enterprises afloat by extending

implicit subsidies through the energy companies.In Russia implicit energy subsidies to manufac-turing enterprises averaged 4 percent of GDP in1993–97. They take the form of arrears and non-cash settlements (including barter, promissory

notes, and tax offsets, where government spend-ing arrears and overdue tax payments are mutu-ally cancelled). In turn, the public utility compa-nies have passed on the costs of the implicittransfers to the general fiscal accounts by run-ning up huge tax arrears and unpaid dues toextrabudgetary funds, so that the burden of sub-sidies has eventually led to the accumulation ofpublic debt.

In Ukraine it is estimated that regional en-ergy companies alone have provided annual fi-nancing to nonpayers on the order of 4 to 5 per-cent of GDP. Moreover, governments faced withtax and expenditure arrears have complicatedthe nonpayments problem by encouraging bar-ter and other nonmonetary instruments to con-duct mutually offsetting operations.1 Much ofthis has continued regardless of government reso-lutions and legal acts forbidding the absorptionof further energy arrears by the budget. Enter-prises, their domestic and foreign suppliers (suchas Gazprom of Russia), and creditors believed,correctly, that despite the statements, some ex-plicit or implicit government support would beforthcoming. The absorption of gas arrears ofpublic institutions by the Moldovan budget in2000 was the fifth cycle of contingent liabilitycreation and moral hazard since the country’sindependence. Payments arrears eventually hadto be absorbed by the budget, often through taxoffsets, or, for energy-dependent CIS countries,through intergovernmental agreements with Rus-sia or Turkmenistan.

Nonpayments problems have been com-pounded in countries whose subnational govern-ments have considerable autonomy, such as Rus-sia and Ukraine. Clearance of tax arrears throughtax offsets gives subnational governments oppor-tunities to increase their retention of shared taxesat the expense of the center. In Russia subnationalgovernments are more actively engaged than thefederal government in clearing tax arrearsthrough offsets and individual tax exemptionsand deferrals for inefficient enterprises.

The consequences of the nonpayment syn-drome go beyond the budget. The failure toharden budget constraints and the opacity ofnoncash payments have weakened incentivesto use existing assets efficiently and to

Imposing Discipline

55

restructure enterprises. Nonpayments reinforceexisting interenterprise relationships. They alsoweaken competition by segmenting productmarkets and dealing with a few key suppliersand customers. Furthermore, by reducing trans-parency in accounting and transactions, theycomplicate monitoring of enterprise managers.Even healthier enterprises have benefited bycolluding with unviable enterprises to divert theimplicit subsidies available in the system ratherthan look for ways to restructure and increaseefficiency.

The widespread use of arrears and noncashpayments led to the proliferation of vertically in-tegrated conglomerates and provided an impetus

for numerous formal and informal financial-industrial groups, thus impeding competition andnew entry. Furthermore, subnational governmentshave protected incumbent enterprises with whichthey have built close relationships throughnonpayments at the expense of new entrants, in-hibiting investments and postponing the resump-tion of growth.

The experience of the 1990s clearly demon-strates that sustainable growth and low inflationrequire enterprise restructuring and exit, new en-try, and fiscal adjustment. But none of these canbe addressed without tackling the nonpaymentsproblem that affects them all. For this, the chal-lenge facing Russia and other CIS countries

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 world—with per capita incomes in 2000ranging from US$170 in Tajikistan to US$610 in Georgia. All five have small economies, narrow export bases, and depend heavily on energyimports. Four of the five are landlocked, and several face natural or conflict-related constraints on international trade. All began economictransition 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 percentof their combined GDP. The net present value of their debt outstanding at the end of 2000 averaged 158 percent of their exports and 358percent of their central government revenues.

Many factors have contributed to the accumulation of debt. Sharp increases in energy prices and the loss of transfers from the centralgovernment of the former Soviet Union after its breakup were massive shocks to these economies. Regional and internal conflicts hinderedeconomic 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 MonetaryFund and World Bank for financial assistance. In the mid-1990s, the international 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 disruptedtheir trade and financial sectors. Four of the five countries (Armenia is the exception) were forced to deeply devalue their currencies, leading tosharp increases in the domestic currency cost of external debt service.

Although the severity of the external debt burden varies among the five countries, the weakness of their external financing and fiscalpositions is likely to be a serious constraint to growth and poverty reduction, even in the medium term. Action is needed on two broad fronts ifthey are to make better progress toward sustained economic growth and poverty reduction with strengthened financial viability.

First, the low-income CIS countries need to improve their policy environments. Fiscal reforms are required to ensure improved efficiency andeffectiveness of public expenditures within the tight overall resource constraint. The pace of reform needs to be accelerated in key sectors, suchas energy, which have strong links to the fiscal and external debt situation. Major reforms are needed to improve the investment climate, notablythe environment for entry by new enterprises, which has proven to be a key success factor in Central and Eastern Europe. Institutional changesto strengthen public and private sector governance will be important in this regard.

Second, they need to work closely with their external partners to secure a volume and blend of financial assistance consistent with theirabsorptive 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 thesecountries 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.


56

differs somewhat from the hardening budget con-straints that the reforming countries of Centraland Eastern Europe faced at the onset of transi-tion. The following actions comprise the mainelements of a solution to the nonpayments crisis(Pinto, Drebentsov, and Morozov 2000):

• The government pays its bills in cash and ontime and refrains from engaging in mutualoffsets to cancel tax and budgetary arrears.

• Energy companies pay their taxes in cash andon time.

• Companies enforce a policy of disconnect-ing those delinquent on payments, thus im-posing a hard budget constraint on thoseenterprises.

• Social assets traditionally provided by enter-prises (such as housing, health clinics, and kin-dergartens) are divested to local governments.

• Benefits are targeted to the low-incomehouseholds most affected by an increase inenergy prices to international levels.

• Some types of one-company towns in distantregions of the CIS, such as distressed settle-ments in isolated areas and settlements witha predominantly retired population, get spe-cial arrangements.

The substantial real devaluation of 1998 hasgiven Russian enterprises some (temporary)breathing space, making it easier to eliminate theimplicit subsidies transmitted through tax andenergy payments. Noncash settlements continueto abate as a result of improved enterprise li-quidity from the devaluation. Although tax ratescontinue to be negotiated, almost 100 percentof current taxes are paid in cash.

Budget constraints are hardening though thevery channels that once transmitted hidden sub-sidies: government budget management, taxes,and energy payments. Cash collections from en-terprises by regional energy companies have in-creased significantly, a development attributedto unremitting pressure from such infrastructuremonopolies as Gazprom and RAO Unified En-ergy Systems. Indeed, these monopolies havebecome potent instruments for hardening bud-get constraints and insisting on cash payments.

Tax offsets are forbidden for the value-addedtax, profits tax, and income tax, and lack of

compliance could lead to federal transfers beingcut. The result is a chain reaction: the federalbudget does not transfer the funds needed forresidential electricity and heat, so the local au-thorities insist that the municipal utility compa-nies foot the bill. The latter in turn insist thatthe households pay. For enterprises, the mainpressure comes from taxes and cash paymentsfor energy. In addition, the restructuring of pastarrears is taking place by mutual consent and oncondition that current obligations are met in full.

Exit Mechanisms—Implement Now,Revise Later

Turning off the spigot of fiscal and quasi-fiscal support and keeping government

agencies current on their obligations is necessary.However, this alone is liable to lead to more accu-mulation of arrears and, without complementarypolicies, will not be enough to lead to restructur-ing or closure. Two such policies are important.First, restructuring and exit will require a transferof responsibility for housing and other social as-sets from enterprises to local governments, an areawhere the roles and responsibilities of local gov-ernment suffer from lack of clarity and absence offinancing. Second, measures of administrative liq-uidation need to be strengthened.

Almost all countries in the region have bank-ruptcy laws. In a number of countries there weresignificant reforms of these laws in 2000. Butthese formal exit mechanisms have not been par-ticularly effective. The European Bank for Re-construction and Development’s legal indicatorsurvey indicates that there is a gap between theadoption or amendment of bankruptcy legisla-tion and its effective implementation (EBRD2000). Experience suggests that it is preferableto implement existing insolvency laws and revisethem once they have been put into practice.

Competition Is Linked to Innovationand Growth

Encouraging competition in product marketsis another important ingredient of discipline.

Because socialist economies were highly con-centrated, exposing enterprises to internal and

Imposing Discipline

57

external competition is important. Results fromthe Business Enterprise Environment and Per-formance Survey (see box 3.1) show that nearly30 percent of state-owned enterprises face nocompetitors, compared with 5–9 percent forprivate enterprises. While more than half ofstate-owned enterprises have more than threecompetitors in their main product market, 80percent or more of privatized and new privateenterprises find themselves in that situation(EBRD 1999).

The survey also shows that enterprises fac-ing an intermediate degree of competition (oneto three competitors) develop new products, re-place managers, or change their organizationalstructure if they are subject to a hard budgetconstraint, with their moderate degree of mar-ket power providing the reward necessary to

innovate. Where monitoring of managerial per-formance by debtors and external shareholdersis weak—as is generally the case in the CIS andSoutheastern Europe—competitive productmarkets can help discipline enterprise manag-ers. The link between competition, innovation,and growth emerges as particularly important.Governments should thus vest the competitionpolicy agency with the authority to enforce com-petition laws strictly.

Note

1. As an illustration of failing to harden budget con-straints by disconnecting nonpayers, the Russian gov-ernment threatened to block oil producers’ access toexport pipelines if producers stopped supplying oil tononpaying domestic refineries.

59

What is needed to encourage the formation, and then growth, of new enterprises? Thischapter identifies administrative barriers to entry as one of the most important obstacles.It 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 ofnurturing new enterprises and monitoring enterprise reform. It also reviews reforms in the tax systemand 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 factorsinfluencing the investment climate into microeconomic variables (including taxes and regula-

tions), macroeconomic variables (including policy instability, inflation, and exchange rates), and lawand 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 mostimportant 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—areseen as serious obstacles. The other obstacles, in order of importance, are inflation, lack of access tofinance, 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 inCentral Europe and the Baltics. The survey also shows that small enterprises across the region pay, onaverage, 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, andUzbekistan. The frequency of bribe payments is about double for small enterprises: 37 percent reportpaying bribes frequently, compared with 16 percent for large enterprises.

New enterprises see corruption and anticompetitive practices as two of the most difficult obstaclesacross all parts of the region. Business licensing, which falls in the category of taxes and regulations, is

6Extending Encouragement


60

seen as particularly troublesome by new enter-prises, given the opportunity for arbitrary be-havior in interactions between officials and en-terprises in granting and renewing licenses. Alsotroublesome are regulations for customs, foreigntrade, foreign currency, foreign exchange, andtaxes. Examples of such obstacles abound.

A survey of Russian enterprises by the WorldBank in 1996 shows that the average new busi-ness applicant had to deal with about 25 differ-ent agencies and complete about 70 registrationforms. There were 30 kinds of licenses for a busi-ness startup. About a third of surveyed enter-prises indicated that they were forced to obtaina license that in their opinion was not required.

In Ukraine an International Finance Corpora-tion study showed that small enterprises endurean average of 78 inspections a year, requiring 68written responses. Dealing with inspections andaudits consumes two days a week of the averagemanager’s time and requires cash outlays ofabout US$2,000.

The administrative problems that new en-trants in Armenia (box 6.1) face in registering,locating, and operating their companies are per-vasive throughout the region. The solutions liein simplifying processes to make them transpar-ent and consistent across all government agen-cies. Similarly, customs procedures need to be re-engineered to make them more transparent and

BOX 6.1.

Reducing the Cost of Entry and Doing Business in Armenia

New enterprises face many problems in Armenia that could be reduced with appropriate measures.

Registration

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 sixmonths.

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.

Permits

Registering ownership rights is a convoluted bureaucratic process, designed more as a data collection system than as a process to register legalrights. Many steps appear irrelevant. Armenia’s land cadastre system records ownership based on street addresses or passport numbers, ratherthan a unique cadastral code established through surveys. Site development is equally difficult, with numerous municipal and state agenciesinvolved in issuing the approvals and clearances required for construction and occupation permits. There is no comprehensive description of therequired 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 various agencies• Introduce application forms for all agencies and attach a reliable and clear guide• Set timetables for each phase.


Extending Encouragement

61

to reduce the scope for arbitrary decisionmakingand abuses of power.

Enterprises Lack Confidence in Legal andJudicial Institutions

The state can create a good climate for in-vestment by protecting the security of prop-

erty and contract rights. Variation in thesecurity of property rights in the region is high(figure 6.1). Fewer than 30 percent of enterprisesin such countries as Croatia, Estonia, and Po-land lack confidence in the security of propertyrights, compared with more than 70 percent inRussia and Ukraine.

Institutions contributing to the security ofproperty rights are diverse: property registries; astable and appropriate body of laws, regulations,and administrative procedures; effective systemsto adjudicate administrative and civil disputes;and legal expertise (lawyers, notaries, judges,prosecutors, police, and bailiffs) to ensure thatthe legal framework is enforced. Data are avail-able on the quality of legal drafting (figure 6.2),which is eventually linked to the stability andappropriateness of the legal framework and tothe quality of the judicial system (figure 6.3).

On the basis of the Business Environmentand Enterprise Performance Survey, more than90 percent of enterprises in 13 of the samplecountries report that they are not adequatelyconsulted in the drafting process for new lawsor policies. More than 80 percent of enterprisesin 13 of the sample countries report that theyare not adequately informed of changes in rulesthat affect them before these rules are adopted.The quality of the judiciary is unbundled intofairness (bias, honesty, consistency), cost (finan-cial and time), and likelihood of enforcement(see figure 6.3). There is wide variation in per-ceptions of the quality of the judiciary. In Hun-gary, Estonia, and Slovenia only a third of en-terprises report that court decisions are unfair,while in Armenia and Lithuania almost 90 per-cent of enterprises do. In Uzbekistan fewer thanhalf of the enterprises complain about the highcost of litigation, while in Poland more than 90percent of enterprises do. In Bulgaria andSlovenia only about a quarter of enterprises

complain about low enforcement rates, whilein Albania almost 90 percent of enterprises do.

Almost two-thirds of the variation in the se-curity of property rights across countries can beattributed to variation in their systems of legaldrafting and the effectiveness of their judiciaries.Thus, attention to the process of drafting andenforcing laws may help ensure that laws aresuited to achieve economic outcomes—in thiscase, secure property and contracts.

A study based on a survey of manufacturingenterprises in Poland, Romania, Russia, the Slo-vak Republic, and Ukraine found that enterpriseswith the least secure property rights investednearly 40 percent less than those with the mostsecure property rights.1 Moreover reinvestment

FIGURE 6.1.

Insecurity of Property Rights in TransitionEconomies, 1999


Hungary

Poland

Czech Republic

Estonia

Slovak Republic

Slovenia

Latvia

Lithuania

Croatia

Kyrgyz Republic

Kazakhstan

BulgariaGeorgia

Moldova

RomaniaArmenia

Albania

RussiaUkraine

Azerbaijan

Uzbekistan

Belarus

806040200

Percent


62

of profits was a bigger source of investment capi-tal than either bank funds or trade credit in allfive countries. In such countries as Russia andUkraine, which have made less progress in tran-sition than Poland and the Slovak Republic, re-forms to secure property rights are more impor-tant for promoting the investment climate thanreforms in the banking sector. Then, as the tran-sition proceeds and legal and judicial reforms

strengthen property rights, a well-functioningfinancial sector is needed to provide credit toenterprises making new investment and requir-ing financing beyond their retained earnings.

Financial Deepening Is Slow but Progressing

Although securing property rights appears tobe more important than reforming the

Voice

Publicity

Hungary

Poland


Slovak Republic

Slovenia

Latvia

Lithuania

Croatia

Kyrgyz Republic

Kazakhstan

BulgariaGeorgia

Moldova

Romania

ArmeniaAlbania

RussiaUkraine

Azerbaijan

Uzbekistan

Belarus

10080706050

Percent

90

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)

Note: This figure shows partial results from two questions on the Business Environment and Enterprise PerformanceSurvey (see box 3.1). The first question measured enterprises’ voice in legal drafting: “In case of important changes inlaws or policies affecting my business operation, the government takes into account concerns voiced either by me or bymy 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 businessesare informed in advance of changes that will affect them.” For both questions the possible responses were “always, mostly,frequently, sometimes, seldom, or never.”



63

banking system in the early stages of transition,financial deepening and development is key tosustaining the growth of the new private sector.While transition economies have come a long

way in developing banks and capital markets,their financial sectors remain underdeveloped byinternational standards. Banking systems in mosttransition economies remain noticeably small in

FIGURE 6.3.

Quality of Judiciary in Transition Economies(percentage of enterprises that complain that courts sometimes, seldom, or never exhibit positive qualities when resolvingbusiness disputes)

Note: This figure shows partial results from the following question on the Business Environment and EnterprisePerformance Survey (see box 3.1): “Now, thinking about our country’s legal system, how often do you associate thefollowing 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 “weak enforcement”bar comprises the descriptions consistent/reliable and able to enforce decisions.


Hungary

Poland

Czech Republic

Estonia

Slovak Republic

Slovenia

Latvia

Lithuania

Croatia

Kyrgyz Republic

Kazakhstan

Bulgaria

Georgia

Moldova

Romania

Armenia

Albania

Russia

Ukraine

Azerbaijan

Uzbekistan

Belarus

Unfair

High cost

Weak enforcement

1006040200

Percent

80


64

international perspective, even more so oncebank credit to the public sector, often reflectingsubsidization of unprofitable activities, is re-moved and attention is confined to the privatesector (figure 6.4).

Similarly, stock market capitalization, a mea-sure of the capacity of securities to provide fi-nance to the real economy, shows that financialdevelopment lags behind other countries of theworld with similar per capita incomes (figure 6.5).Relative to their peers in the region, Croatia, theCzech Republic, and Hungary boasted the high-est stock market capitalization in 1998.2

The Banking Sector Is Moving Forward…

Although the financial sector appears underde-veloped from an international perspective, thisis not necessarily true for financial intermedia-tion in the banking sector. The ratio of operat-ing costs to total assets reveals that several Cen-tral European countries have progressed in recent

years (figure 6.6). In 1997 all Central Europeancountries except Slovenia had operating cost tototal asset ratios that were lower than those incountries with comparable incomes elsewhere.While the Baltic countries had not quite matchedtheir peers in the world, their average ratio ofoperating costs to total assets fell from 7.8 per-cent in 1996 to 5.6 percent in 1997. The remain-ing transition economies with data have ratiosless favorable than their peers, but they too (ex-cept Romania) saw efficiency improve. An alter-native measure of efficiency—the difference be-tween the average interest rate on loans to theprivate sector and the resident deposit rate—alsoshows Central Europe and the Baltics in a favor-able situation, but not Southeastern Europe andthe CIS (figure 6.7).

Although domestic credit to the private sec-tor (as a share of GDP) in the transition econo-mies is generally low in comparison with coun-tries of similar per capita incomes (figure 6.4),in many of the most advanced reformers an

FIGURE 6.4.

Domestic Credit by Deposit Money Banks to the Private Sector, 1998

a. Kyrgyz Republicb. Albaniac. Georgiad. KazakhstanNote: The world fitted line represents the average of all countries in the world in that per capita income range.Source: Eichengreen and Ruehl (2000).

Percentage of GDP100

80

60

40

20

00 10,000 15,000

Per capita income (US$)

HungaryUkraine Russian Federation

Romania

Slovak Republic

Bulgaria Poland

Lithuania

EstoniaLatvia

Czech Republic

5,000

Slovenia

Moldova

Croatia

ArmeniaAzerbaijan

BelarusMacedonia, FYR

a b c d

World fitted line

Central Europe andthe Baltics


Southeastern Europe


65

FIGURE 6.6.

Operating Costs to Total Assets in the Banking Sector, 1997

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).

Percentage of total assets14

12

10

8

6

4

2

00 10,000 15,000


Hungary

Ukraine Russian FederationRomania

Slovak Republic

Bulgaria

Poland

LithuaniaEstoniaLatvia

Czech Republic

5,000

Slovenia

Kazakhstan

World fitted line



Southeastern Europe

FIGURE 6.5.

Stock Market Capitalization and Per Capita Income, 1998


Percentage of GDP70605040302010

0–10

0 10,000 15,000


Hungary

Ukraine Russian FederationRomania

Slovak Republic

BulgariaPoland

LithuaniaEstoniaLatvia

Czech Republic

5,000

Slovenia

KazakhstanMoldova

Croatia

Armenia

Azerbaijan

World fitted line



Southeastern Europe


66

important part of that credit is now flowingto new enterprises. In Estonia, Hungary, andLatvia between two-thirds and three-quartersof total private credit is going to those enter-prises.

The Czech Republic and Hungary provide aninteresting contrast. The Czech Republic has thehighest share of credit to the private sector, about60 percent of GDP, (see figure 6.4), but only one-tenth of it flows to new enterprises. In Hungary,the credit provided to the private sector is muchlower (about 25 percent of GDP), but about three-quarters of it goes to new enterprises. Why? Be-cause in the Czech Republic most banks were stillstate-owned and largely unreformed, and theycontinued to lend to their traditional clients: largesemi-privatized enterprises. This is consistent witha situation of banks not having been positionedto do their job of screening, together with thelimited industrial restructuring and lack of growthin the Czech Republic in recent years.

Hungary, by contrast, sold off almost all itscommercial banks to foreign investors early on inthe transition. These new private banks appliedto their borrowers the hard budget constraintsimposed on them by their headquarters and strictsupervision by the supervisory agencies in

Hungary. Thus comprehensive financial restruc-turing and curtailment of loans to the banks’ tra-ditional clients have been associated with fastergrowth of financial intermediation to new privateclients. The experience of Hungary suggests thatrestructuring the financial sector and developingfinancial intermediation can go hand in hand.

…Can the Discipline-and-Encourage StrategyQuicken the Pace?

Hard budget constraints should apply to banksand enterprises alike. A credible threat of exit isnecessary for a competitive banking system. Buta bank differs from a nonfinancial enterprise, soa decision on liquidation or restructuring, espe-cially for large banks, must take into account thesystemic risk of closure. The efficiency of banksimproves if they are privatized to strategic inves-tors. If such investors are not forthcoming,privatization to concentrated owners should beconsidered, so long as there is a clear separationbetween investors and borrowers from the bank.

Facilitating the entry of foreign financial in-termediaries can accelerate the development ofthe banking sector and in turn help the growthof start-ups not yet ready to tap capital markets.

FIGURE 6.7.

Interest Rate Spreads, 1998


Percentage points50

40

30

20

10

00 10,000 15,000


Hungary

Ukraine

Belarus

Russian Federation

Romania

Slovak Republic

Bulgaria

Georgia

PolandLithuania

Estonia

LatviaCzech Republic

5,000

SloveniaMacedonia, FYRMoldova

Croatia

Armenia

Kyrgyz Republic World fitted line



Southeastern Europe


67

Entry by foreign banks and the acquisition ofdomestic banks by foreign banks are quick waysof importing managerial and supervisory exper-tise, the latter because supervisory responsibilityresides mainly with the home country regulatoryauthority. While encouraging competition is criti-cal for the long-term health of commercial banks,it should be done without relaxing prudentialnorms. These norms should cover minimum capi-tal requirements, regulatory capacity, legal re-course, and the independence of supervisory au-thorities from political interference.

Recent work on the relationship between fi-nance and growth suggests that banks and secu-rities exchanges provide different services, bothrequired in a mature market economy (Levineand Zervos 1998). The financial system in tran-sition economies will most probably remainheavily bank-based for a while. New enterprisesneed to be nurtured and this phase calls for abank-based financial sector capable of provid-ing risk assessment and venture capital services.A later phase will call also for an active and li-quid securities market that is more suited to fund-ing risky projects and offers fuller risk diversifi-cation. This later phase will require even moredemanding institutional developments such as aregulatory framework requiring information dis-closure and preventing insider trading and otherforms of market manipulation.

Privatization Attracts Foreign DirectInvestment, and Positive Spillovers Follow

There are significant differences in the levelsof foreign direct investment flowing to the

transition economies. In the 1991–2000 decade,the CIS received cumulative foreign direct invest-ment per capita of about US$115 compared withUS$800 for the CSB. The Czech Republic andHungary had the highest amounts, almostUS$2000, followed by Estonia, Latvia, andCroatia in the US$300 to US$1000 range.

Cash privatization of large enterprises hasdriven much of these foreign direct investmentflows; cumulative foreign direct investment ishighly correlated with cumulative privatizationrevenues (EBRD 2000). However, foreign di-rect investment has also flown to countries as a

response to improved policies and improvedoverall business environment. Some studies haveshown a clear statistical relationship betweenthe growth of foreign direct investment and theindices of economic liberalization discussed inchapter 2 (Selowsky and Martin 1998). Theability to attract foreign direct investment is thusa good proxy for the attractiveness of acountry’s overall investment climate and is alsocrucial to promote the entry and growth ofSMEs. Actually both are well associated asshown in figure 6.8.

In addition to the positive spillovers of im-proved technology, better management skills, andaccess to international production networks, for-eign direct investment has proved resilient in vola-tile international capital markets. Although theRussia crisis of 1998 led to sizable outflows ofbond and portfolio equity capital, this was offsetby an increase in foreign direct investment in theadvanced reformers in Central Europe and theBaltics. Russia and Ukraine, reluctant to attractforeign participation in their privatization pro-grams, had attracted only modest amounts of for-eign direct investment but substantial portfolioflows. Much of the latter found its way into fi-nancing public sector deficits arising from an over-all climate of soft budget constraints. When thegovernment’s solvency was questioned in the 1998crisis, portfolio flows to Russia quickly reversed.

Tax Reform: Broadening the Base andLowering Rates

The agenda of tax reform in the transitioneconomies is large. Its guiding principle,

based on the lessons of theory and experience, isthat a tax system raising revenue for the govern-ment at least cost to the economy should be stableand broad-based and have low statutory rates.The focus in this section is narrow, restricted toaspects of taxation that bear directly on disci-pline and encouragement.

The key objective of tax reform is to broadenthe base of taxation and reduce statutory rates.This would eliminate tax exemptions enjoyed byfavored enterprises, thus hardening budget con-straints. It would also reduce the burden of taxa-tion on viable enterprises and encourage the


68

informal sector to come within the tax system.The informal sector in 1995 was estimated at 60percent of GDP in Azerbaijan and Georgia, be-tween 40 and 50 percent in Russia and Ukraine,and between 30 and 40 percent in Bulgaria,Kazakhstan, Latvia, and Moldova. In the KyrgyzRepublic tax exemptions, mostly to foreign jointventures, have in the past cost the budget theequivalent of 5–7 percent of GDP, while revenuecollections remain below 16 percent of GDP. InRussia the potentially large tax base was erodedby exemptions granted by subnational authori-ties on federal taxes, estimated at 0.8 percent ofGDP (World Bank staff estimates).

Taxes on small businesses are widely reportedto be a disincentive to creating small enterprisesin the region. The turnover threshold for becom-ing a value-added taxpayer should thus be set highenough to exclude small enterprises, which shouldbe subject to a small-enterprise tax regime, per-haps as simple as a fixed annual fee in lieu of taxesother than wage and social taxes. Larger enter-prises might be subject to either a turnover tax ata moderate rate or to a simplified cash flow tax

using simplified accounting. Such an approachwould relieve the administrative and reportingburden on the taxpayer and reduce the interac-tion between the taxpayer and the tax authority.

For agriculture, a property tax needs to bedeveloped as land privatization continues. Thiscould be the primary method of agriculturaltaxation in the foreseeable future. Parallelingthe earlier discussion, a distinction needs to bemade between small agricultural units andlarger agribusinesses. Smaller agricultural en-terprises should be subject to the small busi-ness tax (or personal income tax, as appropri-ate). Larger enterprises should become morefully integrated into the modern income tax andvalue-added tax systems.

Intergovernmental Fiscal RelationsSupporting Discipline and Encouragement

It is important for the incentives facing federaland local governments to be aligned in sup-

port of the discipline-and-encourage strategy.Subnational governments should have incentives

FIGURE 6.8.

Cumulative Foreign Direct Investment Per Capita and Employment in Small Enterprises, 1998

a. Ukraine.b. Belarus.c. Russian Federation.Source: EBRD (2000); World Bank database on SMEs.

Cumulative foreign direct investment (US$)1,8001,6001,2001,000

800600400200

00 10 20 30 40 50 60

Contribution of small enterprises to employment (percent)

Hungary

Romania

Kazakhstan

BulgariaGeorgia

Poland Lithuania

Estonia

Latvia

Czech Republic

cba


69

to identify with small business. To save jobs andto protect existing political structures these gov-ernments often protect large (often bankrupt)enterprises in their localities, even though jobgrowth generally increases with the developmentof small businesses. However, the latter takes time.

To help subnational governments identifywith new emerging enterprises, small businesstaxes and property taxes could be allocated tothat level of government. Subnational govern-ment officials will then see that the success ofsmall business means more local revenues. Com-pliance with the property tax will be enhanced ifthe tax is tied to improvements in public services,such as police protection demanded by localbusinesses. That in turn increases property val-ues, reinforcing mutual interests.

It is also necessary to ensure that local gov-ernments do not face an incentive to treat en-terprises differently because they continue toprovide social services, such as hospitals, kin-dergartens, and housing recreation facilities. Inmany cases municipalities do not have the re-sources to provide these services. This createsincentives for subnational governments to pro-vide special treatment to unrestructured andsometimes nonviable old enterprises, with theattendant deleterious consequences for thegrowth of new enterprises.

It is important that an effective strategy bedeveloped and implemented for divesting enter-prises of these social services. In some cases thismeans placing the responsibility of service pro-vision on the municipality. In others it meansreassigning responsibilities, or at least financing

the service at a higher level of government. Inyet other cases it means a complete redesign ofthe provision of social services and benefits.

Many countries in the region have shifted so-cial responsibilities to local governments, buthave not ensured that local governments haveenough resources to manage these responsibili-ties, either through bigger transfers or greatertax autonomy. In addition, the divestiture of so-cial services is frequently done in a way that leadsto confusion over which level of government isresponsible for what. Roles and responsibilitiesof subnational governments have to be clarified.In addition, enough resources must be allocatedto the additional responsibilities of absorbingsocial assets from enterprises. That reduces theincentive for localities to favor old enterprisesover new.

Notes

1. The property rights variables used in the analysiscombine an index of property rights security compris-ing extra legal payments for licenses, extra legal pay-ments for services, and payments for protection witha measure of the effectiveness of courts. See Johnson,McMillan, and Woodruff (2000).

2. Several transition economies, notably in CentralEurope, appear to perform better when market liquid-ity is measured by the ratio of turnover (a measure ofthe liquidity of the securities) to market capitaliza-tion. These high turnover ratios, rather than reflect-ing genuine liquidity of the markets, are driven byhigh privatization sales captured (the numerator) inconjunction with low levels of capitalization (the de-nominator).

71

Privatization has been key in the transition from plan to market. Rapid privatization early in thetransition aimed to get the state out of enterprise management, to create a broad constituencyfor reforms, and to arrest “spontaneous privatization” from pretransition reform attempts in

countries as diverse as Hungary, Poland, the Russian Federation, and Ukraine. Privatization was away of imposing hard budget constraints and promoting restructuring. It was also a way of creatingdemand for stronger property rights and institutions of corporate governance, thus contributing toprivate sector development.

What is the evidence on privatization’s effect on restructuring in medium-size and large enterprisesover the past decade in transition economies? Djankov and Murrell’s (2000) recent review of theliterature, 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 notbeen beneficial; indeed, privatization to workers in the CIS has been worse than state ownershipfor restructuring.

A reasonable hypothesis for the ineffectiveness of diffuse or insider ownership is the lack of anadequate institutional framework. That framework would include strong mechanisms of corporategovernance, including rules to protect minority shareholders; rules against insider deals and conflictsof 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 privatebanks. Without such institutions those in control of enterprise assets face few restrictions to preventdiversions for private gain.

Other factors may also have contributed to the poor performance of enterprises privatized toinsiders or diffuse owners. These owners probably had neither the ability to provide the capital tomodernize the enterprise, nor the market experience to cope in a competitive environment, nor thedrive and vision to guide the radical restructuring often required to improve enterprise performance.

7Privatization: Lessons and

Agenda for the Future


72

Furthermore, workers in the CIS, traditionally aweak force in the enterprise, had almost no own-ership or management experience. Also, owner-ship by workers may have done little for restruc-turing if employment had to be reduced sharply.

Concentrated ownership tends to offer thosecontrolling enterprises the incentive to maximizelong-run enterprise profits. Owners can aligntheir interests with minority shareholders andcreditors, reducing the likelihood of asset strip-ping for private gain. But privatization in transi-tion economies shows that concentrated owner-ship alone is not enough for effective governance.

Concentration of ownership must be deepenough to reduce the divergence between theinterests of controlling shareholders and otherstakeholders and to lead controlling sharehold-ers to pursue long-run profit maximization. Anownership control of only 20–30 percent maynot be high enough to align interests.

The type of concentrated ownership also mat-ters. Enterprises controlled by strategic investorshave performed much better than those controlledby investment funds, other stakeholders (hold-ing companies), or other financial institutions.Enterprises controlled by foreign strategic inves-tors have generally been the best performers. Thisis not entirely unexpected. Such investors, ableto provide more resources and skills for restruc-turing, had an advantage over domestic strategicinvestors early in the transition.

The selection of strategic investors matters,too. Enterprises sold through transparent ten-ders or auctions have generally attracted betterowners, outperforming enterprises sold directlyto politically connected parties, frequently athighly subsidized prices.

So, the ideal strategy is to transfer assets asrapidly as possible to individual investors orconcentrated groups of strategic investorsthrough open, fair, and transparent methods.Some countries—such as East Germany, Esto-nia, and Hungary—adopted this strategy, withsatisfactory outcomes. Why did other countriesnot follow suit?

Privatization to diffuse owners and worker-owners was appealing on equity grounds, andin several countries this was the only way to make

private ownership politically acceptable. It wasalso thought that any inefficiency of diffuse own-ership would be only temporary, as secondarymarkets would consolidate share in the hands ofeffective owners. In any event, there were fewconcentrated domestic owners to buy state as-sets. Selling to foreigners was politically difficult,and such investors were sometimes less availableto the CIS countries, given the overall uncertain-ties about the investment climate and lack oftrack record. Other factors included country size,geography, political uncertainty, inadequate le-gal framework, and weak institutional capacity.

Might it not have been preferable to keep theassets in state hands, waiting to identify and thensell the enterprises to reliable strategic investors?Yes, on two conditions. First, the privatizationagency needs the autonomy to discharge its func-tions with transparency and without political in-terference. Second, there has to be enough insti-tutional capacity to prevent asset stripping bystate managers in the interim. In many countriesthese conditions were not met, resulting in “spon-taneous” privatizations by managers when theenterprises were still owned by the state. TakeRussia at the end of the Gorbachev era, whenthe state had lost control over enterprises afterthe collapse of institutions during the dissolu-tion of the Soviet Union.

The challenge in economies with a large un-finished agenda of privatization is for the stateto exercise its control rights to avoid tunnelingand theft by enterprise managers during the yearsit could take to complete privatization and whileenterprises remain in state hands. Yes, divesti-ture to reputable core investors who put theircapital at risk is best practice. However, that isdifficult to achieve on a large scale in a shortperiod. A major question then is whether inter-mediate modes of privatization—such as massprivatization through vouchers, or management-employee buyouts—might eventually lead to thesame result, even if they temporarily worsen en-terprise performance.

Navigating between continued state owner-ship with eroding control rights and a transferto ineffective new private owners with an inad-equate institutional framework is possibly one

Privatization: Lessons and Agenda for the Future

73

of the most difficult challenges confrontingpolicymakers in charge of privatization. Thereis a substantial risk of asset stripping and lossesof economic value in both cases.

How much can tunneling and theft in the in-termediate stages be prevented? How fast canreputable concentrated investors become the ul-timate owners? Does the outcome depend on thepath followed? The present value of the benefitsof a specific privatization strategy depends on theanswers. For example, do rapid intermediatemodalities—such as privatization to diffusedowners or insiders—accelerate or retard the even-tual takeover of the enterprise by the “right” kindof investor? What are the relative magnitudes ofthe gains or losses of these intermediate stages ofownership in relation to state ownership? Theanswers depend on initial conditions and the ca-pacity to reduce tunneling (see box 3.2) and theftand thus enhance the benefits of the strategy.

Traditional Privatization or RapidPrivatization?

If countries choose traditional methods ofprivatization to effective owners, establishing

the necessary legal and institutional frameworkand giving the privatization agency autonomyand resources are essential. However, if coun-tries choose methods of rapid privatization thatlead to diffuse or insider ownership, strengthen-ing and enforcing the regulatory and supervisoryframework are crucial to enhance the account-ability of corporate boards and managers, toprotect the rights of minority shareholders, topromote disclosure, and to ensure that second-ary trading is conducted at fair and transparentprices. Building these institutions requires timeand resources in both cases.

Where court enforcement of contracts isweak, these provisions should be supplementedby stock market regulation for financial inter-mediaries, such as investment funds and bro-kers, which can monitor compliance by otherparticipants in financial markets. This wouldhelp set the stage for medium-size and largeprivatizations in countries that still have a sig-nificant reform agenda. It would also improve

the performance of privatized enterprises byassisting a transparent consolidation of widelyheld shares and new private enterprises. Itwould also facilitate the development of bankand nonbank intermediation.

The success of the two approaches requiresopenness to foreign investors. The importanceof foreign capital in the first approach is obvi-ous. It makes little sense to exclude foreign in-vestors from a privatization program designedto transfer enterprises directly to effective own-ers, as foreign investors are often in the best po-sition to provide the resources and skills for re-structuring state enterprises. However, foreigninvestors are equally important in the secondapproach. They can usually buy large blocks ofshares in the secondary market relatively quickly,while also complying with all capital marketregulations, thus facilitating the concentrationof ownership through fair and transparent sec-ondary trading.

Excluding foreign investors from massprivatization programs using vouchers could forcepolicymakers into difficult choices. For example,the authorities might have to tolerate unfair prac-tices to allow cash-strapped domestic investorsto rapidly gain control. The concentration ofownership in many voucher programs, as in theCzech Republic, owed much to poor capital mar-ket regulation and weak rule enforcement. But ifthe authorities had been willing and able to en-force an adequate regulatory framework, theyprobably would have been forced to accept alonger period of diffuse ownership.1

Why Countries Did What They Did

At the end of the 1980s the best-knownprivatizations were negotiated sales to stra-

tegic investors—or privatization by issuingshares—along the lines applied in Great Brit-ain and a few other OECD countries. The cir-cumstances in transition economies demandeddifferent approaches. For example, all transi-tion economies contained thousands of state-owned small business and service units, of a typeunknown in the West. Most countries begantheir privatization by divesting these small


74

enterprises, usually by simple methods basedon auctions. Small-enterprise privatization is ev-erywhere regarded as a success. But as the sizeof the enterprise increased, so did the complex-ity of privatization solutions.

At the beginning of the 1990s there was ageneral perception of the advantages of a fairlyrapid ownership transfer. If speed was important,what could be done? Quick cash auctions, it wasfeared, would put most assets in the hands ofthe old nomenklatura or the frontmen for for-eigners—with the possibility that the publicwould turn away from privatization and reform.Speedy privatization was also seen as a responseto pretransition attempts at enterprise reform,such as in Russia, that had empowered manag-ers and allowed leasing, cooperatives, and otherarrangements often characterized as “spontane-ous privatization.” Rapid transfer of ownershipwas seen as a way to stop unfair leakage of as-sets to managers.

Vouchers were seen as the answer, for theywould give purchasing power to the populationin a transparent and fair way. Secondary trad-ing, facilitated by investment funds, would al-low transparent consolidation of shares. Russiaand then Czechoslovakia applied voucherschemes universally, and Armenia, Azerbaijan,Georgia, the Kyrgyz Republic, Lithuania, andMoldova among others used them as the princi-pal method of divestiture. Vouchers were a sec-ondary privatization method in several othercountries (see table 7.1).

Russia: From Spontaneous Privatization to DiffuseOwnership

Russia was unique because of the large numberof enterprises to be privatized (about 25,000medium-size to large companies) and the strengthof spontaneous privatization by enterprise man-agers after the dissolution of most branch minis-tries under Secretary Gorbachev. Auctions hadto be conducted across 89 regions over territorycovering six time zones. A decentralized and sys-tematic approach had to be developed at the cen-ter, with autonomous regional implementation.The intent of the reformers was to award only a

minority of shares to insiders, but it was neces-sary to settle for 51 percent preferential subscrip-tion of share ownership by managers and work-ers. When coupled with their vouchers, insidersended up owning on average 66 percent of theshares in privatized enterprises.

Even now, enterprises subject to massprivatization in Russia have done little restruc-turing. A fair number of them were probablyunviable under any form of ownership. Manywere producing goods that people did not needor want, and some were using resources notreadily shifted to the production of other items.However, it is also true that heavy insider controlby managers in an environment of weak or inef-fective shareholder oversight caused significantasset stripping. Delays in implementing a legal andgovernance framework, aggravated by oppositionfrom powerful insiders with a stake in weak cor-porate governance, led to poorly regulated capi-tal markets and weak enforcement of regulationsprotecting investors’ rights. These factors madethe phase of insider control and diffuse owner-ship very costly, raising questions about whatmight have happened with slower privatizationand a longer period of state control (box 7.1).

The Czech and Slovak Republics: Diverging Paths

The Czech and Slovak republics are an interest-ing contrast to Hungary and Poland, neighbor-ing countries where vouchers were used on amuch smaller scale (World Bank 1999a). Themost advanced industrial economy in the formerCommunist bloc, the former Czechoslovakia re-mained firmly in the grip of the Communist Partyright up to the Velvet Revolution. There was littleprice liberalization, little tolerance of small pri-vate enterprise, and little experimentation withworkers’ councils in enterprises. The fear of areversion to communism was perhaps strongerthan in any neighboring country.

Economic reformers believed that nothingcould be expected from a slow or evolutionaryapproach. The government apparatus could notbe turned into a force for positive change. Afast and massive transfer was needed to createnew owners who would support further


75

market reforms. Vouchers thus became the cor-nerstone of the reformers’ approach. Investmentfunds were created to consolidate the diffuseownership of vouchers and to diversify risk.Voucher holders could either bid directly forenterprise shares or buy shares in investmentfunds, which would then use the vouchers tobid for enterprise shares. The funds were ex-pected both to manage a diversified portfolioof shares (equity mutual funds) and to drive

enterprise restructuring through active gover-nance (venture funds and limited partnerships).

The Czech Republic implemented its massvoucher in two stages or “waves,” the first un-der the former Czechoslovak Federation and thesecond after the split. The government that ruledthe Slovak Republic shortly thereafter decidedto abandon the second wave of mass voucherprivatization in the belief that the voucher pro-gram would not be conducive to sound

TABLE 7.1.

Methods of Privatization of Medium-Sized and Large Enterprises

Country Direct sales Vouchers Management-employee buyout

CSB

Albania n.a. Secondary PrimaryBosnia and Herzegovina Secondary Primary n.a.Bulgaria Primary Secondary n.a.Croatia n.a. Secondary PrimaryCzech Republic Secondary Primary n.a.Estonia Primary Secondary n.a.Macedonia, FYR Secondary n.a. PrimaryHungary Primary n.a. SecondaryLatvia Primary Secondary n.a.Lithuania Secondary Primary n.a.Poland Primary n.a. SecondaryRomania Secondary n.a. PrimarySlovak Republic Primary Secondary n.a.Slovenia n.a. Secondary Primary

CISArmenia n.a. Primary SecondaryAzerbaijan Secondary Primary n.a.Belarus n.a. Secondary PrimaryGeorgia Secondary Primary n.a.Kazakhstan Primary Secondary n.a.Kyrgyz Republic n.a. Primary SecondaryMoldova Secondary Primary n.a.Russia Secondary Primary n.a.Tajikistan Primary Secondary n.a.Turkmenistan Secondary n.a. PrimaryUkraine Secondary n.a. PrimaryUzbekistan Secondary n.a. Primary

n.a. Not applicable.Source: European Bank for Reconstruction and Development data.


76

governance. It implemented the second wavethrough direct sales. But the sales were not con-ducted through open and transparent methods,and they favored politically connected parties.

The performance of Czech and Slovak enter-prises privatized through mass voucher methodswas disappointing, for three reasons. First, fewinvestment funds had the resources and the skills

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 criticismis that privatization was implemented too quickly and without first putting in place the institutional and legal framework that would have provided thechecks and balances needed for proper governance 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.

What would 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 massvoucher 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—“insiders”—the incentive to acquire majorityownership of their enterprises’ shares in order to obtain their support for privatization, with managers eventually becoming the most importantshareholders. However, because of a weak legal framework for corporate governance plus major political uncertainties in Russia, managers mayhave had an incentive to maximize their short-term capital gains by selling assets for personal gain (rather than keep the enterprise as a goingconcern and maximize future profits) and thus decapitalizing the enterprise at the expense of other (smaller) shareholders. This may have hadtwo negative effects: one concerning efficiency, the other, equity. First, there was a bias against keeping enterprises operating as a goingconcern. Second, regarding equity, wealth was redistributed from the rest of the population to the managers. Asset stripping may have in the endalso 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 adequateincentives 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 practicereceive the rights of owners. The Law on Cooperative Activities of 1988 allowed the formation of cooperatives (headed by the directors) withinthe enterprises. These cooperatives engaged in the most lucrative activities of the enterprises, while the liabilities remained with the state. As amatter 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 andworkers remain, in the interim, uncertain about how much they will benefit from the eventual privatization of their enterprise in the future. Thatuncertainty is a strong incentive for quick asset stripping. This was aggravated by the collapse of state institutions potentially capable ofexercising 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 wasconsidered at that time, but faced strong political resistance from managers and line ministries. But even if it had been feasible, it may not haveyielded better results in the particular circumstances of Russia. With weak minority shareholders’ rights, the incentives of managers to exploitother shareholders is even higher when the latter hold a larger share of the enterprise, unless they are concentrated and can organize to preventit. While too much share distribution to insiders may prevent strategic outsiders from coming in, too little may accelerate the incentives for assetstripping by these insiders. In the absence of encouragement for foreign strategic investors and with the collapse of state institutions, the choicebetween the mass privatization program and a longer period of state ownership, while developing institutions of corporate governance, is anopen question.

In summary, these and other historical questions can only be answered over the long run in the context of judging Russia’s overall transitionexperience, both economic and political. In contrast, most policy issues in this area today are more straightforward because of changed initialconditions and do not require a resolution of these questions.


77

to drive enterprise restructuring. Second, little wasdone to introduce an adequate regulatory frame-work governing enterprises, investment funds, andcapital market activities, particularly to protectminority shareholders’ rights. Third, surveillanceby creditors was weak because of delays in priva-tizing the largest banks and weak insolvency andcollateral legislation. Indeed, largely unreformedcommercial banks, operating in a climate of weakregulation, controlled many of the largest invest-ment funds. The funds also owned bank shares.This cross-ownership weakened the governancestructure and softened budget constraints further.All these problems resulted in significant “tun-neling”—extraction of assets by enterprise andfund managers (see box 3.2).

In sum, the lack of appropriate accompanyinginstitutional policies and lagging banking sectorreform made mass privatization unnecessarilycostly in equity, transparency, and microeconomicefficiency. It eventually contributed to a largebuildup of contingent fiscal liabilities in the insuf-ficiently reformed, state-dominated commercialbanks. By contrast, the best performers tended tobe enterprises and sectors sold to strategic inves-tors through transparent methods, including oneof the earliest cases of a sale to foreign investors:the Skoda-Volkswagen transaction.

In the Slovak Republic governance of enter-prises privatized by direct sales (during the secondwave) has not been significantly better than that ofthose privatized by voucher methods (during thefirst wave), despite the more concentrated owner-ship, as judged by the overall poor performance ofSlovak enterprises in the second half of the 1990s.The Slovak experience with direct sales provides asobering lesson and reinforces the conclusion thatconcentrated ownership is a necessary, but not suf-ficient condition for effective ownership. A pro-gram of direct sales to concentrated owners thatfollows open and transparent methods is muchmore likely to produce positive outcomes than aprogram of direct sales that favors politically con-nected parties and has the potential for corruption.

Hungary: Bold Moves, Big Gains

In Hungary the rapidly reforming CommunistParty led democratization by making a historic

compromise with the civil opposition, the ma-jority of which had never been underground.This more pragmatic Communist regime hadlargely abandoned central planning, allowing agrowing share of private ventures and speedingreforms in taxation, banking, foreign trade, andcorporate governance during the five years be-fore the political changes.

The new ruling elite in Hungary did not see areturn to communism as a threat. The problemlay with the managerial group that governed thou-sands of more independent, but still formally state-owned, enterprises—and stripped their assets.Transferring ownership titles through massprivatization to the population while leaving gov-ernance to these old managers would not be a so-lution. In addition, Hungary was the most highlyindebted country in the region on a per capita ba-sis, and the government was obliged to think aboutgenerating cash from privatization. These factorsled Hungary to base its privatization on sales tostrategic investors and to open the process entirelyto foreign investors. The full opening of theprivatization program to foreign capital was seenas a bold step that no other government (exceptEstonia) felt able to take at that time. Foreign capi-tal into Hungarian enterprises and banks broughtmuch needed investment, know-how, and compe-tition—the main reasons for Hungary’s goodgrowth in the second half of the 1990s.

By the mid 1990s Hungary had made sub-stantial progress in selling banks to reputableforeign strategic investors and in adopting a strictbanking regulation law and a strict bankruptcylaw. Hungary had also worked out much of thelarge stock of bad loans in the banking system.By contrast, this painful task had not even startedin the Czech and Slovak republics and wouldgain momentum only in the late 1990s.Hungary’s economic performance in the secondhalf of the 1990s shows that creditor disciplinematters as much as good ownership and gover-nance structures in driving economic restructur-ing and microeconomic efficiency.

Poland: Diversify and Discipline

As early as 1990 Poland planned to privatize amass of enterprises through vouchers. But politics


78

delayed the scheme, not any calculated assessmentby reformers that Poland’s promising economicand institutional situation would allow them thelatitude to proceed through other privatizationoptions. From the beginning Poland pursued a richmenu of privatization options, including the firstfive flotations on the Warsaw Stock Exchange andthe sale of assets and selected liabilities of some1,000 medium-sized enterprises through install-ment sales, a process described as “privatizationthrough liquidation.”

Poland’s success in privatization and its goodgrowth performance in the second half of the1990s was also a result of other important comple-mentary reforms. Like Hungary, Poland intro-duced hard budget constraints in the early stagesof the transition, the result of early efforts to re-structure and privatize the banks and address thebad loan problem. Poland reinforces the notionthat creditor discipline matters as much for enter-prise restructuring and performance as effectiveownership and corporate governance.

Slovenia: “Internally Privatized” Enterprises HaveContributed Little to Growth

Slovenia inherited the old Yugoslav concept ofsocial capital and social ownership, with the stateholding title to few industrial enterprises andbanks. Slovenia’s privatization (and that of allthe parts of former Yugoslavia) reflects this start-ing point. The Slovene program allocated themajority of shares to state-owned institutionalinvestors (the pension fund and a developmentfund) and to employees through several subsi-dized schemes. Few of the former socially ownedenterprises ended up in the hands of strategicinvestors. The evidence suggests that the “inter-nally privatized” enterprises have not flourished,though they seem to be doing slightly better thanthe socially owned enterprises that were notprivatized. Most of the growth in industry andservices is explained by new entrants, the recipi-ents of most foreign direct investment.

Other Countries Also Chose Vouchers

Several other CIS countries relied heavily on massprivatization through vouchers. Kazakhstan used

investment funds as vehicles to achieve a con-solidation of shares, requiring that all vouchersbe placed with funds. Moreover, beforeprivatization, the government established sectorholding companies that allowed the state to re-tain a 39 percent residual share in all enterprisesbeing privatized. A legal framework governingthese funds was put together, but was then poorlyenforced. The upshot was a dilution of the valueof shares held by the initial holders of vouchers.Other variants included state and private enter-prise funds in Romania and divestiture by leas-ing to worker cooperatives in Ukraine.

Were Vouchers a Mistake and Other“What Ifs”

Each country pursued its own strategy; therewas no homogenous approach driven by

blueprints. This diversity and considerable dis-appointment raise many questions about the ap-propriateness of strategies. The most fundamen-tal question has two parts: was mass voucherprivatization a mistake? Could that have beenforeseen? Would countries that went throughmass voucher schemes, with disappointing re-sults, have been better off keeping their enter-prises in state hands while trying to accelerateeconomic reform and creating an institutionaland legal framework to attract reputable con-centrated investors?

This approach has been successful in China,which has enjoyed sustained high growth rateswithout, until recently, allowing much in the wayof formal transfer of state ownership. There isconsiderable debate over how China achieved this.Box 4.1 argues that initial conditions in Chinawere different enough to allow growth to be fu-eled by a massive entry of new enterprises whilethe state and subnational governments couldmonitor state enterprises to prevent egregious as-set stripping by managers or their allies. But mostof the CIS started the transition with a major po-litical transformation in parallel with an economicone. This weakened or eliminated many of thepolicy and disciplinary mechanisms required toreplicate this approach, and the recentralizationof political power needed to achieve it would nothave been feasible given the political trends.


79

Answering these broad questions requires ahistorical counterfactual against which to com-pare the mass privatization option (see box 7.1).If political centralization or recentralization wasnot an option, what mechanisms were availableto improve governance of state-owned enter-prises? Would hardening budget constraints—a key accompanying policy that influences en-terprise performance—have been easier understate ownership? Could market and legal insti-tutions supporting property rights evolve in anenvironment of state ownership? The cases ofBelarus (see box 4.3) and Uzbekistan do notsupport this notion.

Only time will permit a fuller assessmentof these fundamental questions. But many use-ful and practical lessons still emerge from moremodest questions. For example, to what extentcan the outcomes of privatization be influencedby the complementary policies generally underthe control of policymakers? In the formerCzechoslovakia, mass voucher privatizationwould have had a better chance of producingmore restructuring and less corruption if thelegal framework governing companies, invest-ments funds, and capital market activities hadbeen sharply enhanced and enforced from thevery beginning. Earlier efforts at privatizingbanks and strengthening creditor rights throughimproved insolvency and collateral legislationmight have aided overall restructuring as well.These reforms would have required intense—but manageable—effort from the Czech andSlovak authorities. Moreover, the efforts prob-ably would have succeeded in attracting largerflows of foreign capital, facilitating secondarytrading, and consolidating ownership by stra-tegic investors.

In Kazakhstan the inability to vigorously en-force the legal framework governing investmentfunds was a major issue. Russia suffered similarweaknesses in the way it implementedprivatization. Managers and local officials try-ing to prevent competition dominated many ofthe auctions to voucher holders. A weak legalframework to protect minority shareholders in-hibited outsider participation in secondary trad-ing, guaranteeing that the managers would con-tinue to control the enterprises.

Summarizing Lessons

This experience suggests the followingagenda:

• Privatization should be part of an overallstrategy of discipline and encouragement.

• Small enterprises under state ownership (gen-erally enterprises with fewer than 50 employ-ees) should be sold quickly and directly tonew owners through an open and competi-tive auction, without restrictions on who maybid for the shares.

• Medium-size and large enterprises shouldtarget sales to strategic outside investors.With a concentrated, controlling interest, theywill have a clear stake to best use the enter-prises’ assets. Although several transactionmethods may be used, including negotiatedsales, this can be brought about most effec-tively through competitive “case-by-case”methods, more deliberative than voucherschemes or rapid, small auctions. They useindependent financial advisors who both pre-pare the enterprise for sale and act as salesagents on behalf of the state.

• Investor protection should be enshrined inthe legal system and enforced, covering rulesto protect minority shareholders; rulesagainst insider dealings and conflicts of in-terest; creditor surveillance accounting, au-diting, and disclosure standards; and take-over, insolvency, and collateral legislation.When court enforcement of contracts isweak, these provisions should be supple-mented by a stock market regulation for fi-nancial intermediaries, such as investmentfunds and brokers, who then have an incen-tive to ensure compliance by other partici-pants in financial markets. This will set thestage for privatization of future medium-sizeand large enterprises. It will also improvethe performance of existing privatized en-terprises by assisting transparent consolida-tion of shares where ownership is diffuse. Itwill also facilitate bank- and nonbank-basedfinancial intermediation.

• Privatization should be accompanied by in-creasing competition in the market for theproducts sold by the enterprise in question


80

and vigorously enforced by the competitionpolicy authority. This can discipline manag-ers in an environment where corporate gov-ernance is weak.

• The cash flow and property rights of thestate should be clarified for enterprises inwhich the state continues to hold an owner-ship stake.

• Divesting enterprises in sectors characterizedby natural monopoly or oligopoly (where av-erage production costs decline continuously asscale increases and the market and society arebest served by one or a few enterprises) mustproceed with great caution, if at all. Advancesin technology have made such sectors increas-ingly rare. But where they exist—as in localdistribution of natural gas—an efficient regu-latory regime that protects the public interestis a prerequisite, lest divestiture transform aninefficient public monopoly into a poorly regu-lated or nonregulated private monopoly.

In Central and Eastern Europe and the Baltics(except the Federal Republic of Yugoslavia), SMEshave been substantially privatized. Voucherprivatization or competitive cash auctions are notrelevant options. Bulgaria, Croatia, and Roma-nia—with large commercial enterprises still to beprivatized—should implement (and, indeed, areimplementing) transparent and competitive case-by-case methods to mobilize strategic investors.Progress in the European Union accession willmake this more likely to succeed. The FederalRepublic of Yugoslavia, now privatizing medium-size enterprises through competitive auctions, willstart the privatization of large enterprises throughcase-by-case methods. A major positive featureof these methods is to restrict the distribution ofsubsidized shares to insiders to a maximum of 30percent of shares, thus making the majority ofshares available for strategic investors.

In such CIS countries as Armenia, Georgia,the Kyrgyz Republic, Russia, and Ukraine, andwhere privatization has been important duringthe past decade, the most critical steps now are

to improve the legal framework, assist a trans-parent consolidation of shares, and develop in-stitutions to help monitor managerial behavior.The remaining privatization agenda must rely ontransparent and competitive case-by-caseprivatization, including the transfer of residualstate ownership.

In countries such as Belarus, Turkmenistan,and Uzbekistan, where most medium-size andlarge enterprises remain in state hands, a ma-jor challenge remains. In Belarus andUzbekistan, as in China, strong central con-trols have prevented egregious asset strippingin state-owned enterprises. However, there hasbeen little restructuring because these enter-prises have been protected through the traderegime and the special allocation of foreign ex-change and credit at subsidized prices. Foreigndirect investment and entry of new enterprisesremain heavily controlled. The challenges inthese countries, when they decide to reform,will be to liberalize, encourage new entry, andassemble a legal framework (for private prop-erty and contract enforcement) and market in-stitutions to monitor managers. Small enter-prises should be quickly privatized throughcompetitive auctions, and medium-size enter-prises through case-by-case methods, with amajority share sold to outsiders. Very large en-terprises should be privatized only when a clearstrategic investor is identified.

Note

1. Capital market regulations in most OECD coun-tries require that a major shareholder or group of con-nected shareholders acquiring 30 percent of the sharesin a company offer to buy out minority shareholdersat the assessed market value. Full implementation ofthis regulation alone would have considerably slowedconsolidation of ownership in the Czech Republic inthe 1990s. Other regulations restricting insider dealsand transactions with connected parties, enhancingdisclosure, and ensuring an integrated pricing mecha-nism would also have slowed the consolidation ofownership. For a review of progress with privatization,see EBRD (2000).

81

8Supportive Social Policies

How can social policies support a strategy of discipline and encouragement? By targetingsocial safety nets to the most vulnerable, such as those affected by the increase in utilityprices 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 ashousing, kindergartens, and clinics, to permit enterprise restructuring to go ahead. And by reformingexpenditures on education and health to allow workers to acquire skills more adapted to new marketrealities 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 tostabilize 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 isto restructure social sector expenditures to make them fiscally more affordable. In the low-incomeCIS 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 theadvanced reformers of Central Europe and the Baltics the ratio of consolidated public expendituresto 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 thenew costs of complying with European Union directives, these countries will need to improve theirefficiency 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 thanin the rest of the CIS. Several of them, facing unsustainable public debt, need an international workout toreduce debt service and domestic efforts to increase tax revenue to maintain social expenditures.

Reforming 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


82

Slovenia public pension expenditures have climbedto more than 13 percent of GDP—twice thepretransition level and not sustainable. The fiscalburden of pension systems increased in the tran-sition partly because benefits eased the social costsin its early years. Indeed, in most CSB countriesthe incidence of poverty for households headedby pensioners is much lower than for other socio-economic groups. However, the aging of the popu-lation and the rising payroll taxes needed to fi-nance more generous benefits undermine thefinancial viability of pension systems and threatenemployment creation. The challenge is to makepension systems more sustainable without under-mining the socially desirable goal of providingadequate income for retired workers.

Large pay-as-you-go pension systems, withtheir attendant large unfunded liabilities, burdenpublic finances and savings in the long term.Higher contribution rates by themselves do notsolve the problem because they raise labor costs,shift incentives toward informal market employ-ment, and undermine job creation, particularlyby new enterprises. In addition, the opportunitiesfor using contractual savings for capital marketdevelopment are lost. Most countries have begunto reform their pension systems by tightening thelink between contributions and benefits, shiftingto notionally defined contribution schemes, rais-ing retirement ages, reducing replacement rates,and changing pension-indexation formulas. Inaddition, several countries in Central and East-ern Europe are taking steps to implementmultipillar pension schemes (table 8.1).

A multipillar pension system allows people todiversify risk across countries, regions, and assets.It provides for a portion of the mandatory contri-butions to the public pension system to fund theindividual accounts of each worker. These contri-butions are managed and invested by private insti-tutions, and pensions are paid according to a de-fined contribution. The eventual pension systemcomprises a downsized pay-as-you-go scheme (thefirst pillar), a benefit from a fully funded scheme(the second), and personal savings (the third).

A key element of these reforms is making ex-plicit at least part of the implicit debt of pensionsystems, which then has to be financed. (The sizeof the full implicit debt varies from country to

country. In Central and Eastern Europe it rangesfrom 120 to 250 percent of GDP.) There are threepossible sources of financing. The first is to limitfirst pillar expenditures to create savings in thestate pension schemes and reduce future liabili-ties. The second is to increase payroll taxes andcontributions. The third is to use fiscal resourcesfrom debt instruments or privatization proceeds.Given the magnitude of the financing needs andthe fact that governments may have largely ex-hausted the first two sources, the resources forpaying obligations to current pensioners will haveto come from the budget.

The up-front fiscal costs of moving to amultipillar system are large. In reforming countriesa second pillar financed by a contribution rate of 6percent of gross wages, as in Hungary, would re-quire resources equal to around 1.9 percent of GDPannually in the first years. Other countries prepar-ing to establish such a scheme—Bulgaria, Croatia,Estonia, and FYR Macedonia—will also need toplan and budget for these costs in ways consistentwith preserving macroeconomic stability. The sizeof the fiscal costs depends on the design of the sec-ond pillar—its size and choice of cohorts—and onthe pace that governments move along this path.However, the institutional requirements of intro-ducing multipillar systems are demanding. Finan-cial markets have to be adequately developed, andgovernments have to regulate and supervise funds.

The CIS typically spends about half what theCSB spends on pensions. But in many cases eventhat spending is fiscally unsustainable, given poortax compliance and high ratios of dependents tocontributors. Pension systems in the CIS have gen-erally done a much poorer job of keeping the eld-erly out of poverty. They need to undertake thebasic reforms to put their pay-as-you-go systemson sounder financial footing. Some low-incomecountries—such as Georgia, where resources areconstrained and tax compliance especially poor—may need to move to a flat benefit structure toprotect the poorest elderly until fiscal conditionspermit a move to a more differentiated structure.

Implementation of these reforms will face po-litical opposition and administrative weakness.For example, flat benefits to the most needy maynot enjoy broad-based support, as Georgia’s re-cent experience showed. Coverage needs to be

Supportive Social Policies

83

TABLE 8.1.

Reform 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 withsupplemented by up-front fiscal costindicator 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 towardand Ukraine) means tested only based benefit targeted social

where local assistanceinstitutions arestrong)

• 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 allAsian 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-tinued and adequately funded. Making the de-sign and implementation of reforms more diffi-cult are weak actuarial capacity to forecastexpenditures and revenues, a lack of auditing ofpension funds, and problems in collection. Aswages begin to rise, tax collection improves, andinstitutional capacity develops, countries canmove toward a multipillar pension system.

Social Assistance Should Protect Childrenand the Most Destitute, Adding More asBudgets Allow

With the transition to market, the guaran-teed employment, retirement security, and

consumer subsidies of the socialist system

diminished considerably. In addition, the real in-comes of households fell. Although the CentralEuropean countries have spent the bulk of theirsocial protection budgets on pensions, these coun-tries allocated somewhat more resources to so-cial assistance spending to address rising poverty.The CIS generally spent less on social assistance,while indirectly protecting employment by not im-posing hard budget constraints on old enterprisesand by maintaining subsidies on housing and utili-ties. The subsidies on utilities have most often beenmet by forcing the utilities to defer maintenance,or by expecting enterprises to finance them evenwhere housing has been divested. Despite the ex-pansion of social assistance, most programs, in-cluding the subsidies on utilities, have not reachedgreat numbers of the poor.


84

The key objectives for social assistance pro-grams—and probably all that the poorest coun-tries can afford—are to help the most destituteand to ensure that children’s mental and physicaldevelopment is not impaired. In addition to en-suring that young children receive essential pre-ventive health care and can afford to attend school,many countries also face the challenge of offeringalternatives to institutions for elderly people un-able to live on their own; adults with physical andmental disabilities; and children disadvantaged bypoverty, ethnicity, or disability. Some countries areintroducing community-based services such ashome and day care and special educational pro-grams for children with disabilities, adapting ap-proaches that have been successful in WesternEurope and the United States.

In addition, to facilitate the enterprise restruc-turing critical to sustained growth, countries maywish to offer some protection to those hurt byrestructuring. The best way of doing so is to re-move barriers to entry of new enterprises. Assis-tance should be targeted to workers whose skillsand experience make them the least likely to beemployed by new enterprises.

For low-income countries with limited re-sources and pervasive poverty, targeting cash ben-efits—with a possible increase in the resource en-velope in some countries in the Caucasus andCentral Asia, where spending is low—may haveto be improved through geographical targeting,community-based identification, or other indica-tors. Given the pervasiveness of the informaleconomy and self-employment, targeting will con-tinue to be imperfect. There may be some scopefor self-targeting through some form of publicworks scheme. There also is considerable scopefor improving the targeting of utility and housingsubsidies. Rather than offer across-the- boardprice subsidies, some countries have introduced“lifeline” tariffs for utility services with meteredconsumption, in which the price subsidy is re-stricted to the initial block of consumption, calledthe basic need level. Lifeline tariffs are easier toadminister than income-tested schemes and areless distortionary than generalized price subsidies.

Higher-income countries may be able to af-ford a more generous safety net. Targetingthrough some forms of means testing may be

less prone to error in these economies, to theextent that the tests are more formalized. Un-employment insurance may also be more fea-sible in these countries, with structural unem-ployment falling and capacity to implement atrue unemployment insurance program grow-ing. But more generous provision of benefitscarries the risk of creating a culture of depen-dency and reducing incentives of workers to findemployment—added, of course, to the ever-present problem of becoming a major fiscal bur-den and a drag on growth.

Severe Cuts Have Compromised the Quality ofEducation

Some of the fiscal adjustment in the CIS hasbeen at the cost of severe cuts in education,

inevitable given the plunge in GDP for all theCIS countries (World Bank 2000b). Publicspending on education ranges from less than 2percent of GDP for Armenia and Georgia toalmost 8 percent of GDP for Uzbekistan. Theaverage for OECD countries—with 10 times theGDP per capita—is about 5 percent of GDP.Several countries have maintained reasonablespending on education relative to their GDP.The challenge is to ensure that resources areallocated to best use—across levels, betweenwage and nonwage expenditures, and acrossregions within each country. But in the poorestcountries of the Caucasus and Central Asiaspending cuts have compromised the ability ofthe education system to prepare students for theemerging market economy (figure 8.1).

The cuts have affected opportunities, access,and coverage (World Bank 2000b). Falling edu-cation budgets and protection of employment inthe sector are squeezing expenditures on text-books, school supplies, and school maintenance.In addition, poorer parts of many countries beara disproportionate share of the adjustment. InGeorgia 43 percent of primary and secondaryschools in urban areas got textbooks for all chil-dren, compared with 27 percent in poorer ruralareas. In the Russian Federation richer regionshave spent more on education, while poorer re-gions have struggled to maintain the basic re-quirements (World Bank 1999b).


85

FIGURE 8.1.

Public Expenditures on Education in TransitionEconomies, 1998


Hungary

Poland


Slovak Republic

LatviaLithuania

Croatia

Kyrgyz RepublicKazakhstan

Bulgaria


Moldova

Romania

ArmeniaAlbania

Russian Federation

Ukraine

Azerbaijan

Uzbekistan

Tajikistan

Belarus

Turkmenistan

100806040200

Percentage of GDP

By not spending enough to meet the basicneeds of the school system, the state effectivelyshifts educational costs to families. Families con-tribute to school maintenance and educationsupplies. In some countries they are asked tosupplement the salaries of education personnelthrough under-the-table payments and varioustutoring schemes—if they want their children topass. At the same time many countries have beenslow to implement cost recovery measures forhigher education, which tends to favor better-off families.

Poorer countries need to take three sets ofmeasures. The first is to ensure universal basiceducation, of particular importance to Armenia,Azerbaijan, Georgia, and Tajikistan, where bud-getary resources for basic education do not pro-vide universal coverage. The second is to ensurethat expansion of tertiary education does notcome at the expense of basic secondary educa-tion. In some countries higher education enroll-ments are growing—appropriately reflecting thehuman capital demands of these emerging econo-mies. This risks absorbing larger shares of edu-cation budgets, unless greater cost recovery,lower unit costs, and more encouragement ofprivate financing and delivery of higher educa-tion services follow.

The third set of measures is to shift resourcesfrom personnel (and energy) into repairingschools and providing adequate educationalmaterial. As materials and maintenance expen-ditures were cut and student enrollments fell, thenumber of teachers rose sharply. In Russia everyregion increased its number of teachers, for anoverall increase of 25 percent between 1989 and1996 (World Bank 1999b). In Central Asia in-creases ranged up to 25 percent (Klugman 1999).Student-teacher ratios are typically low for mostof these countries and can be increased withoutcompromising teaching quality or learning out-comes (table 8.2). Countries will need to ratio-nalize and consolidate their school infrastructure,which is often overdimensioned due to fallingbirth rates and extremely low student-teacherratios. School consolidation would also enablecountries to reduce wage costs. Using OECDstandards as a benchmark, up to a third of theteaching labor force can be reduced. That would

TABLE 8.2.

Student-Teacher Ratios in Basic Education, 1990 and1997(percent)

Countries 1990 1997

Armenia 11.7 8.7Azerbaijan 10.5 9.9Belarus 11.8 10.5Russian Federation 14.0 11.9Turkmenistan 14.0 13.4

Source: UNICEF–ICDC TransMONEE (1999).


86

free budgetary resources to better compensateremaining teachers, to repair schools, and to pro-vide instructional materials.

In addition, energy costs for education needto be reduced. In some CIS countries energy ex-penditures consume between 30 and 50 percentof education expenditures, reflecting the fast risein energy prices in the past 10 years. In the me-dium to long run, energy savings will be realizedas new schools replace old ones. Short-term mea-sures—insulating school walls, double-glazingwindows, and installing meters—can also helpreduce energy bills, but these will need up-frontinvestments from the budget.

Containing Costs Will Make Health CareAffordable for Those Who Need It Most

The public health achievements of the socialistera are being undermined. Many countries

are not spending enough on public healthmeasures to confront the growing threat of HIV/AIDS and drug-resistant tuberculosis.

Inadequate funding has led to under-the-tablepayments. Unpaid or underpaid health workerstap patients for side payments, including pay-ments for drugs and other items in short supply.This is financially onerous for poor families whohave to sell assets or borrow to finance theirhealth care. In the Kyrgyz Republic a third of allpatients seeking inpatient care had to borrowmoney. Similar figures are reported for Georgia,Tajikistan, and Ukraine.

Often the inability to pay means that pa-tients forgo health care. In Russia data for 1997suggest that 41 percent of all Russian patientscould not afford to purchase required drugs,and 11 percent could not afford any kind ofmedical treatment (World Bank 2000b). Simi-larly, 37 percent of pregnant women inTajikistan did not seek prenatal care because itwas unaffordable, and almost a third of birthsoccurred at home, a break from past practicesof hospital births. In several countries a pri-vately financed, unregulated system of healthcare is in a public shell.

In some countries, notably in the Caucasus andCentral Asia, public expenditures are inadequate(figure 8.2). However, other countries are

FIGURE 8.2.

Health Expenditures, 1998

a. Data are for 1997.b. Data are for 1996.c. Data are for 1999.Source: The World Bank’s World Development Indicators

database.

HungaryPoland

Czech Republic

EstoniaSlovak Republic

Latvia

LithuaniaCroatiaa

Kyrgyz RepublicKazakhstan

Bulgaria

Macedonia, FYR

Georgia

Moldovaa

RomaniabArmenia

Albaniaa

Russia

Ukraine

Azerbaijana

Uzbekistana

Tajikistanc

Belarus

Turkmenistana

1086420

Percentage of GDP

Portugal

SpainFinland

United KingdomNetherlands

Canada

United StatesaSloveniaFranceaSweden

Germany


87

spending significant public resources without get-ting the benefits of quality health care. The chal-lenges facing most transition economies are simi-lar to those facing education: the rationalizationof systems with an excess of personnel and

facilities and reallocation of savings on comple-mentary inputs. To reallocate expenditures mosteffectively, however, governments need to decidewhat kind of health care system they want, in-cluding who provides what and who pays for what.

Part 3

The Political Economyof Discipline andEncouragement

91

Why have some governments been able to enact policies to discipline the state sector andencourage new enterprises to create a foundation for sustainable growth? Why have oth-ers maintained far less effective strategies of protecting inefficient enterprises and discour-

aging new entry at considerable social cost? Can these government policy choices be systematicallyrelated to particular institutional characteristics of political systems in transition? Part 3 will try toanswer these questions.

To develop a better understanding of the political economy of reform in transition economies, weextend 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 termas a result of higher unemployment, higher prices, and lower provision of subsidized social services byenterprises. In contrast, the gains associated with the policies of encouragement accrue primarily overthe 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 adjustmentcosts from discipline in the short term for the mere promise of future gains as encouragement promotesgreater investment and growth. How can governments assure the public that those future gains willmaterialize? What guarantees that future governments will not backtrack on reforms before the prom-ised gains materialize? What if the promised gains are directed to groups closely tied to existing elites andnot 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 policiesthat impose clear costs in the short term for the promise of long-term gains inherently subject to highrisks. To win broad support, the government must convince the public that the government will beable 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 (Przeworski1991). Recognizing the different time horizons of costs and benefits and the related risks that might

9The Winners and Losers

from Discipline andEncouragement


92

lead rational individuals to discount those ben-efits at the start of reform is central to the politi-cal economy of market-oriented transitions. Thegovernment must be able to credibly commit toshort-term losers that they will reap the gainsfrom reforms over the long term.

At the same time, a decade of transition ex-perience has shown that some groups realize sub-stantial gains from the no man’s land betweenthe command system and the market economy(Hellman 1998). Uncertainty about propertyrights before privatization allows insider man-agers to tunnel assets from nominally state-owned enterprise to newly created spinoffs un-der managers’ personal control. Partiallyliberalized prices spur arbitrage opportunitiesbetween fixed-price and market-price sectors thatcan generate enormous gains. Incomplete tradeliberalization creates highly profitable monopolyrents, especially in economies rich in natural re-sources. These opportunities to take advantageof the distortions of a partially reformedeconomy are available only to the select few,namely those with control over nominally state-owned assets and those with close ties to politi-cians able to award such advantages. As a re-sult, these gains are highly concentrated.

In theory such gains should be short-term,because as transition progresses many of thesetemporary economic distortions will be eliminatedor competition will arise to dissipate the rents.But experience shows that these short-term win-ners of partial reform can convert a small shareof their gains into political influence that can beused to restrict entry, undermine competition, andpreserve the very distortions that generate theserents. Such constituencies seek to freeze reforminto an equilibrium of liberalization without dis-cipline and selective encouragement, producing ahighly unequal pattern of costs and benefits ofmarket-oriented transition over the long term.

Examples of how the short-term winners ofpartial reform prevent further reforms that wouldimpose discipline and encourage new marketentry and competition are well known through-out the region. Enterprise insiders who gainedminority stakes in privatized enterprises haveopposed improvements in corporate governanceand the security of property rights that would

limit insiders’ ability to tunnel assets abroad.New banks created in the liberalization of finan-cial markets have fought to keep soft govern-ment credits for their enterprise clients, whichcould be recycled in volatile local bond and for-eign exchange markets. Oil and gas exporterswho gained from the opening of foreign trade inpartially liberalized markets have struggled tobuild entry barriers to prevent competition fromdissipating their rents. The ability of these groupsto preserve their extraordinary gains is based ontheir capacity to influence the political process,and in the most extreme cases, to capture keyinstitutions of the state, highlighting the criticalinteraction between politics and economics intransition (Hellman, Jones, and Kaufmann 2000;World Bank 2000c).

Who Wins and Who Loses?

The complex dynamics of the politicaleconomy of reform can be expressed graphi-

cally by tracing the paths of winners and loserswith respect to different levels of reform in thediscipline and encouragement framework. A typi-cal economy at the start of the transition can bedivided into three basic constituencies:

• State sector workers. These are workers fromstate-owned enterprises without the skills tobecome new entrants in the competitive mar-ket. They face significant losses initially be-cause of discipline (unemployment, price in-creases) and are unlikely to realize any gainsfrom encouragement.

• Potential new entrants. These are workersand new entrepreneurs, originally fromstate-owned enterprises, who have the skillsto become new entrants in the competitivemarket. They face initial losses from disci-pline as they adjust to the decline in the statesector. However, they are likely to see gainsfrom entry into the market if encouragementis effectively implemented and sustained.

• Insiders and oligarchs. These are actors whobegin the transition with substantial de factocontrol rights over state assets and close tiesto the political elite inherited from theprevious command system. Insiders and oli-

The Winners and Losers from Discipline and Encouragement

93

garchs benefit immediately from liberaliza-tion and privatization because they can con-vert their existing control over state assetsinto substantial gains. Moreover, insidersand oligarchs can reap further gains fromrent seeking, arbitrage, and asset strippingif liberalization and privatization are notcombined with discipline and encourage-ment. As discipline is imposed and furtherreforms encourage competition from newentrants and the rule of law, these initialgains are dissipated unless new barriers toentry can be erected.

Figure 9.1 presents a stylized depiction ofthe income gains and losses for each of thesethree groups under different levels of economicreform. The state sector workers face a drop inincome as a result of sector downsizing, withlittle hope of substantial recovery as reformadvances. The potential new entrants face a clas-sic J-curve pattern of income, with adjustmentcosts at low levels of reform as they exit the statesector and gains realized only once enoughprogress has been made in institutional reforms

to promote and support new entry into the com-petitive market. The oligarchs and insiders, bycontrast, face an inverted U-curve income pat-tern. Their concentrated gains in the early stagesof reform—associated with opportunities for ar-bitrage, rent seeking, and tunneling—are dissi-pated as further reforms increase competitionand market entry.

For all these constituencies, gains and lossesdepend on how radical the first move in the re-form process is at the start of the transition.The more radical this move, the greater the ini-tial adjustment costs to both state sector work-ers and potential new entrants. Yet new entrantsshould begin to see greater gains at an earlierpoint in the transition if such reforms lead torapid development of the institutions that en-courage entry and competition. For the oli-garchs and insiders, radical reforms generatefewer distortions and imbalances for them toextract rents and strip assets. So such reformsreduce the high concentration of initial gains tothe oligarchs and insiders.

By contrast, less radical reform programs—liberalization and privatization with weak disci-

R0

Oligarchs and insiders

R = Extent of reforms

State sector workers

New entrants

R1 R2

Income gains

0

+

–

FIGURE 9.1.

Winners and Losers from Reform

Note: R0 = no reforms; R1 = point at which income gains of oligarchs and insiders are maximized; R2 = level of reformsthat allows the winners of reforms beyond R1 (new entrants) to compensate for or exercise enough political pressure toneutralize the resistance of oligarchs, insiders, and state sector workers.

Source: Authors.


94

pline and minimal measures to support compe-tition—have the opposite effect. Such partial re-form programs at the start of the transition gen-erate lower initial adjustment costs for both statesector workers and potential new entrants, asstate sector downsizing is limited and the flowof subsidies continues. Oligarchs and insidersenjoy the highest gains from liberalization andprivatization without discipline and encourage-ment.

Given these patterns of gains and losses, eachconstituency prefers a different combination ofreforms. State sector workers prefer the statusquo (R0) and reject all reforms. Oligarchs andinsiders prefer to begin with a partial reform andsustain the reform process through R1, the pointwhere their gains are maximized and beyondwhich implementation of policies of disciplineand encouragement threatens to undermine gainsfrom rent seeking and tunneling. For new en-trants, the process offers sacrifices in the begin-ning for the promise of gains when the reformsare advanced enough to create an environmentconducive to new entry and competition.

For a government to secure the support ofthe potential new entrants for radical reforms atthe start of the transition, the government mustbe able to make a credible commitment that thereforms will be continued until at least R2. Butthe credibility of that commitment will dependon the strength of the oligarchs and insiders, whohave an incentive to invest a share of their initialgains in capturing the state to stop the reformprocess at R1. Thus at the start of transition, thegreater the risk that oligarchs and insiders willcapture the state in the future, the less likely thatpotential new entrants will support a radical re-form program at the outset.

Where the risk of capture by oligarchs andinsiders is high, potential new entrants and statesector workers have an incentive to reject reformsor to accept partial reform programs that reducethe initial adjustment costs. Yet it is precisely suchpartial reforms that make state capture by oli-garchs and insiders a self-fulfilling prophecy, asthese reforms maximize opportunities for rentseeking and theft. This has led to a so-called par-tial reform paradox in many transition econo-mies: governments that lack credibility and are

highly susceptible to state capture cause poten-tial new entrants at the outset of transition tosubstantially discount the potential gains fromany proposed radical economic reforms, leadingthem to support partial reforms that offer lowerinitial costs—even though these partial reformsare more likely to lead eventually to barriers toentry.

The risk of ending up at a low level of re-form (R1) is indeed high. Such partial reforms—liberalization without discipline and with lim-ited encouragement of new entry—are the resultof the joint pressure of oligarchs and state work-ers to prevent further reforms while the gains bynew entrants are not high enough to allow themto exercise enough pressure for more compre-hensive return. Only a minimum commitmentof advancing reforms to at least R2 will generateenough support—where R2 is the level of reformthat allows the winners of reforms beyond R1

(new entrants) to compensate for or exerciseenough political pressure to neutralize the resis-tance of oligarchs, insiders, and state workers.

The political problem posed by initial win-ners from the transition was not foreseen in theearly literature on transition. Few recognized thatoligarchs and insiders would be able to stall thetransition in a state of partial reforms. Yet thishas proven to be one of the most serious ob-stacles to economic reform, particularly in manyCIS countries.

The Government Must Be Credible and Able toConstrain Oligarchs and Insiders

Partial reform—liberalization without disci-pline and selective encouragement—can be

a stable equilibrium of reform if government,at the outset of the transition, lacks the cred-ibility to promise that narrow groups with closelinks to the political elite will not capture thereform process. Public support for radical re-forms, therefore, depends on perceptions ofgovernment credibility.

By recognizing that different combinationsof reforms produce different configurations ofwinners and losers, the discipline and encour-agement framework suggests two political chal-lenges in promoting economic reform:

The Winners and Losers from Discipline and Encouragement

95

• Securing up-front support from potential newentrants for comprehensive reforms

• Preventing early winners of liberalization andprivatization from trying to sustain a partiallyreformed equilibrium of rent seeking andtheft by undermining further reforms.

To meet these challenges, governments musthave the capacity to project credibility to poten-tial new entrants and constrain the oligarchs andinsiders.

What influences the capacity of governmentsin transition economies to make credible com-mitments that the promised gains from reformwill not be expropriated by early winners of re-form? Credibility and constraint are rooted inthe nature of political institutions, which areshaped by the cultural and historical legacies thatguided the exit from communism. Political in-stitutions designed with participation by com-peting groups to foster political contestabilitywithin an agreed set of rules are less likely to becaptured by a corrupt political elite or narrowset of interest groups. The rigors of politicalcompetition increase the costs to politicians ofpleasing narrow constituencies and increase thelikelihood that broader interests will be repre-sented in government over time.

In contrast, political systems designed to con-centrate power and restrict contestability are at

greater risk of being captured by small, power-ful interests. In such systems politicians are heldaccountable to a narrower range of groups, in-creasing the probability that those with accessto political power will expropriate or concen-trate the gains from government policymaking.To prevent the early winners from holding theeconomy in a partially reformed equilibrium, thepolitical system must constrain the power of anynarrow group to capture the state. Expandingthe range of social groups competing for influ-ence over policymaking increases the costs topoliticians of skewing reform in the interests ofa single group. So the government’s capacity tomake a credible commitment to a wide range ofgroups is much greater in political systems thatpromote competition and contestability.

These differences in political systems areshaped by a broad range of factors often looselyreferred to as “initial conditions,” which incor-porate historical, cultural, geographical, andother variables that shape the transition path(de Melo and others 1997). But understandinghow different political systems shape the con-figuration and choices of the winners and los-ers in reform provides important insights intothe relationship between democracy andmarket-oriented reform.

97

To determine the effects of different types of political systems on the capacity of governments toadopt and sustain economic reforms, we first differentiate the range of political systems acrossthe transition economies. One of the most important features of this variation has been the

political contestability in the new regimes, that is, the extent to which key decisions of the politicalprocess—such as choosing political leaders, adopting laws, and making binding policy decisions—aresubject 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 toparticipate freely in the political process and to express one’s views. The extent of such rights canbe measured by indicators of political rights and civil liberties.

• Veto points. The clearest institutional manifestation of political contestability is the right to vetopolitical 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 shapethe 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 regionalcleavages, 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 theirrapid 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 awidely accepted system of annual ratings of political and civil liberties (figure 10.1). On the basis ofthese ratings, the following country groups can be developed:

10Classifying Political Systems

in Transition


98

• Competitive democracies have from the startof transition maintained both a high level ofpolitical rights to compete in multiparty demo-cratic elections and an extensive range of civilliberties. They include the Czech Republic,

Estonia, Hungary, Latvia, Lithuania, Poland,and Slovenia.

• Concentrated political regimes conduct mul-tiparty elections, but for some period of thetransition they have either curtailed full rights

FIGURE 10.1.

Classifying Political Systems in Transition Economies, 1990–99

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).

HungaryPoland

Czech Republic

Estonia

Slovak Republic

Latvia

Lithuania

Croatia

Kyrgyz Republic

Kazakhstan

Bulgaria

Macedonia, FYR

Georgia

Moldova

Romania

ArmeniaAlbania

RussiaUkraine

Azerbaijan

Uzbekistan

Tajikistan

Belarus

Turkmenistan

86420

Average rating of political and civil liberties

Noncompetitive political regimes

War-torn regimes

Concentrated political regimes

Competitive democracies

Average

Average

Average

Average

Slovenia

Classifying Political Systems in Transition

99

to participate in those elections or otherwiselimited political competition through con-straints on civil liberties. The result has been aconcentration of political power, often in theexecutive branch of government, within theframework of a multiparty electoral system.These countries include Bulgaria, Croatia, theKyrgyz Republic, FYR Macedonia, Moldova,Romania, the Russian Federation, the SlovakRepublic, and Ukraine. The countries span awide range of political systems. Bulgaria andthe Slovak Republic have a high degree ofpolitical contestability over the selection ofgovernments, though some of these govern-ments have in turn concentrated politicalpower to limit contestability in policymaking.

• Noncompetitive political regimes constrainentry of potential opposition parties into theelectoral process and sharply restrict politi-cal participation through the exercise of civilliberties. Such systems tend to have few in-stitutionalized limitations to check the execu-tive. These countries include Belarus,Kazakhstan, Turkmenistan, and Uzbekistan.

• In addition to political liberties and civilrights, war and political violence define afourth group of transition economies, whereexternal conflicts or extreme internalcontestability has strained the state at certainperiods of the transition.1 War-torn politicalregimes have engaged in prolonged wars orcivil conflicts over the past decade. Such con-flicts have generally been rooted in ethnic orterritorial divisions. In these countries thereis political contestability over the boundariesof the community to be governed or who hasthe right to select leaders and make bindingrules for that community. Such conflicts haveplaced severe strains on the capacity of thestate, resulting in some of the countries in aprolonged loss of political order and controland serious weaknesses in the provision ofbasic public goods. They include Albania,Armenia, Azerbaijan, Bosnia-Herzegovina,Georgia, and Tajikistan.2

Like all ideal types, these four categories arenot intended to be absolutes. Several transition

economies straddle different categories. Bulgariaand the Slovak Republic lie on the border betweencompetitive and concentrated political regimes.Azerbaijan has characteristics of both war-tornand concentrated political regimes. Moreover,political systems are still evolving, so countrieshave been shifting over time across different idealtypes. Croatia and the Slovak Republic havemoved sharply toward a more competitive demo-cratic system following critical elections that endedthe lengthy regimes of once powerful political lead-ers. Romania has shown steady improvements inextending political rights and civil liberties. Suchchanges suggest that countries are not locked intoany particular path of political transition. Butbecause this report summarizes the entire transi-tion path, we rely on measurements averagedacross the entire transition period.

Competitive Democracies Have High PoliticalContestability…

Political contestability is a function of therights to participate in the political system

and the extent to which competing institutionalactors have the power to influence politicaldecisionmaking. It is generally measured by thenumber of political actors whose approval isnecessary to adopt binding decisions on the pol-ity. As the number of political parties in a coa-lition government increases, the number of po-tential veto points for policy decisions alsoincreases. Similarly, presidential systems inwhich competing political parties dominate theexecutive branch and legislature (that is, dividedgovernment) should also exhibit highercontestability than those where the president’sparty also controls the legislature. A simplemeasure of political contestability based onthese veto points has been developed by Roubiniand Sachs (1989). The index measures the num-ber of competing political parties in governingcoalitions in both parliamentary and presiden-tial systems (data for the transition economiescan be found in Frye and Hellman 2001; seealso figure 10.2).

Competitive democracies have the highestnumber of veto points among countries in the


100

FIGURE 10.2.

Veto Points Index, 1989–99

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).

HungaryPoland


Slovak Republic

Latvia

Lithuania

Croatia

Kyrgyz Republic

Kazakhstan

Bulgaria


Moldova

Romania

Armenia

Albania

Russian FederationUkraine

Azerbaijan

Uzbekistan

Tajikistan

Belarus

Turkmenistan

43210

Veto points index


War-torn regimes



Average

Average

Average

Average

Slovenia

00

Classifying Political Systems in Transition

101

region. All of them are parliamentary orsemipresidential systems governed by multipartycoalitions. Lithuania and Poland could be char-acterized as semipresidential systems, which com-bine parliamentary government with a directlyelected president. Indeed, one-party majority gov-ernments have been rare in most of these coun-tries. Six of the seven competitive democracieshave had prolonged periods of coalition govern-ments consisting of three or more political par-ties. In Estonia, Poland, and Slovenia coalitiongovernments of up to five political parties havenot been unusual. While the literature on politi-cal economy generally takes a negative view ofcoalition governments, the most reformist gov-ernments in transition economies have tendedto be multiparty coalitions (Alesina andRosenthal 1995; Roubini and Sachs 1989).

Concentrated political regimes generallyhave fewer veto points on average (see figure10.2). Five of the eight concentrated politicalregimes are presidential systems. Bulgaria andthe Slovak Republic, on the border between con-centrated and competitive political systems,have multiparty coalition governments moresimilar to a competitive system than a concen-trated one. Several of the presidential systemsappear to have been ruled by divided govern-ments for much of the transition, mainly be-cause presidents in the Kyrgyz Republic, Rus-sia, and Ukraine did not belong to any politicalparty for extended periods.

Noncompetitive political regimes have thelowest scores on the veto points index. Powerfulpresidents in Belarus, Kazakhstan, Turkmenistan,and Uzbekistan have essentially created one-party systems with strong restrictions on oppo-sition parties. Similar one-party systems have alsobeen developed in the war-torn political systems,such as Tajikistan, but these countries have gen-erally had much less stable governments as a re-sult of the conflicts.

…and High Government Turnover

Countries with greater political contestabilityalso have, as might be expected, more fre-

quent changes of government (figure 10.3). In-deed, the countries most advanced in economicreform have tended to have the most frequentchanges in government, contrary to the conven-tional view that such turnovers create an un-certain environment that undermines reform.There have been nine governments in Poland,seven in Estonia, and five in Hungary since thestart of transition. As a group the competitivedemocracies have had an average of six gov-ernment turnovers since the collapse of the So-viet bloc. This contrasts sharply with the othercountry groups. Concentrated political regimes,except borderline Bulgaria, have tended to havefewer government turnovers, averaging justmore than three for the group. In all of the non-competitive countries, except Belarus, the leaderat the time of the dissolution of the Soviet Unionhas ruled continuously throughout the transi-tion. Political continuity has not positively af-fected the government’s propensity to adopteconomic reforms.

Notes

1. This report defines political violence as governmentturnovers or attempted turnovers through violence.

2. This group includes all countries in the Europeand Central Asia region that have been engaged inprolonged military conflicts, except Croatia and Rus-sia. In Croatia the military conflict did not sparkdomestic political violence that threatened the in-cumbent regime of Franjo Tudjman. Russia engagedin a long-term territorial conflict in the breakawayrepublic of Chechnya and survived a violent attemptto overthrow Boris Yeltsin. However, these twoevents were unrelated and did not lead to disorderand political fragmentation, as in other countries inthe group.


102

FIGURE 10.3.

Main Political Executive Turnovers, 1989–99

Note: This figure measures the number of times there has been a change of the country’s lead executive—president inpresidential systems and prime minister in parliamentary or semipresidential systems—since the country gained independence.

Source: Frye and Hellman (2001).

Hungary

Poland

Czech Republic

Estonia

Slovak Republic

Latvia

Lithuania

Croatia

Kyrgyz Republic

Kazakhstan

Bulgaria


MoldovaRomania

Armenia

Albania

Russian Federation

Ukraine

Azerbaijan

Uzbekistan

Tajikistan

BelarusTurkmenistan

106420

Number of turnovers


War-torn regimes



Average

Average

Average

AverageSlovenia

8

103

The simple classification of political systems in transition economies allows us to examine therelationship among different types of political institutions and the range of outcomes on thespectrum of discipline and encouragement discussed in part 2.

Competitive democracies have proven to be among the most advanced economic reformers in theEurope and Central Asia region, pursuing policies that have promoted SMEs and maintaining hardbudget constraints on both new and old enterprises. Concentrated political regimes have been morelikely to sustain a pattern of partial reforms that protect old enterprises and create barriers to marketentry. Yet in these countries the combination of liberalization and privatization with continued softbudget constraints and a weak rule of law have encouraged even new enterprises to focus their effortson 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 com-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 seekingand tunneling prominent in the concentrated political regimes, thus avoiding the sharp contractionscommon among other reform-minded countries. War-torn countries have been characterized by weakstate 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 andDevelopment’s transition indicators (EBRD 2000), which evaluate annual progress in transition ineight 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 madethe 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 economic

11Political Systems Influence

the Choice of EconomicReforms


104

FIGURE 11.1.

Political Systems and Economic Reform Outcomes, 2000

Note: The European Bank for Reconstruction and Development measures annual cumulative progress in transitioneach year in eight different categories of market-oriented reform on a scale from 1 (little or no reform) to 4.3 (standardstypical of advanced industrial economies). Each bar represents the country average across all eight categories. The linerepresents the variance, which for each country measures the dispersion around the mean across all eight reform categoriesof the transition indicators.


Hungary

PolandCzech Republic

Estonia

Slovak Republic

LatviaLithuania

Croatia

Kyrgyz Republic

Kazakhstan

Bulgaria


Moldova

Romania

ArmeniaAlbania


Azerbaijan

Uzbekistan

Tajikistan

BelarusTurkmenistan

Variance


War-torn regimes



Average

Average

Average

Average

Slovenia

4321

Mean scores

Bosnia and Herzegovina

1.20.80.40

Political Systems Influence the Choice of Economic Reforms

105

reform across the world, the results are still largelyinconclusive (for a review of this literature, seeHaggard and Webb 1993). Though the nature ofthe political system structures the incentives ofpoliticians to adopt economic policy choices, thereform choices themselves shape the configura-tion of social groups and the distribution of powerthat affect the structure and functioning of thepolitical system. For example, economic reformsthat enable new entry also strengthen the con-stituency of SMEs that build support for increas-ing political competition. Thus, the developmentof the political system and progress of economicreforms are closely inter-related.

Yet given the sharp break with communismand the disintegration of the Soviet Union,choices about the structure of the political sys-tem in transition economies were generally madebefore decisions about the nature and pace ofeconomic reform. As a result, a stronger case canbe made for identifying the direction of causa-tion from political choices to economic choices.Moreover, only a few countries (Belarus, Croatia,and the Slovak Republic) have seen a majorchange in political regime since the start of tran-sition. This suggests that while the pace and di-rection of economic reform may have reinforcedinitial choices about the structure of the politi-cal system, economic reforms have yet to deci-sively shift the course of political transition.

However, after only a decade of transition,these political and economic systems are still intheir infancy. As in all countries, one might ex-pect that over the long term, there will be a stronginteractions among economic reform, economicperformance, and political evolution (Przeworski1991). Yet in the early stages of transition, po-litical choices appear to be the driving force ofchange in economic reform.

Political Systems Create Rent-SeekingOpportunities

An important indicator of the opportunitiesfor rent seeking in the transition process

comes not just from the overall measure ofprogress in market-oriented reform, but fromdifferent progress across different componentsof the reform agenda. Imbalances in the reform

process—such as price and trade liberalizationwith continued restrictions on entry, privatizationwith soft-budget constraints, and rapid creationof banks without a sufficient regulatory frame-work for financial markets—create opportuni-ties for rent seeking and theft. One crude indica-tor of these imbalances comes from comparingthe different rates of progress across the eightdifferent components of the European Bank forReconstruction and Development’s transition in-dicators. Taking the variance (a measure of thedispersion around the mean) of the ratings foreach country across these eight reform compo-nents gives a rough measure of the imbalancesin the reform that give rise to rent seeking (seefigure 11.1).

The variances tend to be lowest in the mostreformist and least reformist countries. In com-petitive democracies economic reforms have gen-erally progressed across the board despite dif-ferences in the sequencing of reform measuresacross countries. Similarly, noncompetitive po-litical systems have generally made little or noprogress in all of the key areas of economic re-form, maintaining substantial continuity with theprevious command system. In contrast, the con-centrated political regimes and war-torn regimeshave tended to advance rapidly in liberalizationand privatization with much slower progress inthe institutional reforms to support effectivelyfunctioning markets, generating much highervariances in their reform scores.

These asymmetries in the reform process cre-ate a wide range of arbitrage and rent-seekingopportunities available to a small group, usu-ally with close ties to the government or the ex-isting state-owned sectors. The gap betweenprogress in liberalization and privatization anddevelopment of a proper legal and regulatoryframework also provides opportunities for theftof both state and private assets through expro-priation of minority shareholders.

The pace and symmetry of market-orientedreforms suggest variation across the different po-litical systems in rent seeking and asset stripping.A good proxy measure of such phenomena canbe found in recent attempts to develop more sys-tematic, survey-based indices of corruption andstate capture. (State capture refers to the efforts


106

of enterprises to influence laws, decrees, regula-tions, and the like through private payments topublic officials.) By capturing state institutionsenterprises seek to extract rents from the statethrough a myriad of preferences, exemptions, andanticompetitive practices. The Business Environ-ment and Enterprise Performance Survey (see box

3.1) provides an index of state capture based onthe share of enterprises that reported a direct im-pact on their business from private influence pay-ments to public officials (figure 11.2; see chap-ter 3).1 The extent of state capture can becompared across countries with different typesof political regimes.

FIGURE 11.2.

State Capture Index, 1999

Note: The state capture index is based on the Business Environment and Enterprise Performance Survey. Enterpriseswere 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 eachcountry represents the share of enterprises that reported a significant impact averaged across the six institutions. As themeasurement of this index is subject to a margin of error, any efforts to rank countries would be inappropriate. The datawere collected in 1999 and do not reflect the impact of reforms since that time.

Source: Hellman, Jones, and Kaufmann (2000); World Bank (2000c).

Hungary

PolandCzech Republic

Estonia

Slovak Republic

Latvia

Lithuania

Croatia

Kyrgyz Republic

Kazakhstan

Bulgaria

Georgia

Moldova

Romania

ArmeniaAlbania


Azerbaijan

UzbekistanBelarus

5040200

Index


War-torn regimes



Average

Average

Average

Average

Slovenia

3010


107

There is a stark contrast in the extent andimpact of state capture across different politicalsystems. Concentrated political regimes exhibitconsistently higher state capture, affecting onaverage more than twice as many enterprises asin competitive or noncompetitive political sys-tems. Except Latvia, only a small share of enter-prises in competitive democracies report a sig-nificant impact from state capture. State captureis also low in noncompetitive political regimes,reflecting the weakness of the private sector rela-tive to a highly authoritarian state. War-torncountries have both high and low state capture:low in countries with high domestic instability,such as Albania, Armenia, and Bosnia and highin countries where powerful leaders have begunto consolidate political power, as in Azerbaijanand Georgia. The concentration of politicalpower appears to be an important determinantof the extent of state capture in transition econo-mies (Hellman, Jones, and Kaufmann 2000).

How Do Political Systems Affect EconomicReform?

The data suggest clear links among differenttypes of political systems and alternative

paths of economic reform as defined in the disci-pline and encouragement framework. Of course,the correlation between political regimes and dif-ferent economic reform paths need not imply cau-sation. There may be other factors, such as his-torical and cultural legacies, the structure of theeconomy, and even geographical position that af-fect both the choice of political regime and thecourse of economic reform. Such factors are allhighly correlated, making it impossible to disen-tangle empirically the lines of causation. However,the strong similarities in the transition process incountries with similar political systems suggest thatpolitical institutions affect choices about economicreform. What are these similarities?

In Competitive Democracies, Inclusion PromotesSupport for Comprehensive Reform

In the aftermath of popular revolutions againstcommunist rule, the political institutions in com-petitive democracies were forged in roundtable

negotiations by popular fronts and a wide rangeof other organized interests from trade unionsto religious representatives. Guided partly by theexample of Western European democracies, theroundtables produced political institutions thatgenerally tended toward parliamentary systems,promoting party competition and constrainingexecutive authorities.

The inclusive process for creating these po-litical institutions and the broad range of politi-cal groups that could compete for power in thenew system enhanced the capacity of govern-ments to make credible commitments that thepromised gains of economic reform would notbe expropriated or otherwise restricted to par-ticular vested interests. This contributed to awider social consensus on the main directions ofreform, despite differences among parties on thesequencing and pace of reforms. It also led to agreater mobilization of organized interest groupsin civil society (such as independent labor unions)that would enhance political accountabilitythroughout the transition. So governments mo-bilized broad public support for comprehensivereform programs early in the transition.

A key factor in building and sustaining thisbroader consensus on reform has been the his-torical ties and geographical links of these coun-tries to Western and Northern Europe. The pullof European accession generated strong incen-tives for a common institutional framework, bothfor the economy and the polity, creating a focalpoint for the reform agenda that mitigated pos-sible disputes over alternative institutional de-signs and regulatory frameworks.

By taking a comprehensive approach early on,such programs generated far fewer opportunitiesfor rent seeking and theft, lowering the returns tosuch activities in the short term. This weakenedthe capacity of short-term winners to restrict com-petition and preserve rent-generating market dis-tortions. Moreover, the regular succession of dif-ferent coalition governments created genuinecompetition among groups for political influence.That led to equally fierce competition over rents,quickly dissipating any efforts to concentrate rentflows and preventing theft on a massive scale.

As already suggested, this competition oc-curred within a broad consensus on the direction


108

and goals of the reform. Even the return to powerof communist-successor parties in such countriesas the Czech Republic, Hungary, and Poland didnot derail reform. Government credibility at theearly stages of transition enabled a first move onthe reform path that limited the rents from arbi-trage opportunities in distorted markets. It quicklycreated new constituencies with an incentive topush for further reform. It also signaled to inves-tors the government’s commitment to more long-term structural reforms. Some groups gained im-mediate advantages from liberalization andprivatization. But they could not convert the gainsinto enough political influence to erect barriers tocompetition and entry that would have preservedlimited opportunities for rent seeking and theftalong the way.

From their first moves in the reform process,new governments tended to focus on promotingnew constituencies of winners, removing entrybarriers, and quickly tackling severe macroeco-nomic instability. They also supported the losersfrom the dislocations of the reform by maintain-ing adequate social protection (Orenstein 2001).Early efforts by enterprise insiders to spontane-ously privatize enterprises sparked a substantialpolitical backlash in the Hungarian and Polishelectoral arenas (and much later in the Czechelections), curbing the practice with varying ef-fectiveness. As a result the concentration of eco-nomic power in the early stages of transition inthese competitive democracies was far less thanin other parts of the region (evident in the dataon inequality). As reforms progressed to promoteentry and improve the enabling environment,they strengthened the constituencies with a stakein moving the reform agenda forward in the dif-ficult areas of structural and institutional change.

What enabled this virtuous circle? One keyfactor was the much higher state capacity thananywhere else in the region. The fairly peacefulexit from communism did not destroy key stateinstitutions. Facilitating more effective imple-mentation of reforms, the legacy of public ad-ministration provided important preconditionsfor promoting new entry, such as greater secu-rity of property and contract rights and betterpublic infrastructure.

In Concentrated Political Regimes, Oligarchs andInsiders Capture the State

In concentrated political regimes the collapse ofcommunism was more a result of the contestamong competing elites than a broad socialmovement. The new political regimes were notforged though roundtable negotiations amongpotential competitors, rather, the regimes weredesigned by incumbent leaders to consolidatetheir power. These new regimes tended to bepresidential systems with power concentrated inthe executive branch. Political parties were weakand did not represent a broad range of socialinterests. Moreover, the old nomenklatura re-mained strong, especially in the economy.

Though comprehensive reforms were pro-posed in some concentrated political regimes inthe early stages of transition (Russia in 1991),they were rarely adopted in full, and the onesadopted were poorly implemented. The regimeslacked the credibility to build and sustain broadpopular support for such reforms. Instead, theytended to fall back on partial liberalization andprivatization. The soft budgets and remainingbarriers to entry generated tremendous oppor-tunities for rent seeking and theft, especially ineconomies rich in natural resources.

Such exits from communism also led to muchsharper deteriorations of state capacity thanamong the competitive democracies of CentralEurope. In former Soviet states, secessions wereconsciously intended to weaken federal controlby radically decentralizing authority over publicbureaucracies, sparking a wave of asset stripping.In Bulgaria and Romania, for example, the col-lapse of incumbent regimes caused greater uncer-tainty than in the negotiated transitions of Cen-tral Europe, eroding state control. Furthermore,the deterioration of public administration pro-foundly reduced the state’s capacity to raise rev-enue, implement proposed economic reforms, andbuild broader consensus around reform goals.

The partial reforms merely increased andconcentrated rents and the opportunities fortunneling and theft. Comprehensive price lib-eralizations were often followed by a stream ofexecutive orders and decrees (with exceptions


109

for some sectors and goods). Licensing andother regulatory barriers preserved trading mo-nopolies. Explicit controls on foreign competi-tion were set.

The winners from these partial reformsreaped spectacular gains. Those with controlrights over valuable assets and political accessto rent-seeking opportunities could easily priva-tize those assets in an environment of poorlydefined property rights, nonexistent corporategovernance, and an inadequate regulatory frame-work. As long as the stock of rents was not de-pleted, these winners had a strong incentive topreserve their advantages, using their consider-able resources to block reforms that threatenedthose advantages.

Countervailing pressures from competinggroups were weak, and without genuinely broad-based social movements, no clear goals were ar-ticulated at the start of transition. Nor was theremuch social mobilization from the collapse ofcommunism. As a result, these countries em-barked on transition without building a broadsocial consensus on the goals of reform and with-out a means of organizing the public behind thesegoals. Effective political parties never material-ized to mobilize social support, and residual sup-port for the Communist Party remained high.That set the stage for much greater political po-larization over economic reform, and that po-larization in many cases became a pretext forfurther centralization of political power and lim-its on political competition (EBRD 1999).

Lacking strong social support and a solid po-litical base, incumbent politicians in concentrateddemocracies sought alliances with powerful in-cumbent enterprises for funding and support. Thisnaturally made the state highly subject to capture.The early winners from the mass of arbitrage op-portunities had a strong interest in using their po-litical influence to preserve rent-generating dis-tortions that intensified the winners’ economicpower. Direct barriers to entry solidified theirgains, weakening new constituencies that mightcounterbalance these powerful incumbents.

The social costs of state capture have been highin these countries. But the direct costs to politi-cians of poor policy choices were low thanks to

limited institutional restraints to promote account-ability and weak intermediate organizations tochannel the dissatisfaction of losers. Many con-centrated democracies languished in an equilib-rium trap of partial economic reforms in whichthe concentration of both political and economicpower preserved market distortions generatinghighly concentrated gains to narrow vested inter-ests at considerable social cost.

In Noncompetitive Political Systems, LeadersSavor Status Quo and Economic Stability

In noncompetitive political regimes leaders fromthe Soviet era generally pursued economic sta-bility, while securing their dominance of the post-Soviet political system. Concerned about rivalsources of power in both the economy and thepolity, they largely rejected market-oriented re-forms. They feared the opening to global mar-ket forces and the rise of oligarchs in partiallyreformed economies. Instead, they preferred tomaintain continuity with key elements of theprevious command system to maintain the state’s(and the leaders’) predominant role and to avoidany destabilizing adjustment costs associatedwith reform.

In these cases economic reform was drivennot by the winners or losers of reform, but byauthoritarian political leaders trying to maintainpolitical control and ensure some economic sta-bility. Limited economic reform went hand inhand with limited political reform, as incumbentleaders severely restricted political opposition.

The political imperative to maintain statepower and the lack of economic reform avoidedthe disintegration of state capacity that plaguedmany other transition economies. So althoughthe state in these countries does not provide theinstitutional foundation for the market economy,it still provides public goods—but at levels simi-lar to the communist era. This suggests that if anew government introduced comprehensive eco-nomic reforms it would have stronger capacityto implement them and would begin with a lessdaunting decline in living standards.

Despite these possible advantages, incumbentpolitical leaders face little pressure to pursue


110

reforms. By maintaining dominant state ownershipin the economy and considerable discretionary ca-pacity to intervene in economic affairs, these lead-ers ensure their positions remain powerful. Overtrestrictions on political competition reduce evenfurther the risk of challenges to incumbents’ power.

In these systems few viable constituencies havethe capacity to push for economic reforms. Theeconomic elites tend to be closely allied with (orco-opted by) incumbent political leaders. Thereis no critical mass of new actors with an interestin reforms. Trade unions, business associations,and other civil groups that might representbroader social interests are circumscribed or tiedto the state. In addition, the lack of reform iso-lates these countries from the global economy,leading to autarky that undermines what exter-nal constituencies might do in promoting reform.

The extent of economic reform is thus highlydependent on the political leader’s preferences.Although many authoritarian leaders in otherregions adopted comprehensive economic re-forms, this has not yet happened in any of thetransition economies. Why? Authoritarian lead-ers in the region have tended to inherit their powerand support from the surviving structures of theformer communist system. This has created astrong link between authoritarian political powerand command administrative methods in the eco-nomic sphere in transition economies.

In War-Torn Regimes, Violence and Lack ofCredibility Prevent Reform

In war-torn political systems, efforts to promotecomprehensive reforms at the early stages of tran-sition were thwarted by contestability over whohad the rights to make binding decisions for thecommunity and over the definition of the com-munity. Societies with deep-rooted ethnic divi-sions face a high risk that governments domi-nated by one group will seek to expropriatewealth and resources from rivals, while encod-ing advantages for themselves into the develop-ing institutional structure.

Facing such threats, minority ethnic groups inseveral countries in the region have violently chal-lenged the legitimacy of the state. Croats in Bosnia-Herzegovina, Armenians in the Azerbaijani enclave

of Nagorno-Karabakh, Albanians in the Kosovoregion of Yugoslavia, Abkhaz and Meshketians inGeorgia, fundamentalist Muslims in Tajikistan—all are cases of minority groups seeking to pre-emptthrough violent means the creation of a state domi-nated by the ethnic majority. Governments in suchenvironments could not make credible commit-ments about the distribution of future gains fromreforms or about not expropriating wealth.

During periods of peace and relative stabil-ity, governments in these countries have tried toadopt comprehensive reform programs, but theycontinue to be undermined by severe credibilityproblems. War and instability radically shortentime horizons, so ethnic groups are not preparedto accept short-term sacrifices for the promiseof long-term gains.

Prolonged conflict also sharply reduces out-put, living standards, and the resources of thestate. Physical and human capital deteriorate.Poverty increases. Even the capacity of the stateto provide the most basic public goods falls apart.

War also can concentrate economic powerin the hands of smugglers, arms traders, andparamilitary groups, who use their power inpeacetime to secure their positions in theeconomy. These “winners” use the instability ofwar to centralize control over state assets anddistribution networks, particularly energy. Theyalso tend to maintain close relationships withpolitical leaders, who depend on them to fundand supply the war. Given the weakness ofcountervailing interests, constraining the powerof such groups during peacetime is difficult.Manipulating privatization to enhance their con-trol over assets, the winners use their influenceto preserve exemptions and other preferences thatundermine competition and to weaken the de-velopment of the rule of law.

This combination of political instability, aweak state, and powerful economic groupsrooted in illicit and opaque trade has underminedreform in most war-torn regimes.

Note

1. Hellman, Jones, and Kaufmann (2000) presentevidence from the Business Environment and Enter-prise Performance Survey to show that enterprises thatengage in state capture get substantial rents.

111

After a decade of transition, the political challenges of pressing ahead with the remainingreform 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-termadjustment costs for the potential new entrants and the high concentration of gains for the short-termwinners such as oligarchs and insiders. Also needed are political changes to enhance the government’scredibility and capacity to constrain the power of constituencies seeking to sustain partial reformsregardless 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 thedislocations 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 enhancegovernment credibility.

To smooth the short-term winners’ curve, governments need to reduce the incentives that leadoligarchs 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 implementingsuch schemes is small. So reducing the concentration of gains to oligarchs and insiders must be basedon changes in the structure of political power. Mobilizing collective action by a broader range of socialgroups that will gain from further reforms could pressure politicians to reduce state capture. Reform-ing the political system—to increase participation and enhance political competition—can break thelink between highly concentrated economic power and political power.

Increasing transparency and accountability in government increases the costs to politicians of skewingeconomic reform in the interest of narrow constituencies. Of course, the oligarchs and insiders willoppose 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 changesin the underlying correlation of power in the system.

12Confronting the Political

Challenge


112

Not all the transition economies battle thesesame problems. A decade of variation in reformpaths and political developments has left coun-tries with very different political challenges.

For Concentrated Political Regimes,Mobilizing Potential Winners

Breaking the political economy equilibriumunderlying state capture and partial reforms

is the most important and difficult challenge inadvancing the transition in countries with con-centrated political regimes. The vested interestsunderlying soft budget constraints, barriers toentry, opaque regulatory frameworks, and landreform have accumulated considerable economicand political power, raising more barriers to po-litical entry for the losers of such policies. Statecapture at all levels of the political system andthe lack of accountability for politicians andpublic officials make it difficult for even the mostcommitted reformers to overcome the powerfulvested interests against reform. In addition, weakstate capacity has limited the state’s autonomyto tackle these groups effectively.

Reforms have proven most difficult politicallywhere the rents have been most concentrated,particularly in energy and finance. In nearly allthe concentrated democracies with substantialnatural resources, energy lobbies have been amongthe most formidable opponents of reforms. Theirpolitical power, often exceeding that of majorpolitical parties, cannot be underestimated.

The political leverage of the power and energysectors comes not just from their resources, butthrough the vast network of nonpayments. As stud-ies of the virtual economy suggest, dominant powerand energy monopolies are the key sources of valuein the complex web of nonpayments because themonopolies continue to provide oil, gas, andelectricity to loss-making enterprises, often in ex-change for overvalued barter goods (Gaddy andIckes 1999). Being the main conduits of the state’ssoft-budget constraint gives them considerable po-litical influence over the state.

Many other groups could be considered win-ners from the distortions of partial reforms.Natural resource extraction and export enter-prises have gained from distorted domestic prices,

subsidized inputs, and monopoly production anddistribution rights. Commercial banks have takenadvantage of macroeconomic volatility, ineffi-cient financial markets, and lax regulatory struc-tures to reap gains from arbitrage. Politicians andbureaucrats have used their discretionary pow-ers to intervene in the economy to extract bribesand other advantages from enterprises andhouseholds. Enterprise insiders have manipulatedunclear property rights and weak corporate gov-ernance to divert enterprise assets and cash flowinto offshore companies and other subsidiariesunder their direct ownership. These practiceshave been largely at the expense of unprotectedsmall shareholders.

The winners from these market distortionsand inefficiencies have powerful incentives topreserve them and the associated rent flows, re-gardless of the social costs. Large exporters lobbyfor entry barriers. Commercial banks oppose sta-bilization programs and proposals to enhancethe central bank’s regulatory powers. Insider-owners undermine efforts to clarify corporategovernance and to introduce greater transpar-ency into the distribution of property rights.Public officials resist deregulation and efforts tolimit discretionary interventions. Although thegains from such distortions tend to be highlyconcentrated, the losses are dispersed amongconsumers, savers, minority shareholders, start-up companies, small and medium-size businesses,and foreign investors.

This pattern of dispersed losses and concen-trated gains holds the key to designing politi-cally feasible strategies of economic reform. Inmany concentrated democracies, reformers of-ten attempt to overcome political obstacles byaugmenting executive power to counter thepower of vested interests. Yet given the statecapture in these countries, such efforts tend tofail. At various times the presidents of bothRussia and Ukraine were granted extraordinarydecree-making powers to push through eco-nomic reforms that did not break the stalemateon further reforms. Instead, concentrated gainsand widely dispersed losses suggest the mobili-zation of the losers in the existing low-level equi-librium through greater political inclusion. Howsuch a strategy should be designed depends on

Confronting the Political Challenge

113

factors specific to each country. But some gen-eral approaches have worked in other transi-tion economies.

Mobilize the electorate. Electoral appeals tomobilize the losers of partial reform have builtsupport for macroeconomic stabilization and forbanking reform. Given the wide and generallyregressive impact of high inflation, political par-ties in several transition economies have mobi-lized enough electoral support for macroeconomicstabilization to overcome the opposition of pow-erful commercial banks and other actors thatgained from macroeconomic volatility. In addi-tion, banking crises in the Czech Republic andHungary sparked electoral appeals to disgruntledsavers, helping to break the stalemate over bank-ing privatization and regulatory reform.

Mobilize excluded enterprises. Another im-portant strategy for weakening the oppositionof narrow vested interests to reforms is mobiliz-ing collective action among enterprises excludedfrom these concentrated gains. SMEs, new en-terprises, and second tier enterprises suffer mostfrom existing weaknesses in the enabling envi-ronment, from discretionary taxation and regu-lation and from anticompetitive barriers. How-ever, they lack vehicles of collective action andinfluence with the government. In Central Eu-rope business associations have strengthened thevoice of this tier of the economy, constrainingpublic officials and checking influential enter-prises and financial-industrial groups. In the con-centrated democracies, such associations haveremained weak, and political parties have notsought strategic alliances with such actors as analternative basis for support and funding.

The Business Environment and EnterprisePerformance Survey (see box 3.1) showed vastdifferences across transition economies in howmany enterprises were members of business as-sociations and how they used these associationsto seek remedies to problems with the govern-ment. In Hungary more than 75 percent of theenterprises surveyed reported membership in atrade association, and 60 percent said they wouldrely on such associations first in handling prob-lems with the government. But in Russia fewer

than 20 percent of the enterprises reported mem-bership in business associations, and less than10 percent relied on them for resolving prob-lems with the government (World Bank 2000c).

The unofficial sector is a large and poten-tially influential constituency that has not beeneffectively mobilized in the concentrated democ-racies. Enterprises in the unofficial economy gen-erally suffer most from the weakness in the en-abling environment, especially in the opportunitycosts associated with limitations on their capac-ity to expand operations. Tax reforms that lowermarginal rates can promote entry from the un-official to the official economy. This would pro-mote growth and possibly crystallize an impor-tant political constituency to remove obstaclesin the business environment that work only tothe advantage of a narrow group of powerfulenterprises. Given the size of the informal sectorin many of these countries, the potential politi-cal consequences of mobilization could seriouslyshift the balance of power away from incum-bent vested interests to a much broader collec-tion of economic actors. However, there are fewexamples of rapid shifts from the informal tothe formal sector, or of the political mobiliza-tion of such economic actors.

One way to promote new enterprises andthose in the unofficial sector is to align the in-centives of local governments to increase entry.Tax-sharing schemes between central and localgovernments can be modified so that propertytaxes on small business are assigned exclusivelyto local governments, encouraging them to re-form the enabling environment.

Ensure—and use—a free media. The emer-gence of a free press and broadcast media canovercome the coordination dilemmas associatedwith reforms that generate concentrated winnersbut highly dispersed losers in many sectors. Themedia can promote collective action among thesedispersed losers of partial reforms and increasethe political pressure on those who gain fromthe general equilibrium underlying these econo-mies. The challenge for reformers in concentrateddemocracies is to use the media to make clearthe links between the rents from partial reformsand the direct costs to society.


114

But the problem in many of the concentratedpolitical regimes is that ownership of the media isclosely intertwined with powerful financial-industrial groups. In other concentrated politicalregimes the media remains largely under the con-trol of the state. This naturally limits what themedia can do in breaking the partial reform trap.

Make obvious what is hidden. Tax arrearsand nonpayments need to be linked in the pub-lic mind with delayed public sector wages andpensions and poor provision of social services.The complex web of hidden subsidies to power-ful business needs to be exposed, making clearthat such subsidies tend to benefit incumbentmanagers (often through offshore accounts)rather than workers. Barter and arrears in theenergy sector need to be linked to domestic powerand fuel shortages and outages that plague manyof these countries.

Converting hidden and discretionary subsidiesto enterprises into budgetary subsidies to supportworker training, severance schemes, and better lo-cal services in communities affected by downsizingis vital for discipline and encouragement.

The potential for mobilizing these dispersedlosers is particularly high in countries whose so-cial sectors need resources, but whose high-profile conglomerates in key sectors enjoy a rangeof explicit tax and duty exemptions and main-tain high tax arrears. This is particularly true insome of the small, fairly open economies, suchas Georgia or Moldova, where trade flows are amajor source of the tax base and powerful inter-ests in control of those flows are the main taxevaders. Explaining to the public the costs ofevasion should be an important element of anystrategy to mobilize support for further reforms.Having made clear who benefits from partialreforms—and how those benefits come directlyat the expense of large but dispersed domesticconstituencies—reformers can begin to buildpolitical coalitions to marshal the potential win-ners from further reforms.

Allow political competition and economiccompetition to reinforce each other. Enhancingpolitical contestability by mobilizing civil society,

strengthening political parties and other organi-zations representing the collective interests of al-ternative constituencies, and developing institu-tional restraints will expand political accessbeyond vested interests and increase the costs topoliticians of maintaining partial reforms. Butbreaking the equilibrium of partial reforms is achallenge that can be addressed only through si-multaneous economic and political reforms.Though exogenous shocks often create windowsfor decisive action by committed reformers, theircapacity to spur profound changes in the struc-ture of political and economic power simulta-neously is highly constrained, and their agendahighly overloaded.

Our analysis suggests that promoting entryand competition on the margin, whenever andwherever possible, creates the necessary precon-ditions for a gradual move out of the partial re-form trap. As more enterprises enter the marketeconomy, the competition for rents should alsopromote competition for political influence, whichweakens the capacity of powerful enterprises andsectors to capture the state and oppose reformsthat might undermine their advantages. Promot-ing entry and competition on the margin will notaccomplish radical changes in the short run, butit may be the most effective means of creatingdemand for reforms in the long run.

For War-Torn Political Systems, RestoringStability and Reducing Uncertainty

Though the magnitude of the challenges facedby these countries in light of the extreme

degradation of state capacity and concomitantsharp decline in living standards may be amongthe most serious in the region, the preconditionfor success is clear: without resolving the un-derlying divisions behind political fragmenta-tion and violence, further reforms are unlikelyto be successful.

Once some measure of stability is restored anduncertainty reduced, these countries need to re-build basic state capacity and restore public goodsfor the functioning of the market. Attention hasto go to strengthening the state’s capacity to col-lect tax revenue to meet the considerable fiscal

Confronting the Political Challenge

115

challenge. Rapid liberalization in the immediateaftermath of conflict is critical to eliminating therents associated with controls on prices and dis-tribution networks that fuel the winners of thewartime economy. In addition, comprehensivestructural reforms, focused on privatization anddemonopolization, are needed to prevent thesewartime winners from solidifying their positionthrough state capture in the postwar economy.

Given deteriorated state capacity in most ofthese countries, direct assistance and participa-tion by bilateral and multilateral donors will becritical in generating the resources, providing thenecessary technical assistance, and buttressing thepolitical commitment for these fundamental tasksof state building.

For Noncompetitive Political Systems, TakingAdvantage of State Capacity

A change in political regime could create op-portunities for implementing a comprehen-

sive reform strategy. Regime change is often ac-companied by a resurgence of political competi-tion, a strengthening of political parties, and arejuvenation of civil society that can pressure newpolitical leaders to pursue policy innovation andimprove economic performance. As suggestedearlier, these regimes can take advantage ofhigher levels of state capacity to implement re-forms. This suggests that there might be someadvantages of tardiness (in the spirit of AlexanderGershenkron’s famous phrase) that would en-able these countries to shift from a minimal re-form equilibrium to a more comprehensive setof reforms without the same deterioration of statecapacity that marked countries that began thetransition with partial reforms.

For Competitive Democracies, UsingMomentum to Build Coalitions for Reform

As advanced democracies around the worldhave amply demonstrated, multiparty com-

petition can pose periodic risks to macroeco-nomic stability often aligned with the electoralcycle, or what is referred to as the political-business cycle (Alesina, Roubini, and Cohen

1997). Incumbent parties can win support by ex-panding fiscal expenditures before key elections,possibly requiring sharp contractions after theelections. Some competitive democracies (Hun-gary and Poland) have faced or are now grap-pling with high budget deficits, with possibly de-stabilizing consequences.

Some competitive democracies have com-bined budget deficits with high current accountdeficits. The rapid inflow of foreign capital hasrisks and benefits. Banking systems in many ofthese countries are not sufficiently developed andregulated to handle these inflows. A key chal-lenge is to prevent a mismatch between demandfor private sector borrowing and the capacity ofthe domestic banking system to make credit al-location decisions, monitor borrowers, and en-force discipline on delinquent borrowers. Therisk of political interference in banking systemsremains and must be strenuously avoided. Pre-serving and enhancing the independence of regu-latory and supervisory bodies in the financialsystem is crucial to preventing recurrent crises.

Success should not breed complacency inmeeting the remaining challenges of structuralreform, particularly in the public sector and inpolitically sensitive areas of the economy. Main-taining the main pillars of the social safety netas a cushion in the beginning of reform was im-portant in the success of these countries. How-ever, it has also left them with highly overstaffedbureaucracies and high public sector wages,dampening their growth potential.

Strong political resistance to downsizing andwage cuts can be expected from public sectorworkers. Local governments will oppose anypolicies that reduce government control overthe provision of health services and education.Resistance from these constituencies will be dif-ficult to overcome, just as in many advancedindustrial democracies. Countering this oppo-sition means developing mechanisms to mobi-lize politically the new and rapidly expandingprivate sector, especially SMEs, whose interestsare directly affected by the discretionary powerof public sector bureaucracies.

Beyond the public sector, many of the com-petitive democracies still need to restructure


116

politically sensitive sectors, such as agriculture,coal, mining, railways, shipbuilding, and steel.The dynamics of political coalition-building thatfoster a broader consensus on the course of eco-nomic reforms also tend to give these sectors con-siderable power to demand subsidies, protection-ist measures, and other advantages as a conditionfor their political support. The agricultural lobby,in particular, is a powerful obstacle to reform.

As the reform agenda loses urgency over time,government capacity to take on powerful lob-bies will diminish. But the lack of reforms in theseimportant sectors holds back growth in the newprivate sector, weakening the overall perfor-mance of these economies. One continuingsource of pressure for reform in some of thesesectors is European Union accession. Again, it isimportant to mobilize domestic constituencies—especially the vibrant new sector, which bearsthe brunt of the costs of postponed reforms—tobuild a coalition with the strength to overcomethe opposition to further reforms.

Conclusion

The key challenge of the political economyof reform is to create the conditions that

will generate incentives for new market entry andshift the emphasis of enterprises from rent extrac-tion to entrepreneurship and productive invest-ment. Though initial conditions cannot bechanged, measures to compensate for the struc-tural peculiarities of different economies can besuggested. Though the “first move” of the reformprocess cannot be undone, coalitions of winnersand losers can be fostered to weaken the grip ofvested interests and deconcentrate monopolypower. Though the concentration of economicpower tends to be supported by a concentration

of political power, reforming political institutionscan alter the incentives of public officials to limitor even reverse these imbalances. Such changesare crucial for shifting the incentives of old enter-prises and promoting the proper development ofnew ones.

Initial conditions and political institutionsaffect the likelihood that some countries will fol-low particular reform paths. But these structuralfactors can never predetermine outcomes in suchcomplex and multifaceted processes as transition.A decade of transition shows the critical role ofpolitical leadership in shaping reform. A thor-ough political economy analysis of winners andlosers from reform can set the parameters forunderstanding the likely pressure points in anysystem and provide guidance for crafting feasiblereform strategies. But it cannot predict the qual-ity or strength of the leadership of the reformprocess that will motivate the pace and direc-tion of reforms.

Political economy analysis has an inherentstatus quo bias (Fernandez and Rodrik 1991).But experience from around the world showsthat talented political leaders can maneuvercountries out of so-called reform traps and shiftequilibrium paths. Critical elections can breaklong-term stalemates on reform. New leaderscan mobilize alternative coalitions and sparkcollective action that tips the balance of powerbetween the potential winners and losers fromfurther economic reforms. Clever reformers candevise win-win strategies that co-opt their op-ponents to build support for reform. We can-not predict the “quality of reform-mongering,”to use Albert Hirschman’s phrase (1963, p.225), either within or across countries. How-ever, it is the essential ingredient in understand-ing the politics of economic reform.

117

The simultaneous political and economic transitions in Eastern Europe and the former SovietUnion have spawned an enormous literature across a wide variety of disciplines. Thoughnumerous 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.

Selected Bibliographic Guideto the Political Economy

of Transition

The Socialist System: Reform and Collapse

Åslund, Anders. 1991. Gorbachev’s Struggle forEconomic Reform, 2nd ed. Ithaca, NY:Cornell University Press.

Easterly, William, and Stanley Fischer. 1995.“The Soviet Economic Decline.” WorldBank Economic Review 9(3): 341–71.

Ericson, Richard. 1991. “The Classical Soviet-Type Economy: Nature of the System andImplications for Reform.” Journal of Eco-nomic Perspectives 5(4): 11–27.

Fischer, Stanley, and Alan Gelb. 1991. “The Pro-cess of Socialist Economic Transforma-tion.” Journal of Economic Perspectives5(4): 91–105.

Garton Ash, Timothy. 1983. The Polish Revolu-tion: Solidarity 1980–82 . London:Jonathan Cape.

_____. 1990. We the People: The Revolution of’89 Witnessed in Warsaw, Budapest, Ber-lin & Prague. Cambridge, MA: GrantaBooks.

IMF (International Monetary Fund), WorldBank, OECD (Organisation for Eco-nomic Co-operation and Development),and EBRD (European Bank for Recon-struction and Development). 1991. AStudy of the Soviet Economy. vol. 3.Washington, D.C.

Kornai, János. 1992. The Socialist System: ThePolitical Economy of Communism.Princeton, NJ: Princeton University Press.

Murrell, Peter, and Mancur Olson. 1991. “TheDevolution of Centrally Planned Econo-mies.” Journal of Comparative Econom-ics 15(2): 239–65.


118

Comparative Economic Performance

The Nature and Causes of the Transitional Recession

Blanchard, Olivier, and Michael Kremer. 1997.“Disorganization.” Quarterly Journal ofEconomics 112(4): 1091–127.

Calvo, Guillermo, and Fabrizio Coricelli. 1993.“Output Collapse in Eastern Europe: TheRole of Credit.” International MonetaryFund Staff Papers 40(1): 32–52.

Ericson, Richard E. 1999. “The Structural Bar-rier to Transition Hidden in Input-OutputTables of Centrally Planned Economies.”Economic Systems 23(3): 199–224.

Gomulka, Stanislaw. 1998. “Output: Causes ofthe Decline and Recovery.” Working Pa-pers Series No. 8, Center for Social andEconomic Research, Central EuropeanUniversity, Warsaw, Poland.

Kornai, János. 1993. “Transformational Recession:A General Phenomenon Examined throughthe Example of Hungary’s Development.”Economie Appliqué e 46(2): 181–227.

Rodrick, Dani. 1992. “Making Sense of theSoviet Trade Shock in Eastern Europe: AFramework and Some Estimates.” InMario Blejer, Guillermo A. Calvo,Fabrizio Coricelli, and Alan H. Gelb,eds., Eastern Europe in Transition: FromRecession to Growth? World Bank Dis-cussion Paper No. 196, World Bank,Washington, D.C.

Roland, Gerard, and Thierry Verdier. 1999.“Transition and the Output Fall.” Econom-ics of Transition 7(1): 1–28.

Explaining Variation in Economic Reform Outcomes

Bokros, Lajos. 2000. “Visegrad Twins’ Diverg-ing Path to Relative Prosperity. Compar-ing the Transition Experience of the CzechRepublic and Hungary.” Conference at theCzech National Bank, Prague. September.

de Melo, Martha, and Alan Gelb. 1996. “AComparative Analysis of Twenty

Transition Economies in Europe andAsia.” Post-Soviet Geography and Eco-nomics 37(5): 265–85.

de Melo, Martha, Cevdet Denizer, and AlanGelb. 1996. “From Plan to Market: Pat-terns of Transition.” World Bank Eco-nomic Review 10(3): 397–424.

de Melo, Martha, Cevdet Denizer, Alan Gelb,and Stoyan Tenev. 2001. “Circumstancesand Choice: The Role of Initial Condi-tions and Policies in Transition Econo-mies.” World Bank Economic Review15(1): 1–31.

EBRD (European Bank for Reconstruction andDevelopment). Various years. TransitionReport. London.

IMF (International Monetary Fund). 2000.World Economic Outlook 2000: Focus onTransition Economies. Washington, D.C.

World Bank. 1996. World Development Report1996: From Plan to Market. Washington,D.C.

Key Topics on Economic Reform

General

Åslund, Anders. 1995. How Russia Became aMarket Economy. Washington, D.C.:Brookings Institution.

Blanchard, Olivier. 1997. The Economics ofTransition in Eastern Europe. Oxford,England: Clarendon Press.

Blanchard, Olivier, Rudiger Dornbusch, PaulKrugman, Richard Layard, and LawrenceSummers. 1991. Reform in Eastern Eu-rope. Cambridge, MA: MIT Press.

Clague, Christopher, and Gordon C. Rausser,eds. 1992. The Emergence of MarketEconomies in Eastern Europe. Cambridge,MA: Blackwell Publishers.

Coricelli, Fabrizio. 1998. Macroeconomic Poli-cies and the Development of Markets inTransition Economies. Budapest: CentralEuropean University Press.

Selected Bibliographic Guide to the Political Economy of Transition

119

Dabrowski, Marek, Stanislaw Gomulka, andJacek Rostowski. 2000. “Whence Reform?A Critique of the Stiglitz Perspective.” Cen-tre for Social and Economic Research,Warsaw. Processed.

Klaus, Vaclav. 1997. Renaissance: The Rebirthof Liberty in the Heart of Europe. Wash-ington, D.C.: Cato Institute.

Kornai, János. 1990. The Road to a FreeEconomy. Shifting from a Socialist System:The Example of Hungary. New York: W.W. Norton.

_____. Forthcoming. “Ten Years After ‘The Roadto a Free Economy,’ The Author Self-Evalu-ation.” In Boris Pleskovic and NicholasStern, eds., Annual World Bank Conferenceon Development Economics. Washington,D.C.: World Bank.

McMillan, John, and Barry Naughton. 1992.“How to Reform a Planned Economy: Les-sons from China.” Oxford Review of Eco-nomic Policy 8(1): 130–43.

Poznanski, Kazimierz Z., ed. 1993. Stabilizationand Privatization in Poland. Boston, MA:Kluwer Academic Publishers.

Qian, Yuan, Gerard Roland, and Chenggang Xu.1999. “Why is China Different From East-ern Europe: Perspectives from Organiza-tion Theory.” European Economic Review43(4–6): 1085–94.

Sachs, Jeffrey. 1996. “The Transition at MidDecade.” American Economic Review Pa-pers and Proceedings 86(2): 128–33.

Stiglitz, Joseph E. 1999. “Whither Reform? TenYears of the Transition.” Keynote address,World Bank Annual Bank Conference onDevelopment Economics, Washington,D.C.

Speed and Sequencing of Reforms

Aghion, Philippe, and Oliver Blanchard. 1994.“On the Speed of Transition in CentralEurope.” In Stanley Fischer and J.Rotemberg, eds., National Bureau of

Economic Research Macroeconomic An-nual 1994. Cambridge, MA: MIT Press.

Åslund, Anders, Peter Boone, and SimonJohnson. 1996. “How to Stabilize: LessonsFrom Post-Communist Countries.”Brookings Papers on Economic Activity26(1): 217–313.

Dewatripont, Mathias, and Gerard Roland.1992a. “Economic Reform and DynamicPolitical Constraints.” Review of EconomicStudies 59(4): 703–30.

_____. 1992b. “The Virtues of Gradualism andLegitimacy in the Transition to a MarketEconomy.” Economic Journal 102(411):291–300.

_____. 1995. “The Design of Reform Packagesunder Uncertainty.” American EconomicReview 85(5): 1207–23.

Fischer, Stanley, Ratna Sahay, and Carlos A.Vegh. 1996a. “Stabilization and Growthin Transition Economies: The Early Expe-rience.” Journal of Economic Perspectives10(2): 45–66.

_____. 1996b. “Economies in Transition: TheBeginnings of Growth.” American Eco-nomic Review 86(2): 229–33.

Kornai, János. 1990. The Road to a FreeEconomy. Shifting from a Socialist System:The Example of Hungary. New York, NY:Norton.

Lipton, David, and Jeffrey D. Sachs. 1990. “Cre-ating a Market in Eastern Europe: The Caseof Poland.” Brookings Papers on EconomicActivity 20(1): 75–147.

McKinnon, Ronald. 1992. The Order of Eco-nomic Liberalization: Financial Control inthe Transition to a Market Economy. Bal-timore, MD: John Hopkins UniversityPress.

Murphy, Kevin M., Andrei Shleifer, and RobertW. Vishny. 1992. “The Transition to aMarket Economy: Pitfalls of Partial Re-form.” Quarterly Journal of Economics107(3): 889–906.


120

Murrell, Peter. 1992a. “Conservative PoliticalPhilosophy and the Strategy of EconomicTransition.” East European Politics andSocieties 6(1): 3–16.

_____. 1992b. “Evolutionary and Radical Ap-proaches to Economic Reform.” Econom-ics of Planning 25(1): 79–95.

Sachs, Jeffrey. 1993. Poland’s Jump to the Mar-ket Economy. Cambridge, MA: MIT Press.

_____. 1994. “Life in the Economic EmergencyRoom.” In John Williamson, ed., The Po-litical Economy of Policy Reform. Wash-ington, D.C.: Institute for InternationalEconomics.

Privatization and Corporate Governance

Aghion, Philippe, and Olivier Blanchard. 1998.“On Privatization Methods in Eastern Eu-rope and their Implications.” Economicsof Transition 6(1): 87–99.

Amsden, Alice, Jacek Kochanowicz, and LanceTaylor. 1994. The Market Meets Its Match:Restructuring the Economies of EasternEurope. Cambridge, MA: Harvard Univer-sity Press.

Balcerowicz, Leszek, Cheryl Gray, and IrajHoshi, eds. 1998. Enterprise Exit Processesin Transition Economies. Budapest: Cen-tral European University Press.

Black, Bernard S., Reinier Kraakman, and AnnaTarassova. 2000. “Russian Privatizationand Corporate Governance.” StanfordLaw Review 52(6): 1731–808.

Blasi, Joseph R., Maya Kroumova, and DouglasKruse. 1997. Kremlin Capitalism: ThePrivatization of The Russian Economy.Ithaca, NY: Cornell University Press.

Boycko, Maxim, Andrei Shleifer, and RobertVishny. 1995. Privatizing Russia. Cam-bridge, MA: MIT Press.

_____. 1996. “A Theory of Privatization.” Eco-nomic Journal 106(435): 309–19.

Djankov, Simeon. 1999. “The Restructuring ofInsider Dominated Firms.” Economics ofTransition 7(2): 467–79.

Earle, John S., Roman Frydman, and AndrzejRapaczynski, eds. 1993. Privatization in theTransition to a Market Economy. Budapest:Central European University Press.

Frydman, Roman, Cheryl W. Gray, and AndrzejRapacszynski, eds. 1996. Corporate Gov-ernance in Central Europe and Russia, vols.1–2. Budapest: Central European Univer-sity Press.

Frydman, Roman, and others. 1993a. ThePrivatization Process in Central Europe.Budapest: Central European University Press.

_____. 1993b. The Privatization Process in Rus-sia, Ukraine, and the Baltic States. Budapest:Central European University Press.

Lipton, David, and Jeffrey D. Sachs. 1990.“Privatization in Eastern Europe: The Caseof Poland.” Brookings Papers on EconomicActivity 20(2): 293–341.

Nellis, John. 1999. “Time to RethinkPrivatization in Transition Economies?”Finance and Development 36(2): 16–9.

Roland. Gerard. 1994. “On the Speed and Sequenc-ing of Privatization and Restructuring.” Eco-nomic Journal 104(426): 1158–68.

Roland, Gerard, and Thierry Verdier. 1994.“Privatization in Eastern Europe: Irrevers-ibility and Critical Mass Effects.” Journalof Public Economics 54(2): 161–83.

Stark, David, and Laszlo Bruszt. 1998. PostsocialistPathways: Transforming Politics and Prop-erty in East Central Europe. Cambridge, MA:Cambridge University Press.

Empirical Assessments of Privatization

Blasi, Joseph, and Andrei Shleifer. 1996. “Cor-porate Governance in Russia: An InitialLook.” In Roman Frydman, Cheryl W.Gray, and Andrzej Rapacszynski, eds.,


121

Coporate Governance in Central Europeand Russia, vol. 2. Budapest: Central Eu-ropean University Press.

Brown, David, and John Earle. 2000. “Privatizationand Enterprise Restructuring in Russia: NewEvidence from Panel Data on Industrial En-terprises.” Working Paper, Russian-EuropeanCentre For Economic Policy, Moscow.

Djankov, Simeon, and Peter Murrell. 2000. TheDeterminants of Enterprise Restructuringin Transition: An Assessment of the Evi-dence. Washington, D.C.: World Bank.

Earle, John, and Saul Estrin. 1998. “Privatization,Competition, and Budget Constraints: Dis-ciplining Enterprises in Russia.” WorkingPaper No. 128, Stockholm Institute of Tran-sition Economics, Stockholm.

Frydman, Roman, Cheryl W. Gray, MarekHessel, and Andrzej Rapaczynski. 1997.“Private Ownership and Corporate Perfor-mance: Evidence from Transition Econo-mies.” EBRD Working Paper No. 26, Eu-ropean Bank for Reconstruction andDevelopment, London.

_____. 1999. “When Does Privatization Work?The Impact of Private Ownership on Cor-porate Performance in the TransitionEconomies.” Quarterly Journal of Eco-nomics 114(4): 1153–92.

Lieberman, Ira W., Stilpon S. Nestor, and Raj M.Desai, eds. 1997. Between State and Mar-ket: Mass Privatization in Transition Econo-mies. Washington D.C.: The World Bankand the OECD (Organisation for EconomicCo-operation and Development).

Megginson, William E., and Jeffry M. Netter.2001. “From State to Market: A Survey ofEmpirical Studies on Privatization.” Jour-nal of Economic Literature 39(2): 321–89.

Barter, Arrears, and Noncash Payments

Commander, Simon, and Christian Mumssen.1998. “Understanding Barter in Russia,”

EBRD Working Paper No. 37, EuropeanBank for Reconstruction and Development,London.

Gaddy, Clifford, and Barry Ickes. 1999. Russia’sVirtual Economy. Washington, D.C.:Brookings Institution.

Ivanova, Nadezhda, and Charles Wyplosz. 1999.“Arrears: The Tide that is Drowning Rus-sia.” Working Paper, Russian-EuropeanCentre for Economic Policy, Moscow.

Johnson, Simon, Daniel Kaufmann, JohnMcMillan, and Christopher Woodruff. 2000.“Why Do Firms Hide? Bribes and UnofficialActivity After Communism.” Journal of Pub-lic Economics 76(3): 495–520.

McKinnon. Ronald I. 1991. The Order of Eco-nomic Liberalization: Financial Control inthe Transition to a Market Economy. Balti-more, MD: John Hopkins University Press.

Tanzi, Vito, ed. 1992. Fiscal Policies in Econo-mies in Transition. Washington, D.C.: In-ternational Monetary Fund.

Woodruff, David. 1999. Money Unmade: Bar-ter and the Fate of Russian Capitalism.Ithaca, NY: Cornell University Press.

Social Safety Nets and Reform

Commander, Simon, and Fabrizio Coricelli, eds.1995. Unemployment, Restructuring, andthe Labor Market in East Europe and Rus-sia. Washington, D.C.: World Bank.

Cook, Linda, Mitchell Orenstein, and MarilynRueschemeyer, eds. 1999. Left Parties andSocial Policy in Postcommunist Europe.Boulder, CO: Westview Press.

Graham, Carol. 1995. “The Politics of SafetyNets.” In Larry Diamond and Marc F.Plattner, eds., Economic Reform and De-mocracy. Baltimore, MD: Johns HopkinsUniversity Press.

Kornai, János. 1997. “Reform and the WelfareSector in the Post-Socialist Countries: A


122

Normative Approach.” In Joan Nelson,Charles Tilly, and Lee Walker, eds., Trans-forming Post-Communist Political Econo-mies. Washington, D.C.: National Acad-emy Press.

Kramer, Mark. 1997. “Social Protection Policiesand Safety Nets in East-Central Europe:Dilemmas of the Post-Communist Trans-formation.” In Ethan Kapstein and MichaelMandelbaum, eds., Sustaining the Transi-tion: The Social Safety Net inPostcommunist Europe. New York, NY:Council on Foreign Relations.

Milanovic, Branko. 1998. Income, Inequality,and Poverty during the Transition fromPlanned to Market Economy. Washington,D.C.: World Bank.

World Bank. 2000. Making Transition Work forEveryone: Poverty and Inequality in Eu-rope and Central Asia. Washington, D.C.

Accession to the European Union

Begg, Iain. 1999. “EU Investment Grants Re-view.” World Bank Technical Paper No.435, World Bank, Washington, D.C.

Drulák, Petr, ed. 2001. National and EuropeanIdentities in EU Enlargement: Views fromCentral and Eastern Europe. Prague: In-stitute of International Relations.

European Commission and the World Bank.1998. European Union Accession: TheChallenges for Public Liability Manage-ment in Central Europe. Washington, D.C.

_____. 2000. European Union Accession: Op-portunities and Risks in Central EuropeanFinances. Washington, D.C.

Funck, Bernard, and Lodovico Pizzati, eds. 2001.Labor, Employment, and Social Policies inthe EU Enlargement Process. Washington,D.C.: World Bank.

Garibaldi, Pietro, Mattia Makovec, andGabriella Stoyanova. 2001. “From Tran-sition to EU Accession: The Bulgarian La-bor Market during the 1990s.” World Bank

Technical Paper No. 494, World Bank,Washington, D.C.

Kaminski, Bartlomiej. 1999a. “Hungary: ForeignTrade Issues in the Context of Accessionto the EU.” World Bank Technical PaperNo. 441, World Bank, Washington, D.C.

_____. 1999b. “The Role of Foreign Direct In-vestment and Trade Policies in Poland’sAccession to the European Union.” WorldBank Technical Paper No. 442, WorldBank, Washington, D.C.

Nabli, Mustapha. 1999. “Financial Integration,Vulnerabilities to Crisis, and EU Accessionin Five Central European Countries.”World Bank Technical Paper No. 439,World Bank, Washington, D.C.

Nunberg, Barbara. 2000. “Ready for Europe:Public Administration Reform and Euro-pean Union Accession in Central and East-ern Europe.” World Bank Technical PaperNo. 466, World Bank, Washington, D.C.

Petkova, Ivanka, Elitsa Markova, Irena Mladenova,and Ruslan Stefanova, eds. 2000. RegionalCooperation in Central and Eastern Europe.Sofia: Economic Policy Institute.

Tang, Helena, ed. 2000. Winners and Losers ofEU Integration: Policy Issues for Centraland Eastern Europe. Washington, D.C.:Bertelsmann Foundation and the WorldBank.

Ulgenerk, Esen, and Leila Zlaoui. 2000. “FromTransition to Accession: Developing Stableand Competitive Financial Markets in Bul-garia.” World Bank Technical Paper No.473, World Bank, Washington, D.C.

World Bank. 1997. “Poland—Reform andGrowth on the Road to the EU.” Countrystudy, Washington, D.C.

_____. 1998. “Slovak Republic—A Strategy forGrowth and European Integration.” Coun-try study, Washington, D.C.

_____. 1999a. “Slovenia—Economic Transfor-mation and EU Accession.” Country study,Washington, D.C.


123

_____. 1999b. “Hungary—On the Road to theEuropean Union.” Country study, Wash-ington, D.C.

_____. 1999c. “Estonia—Implementing the EUAccession Agenda.” Country study, Wash-ington, D.C.

_____. 1999d. “Czech Republic—Toward EUAccession.” Country study, Washington,D.C.

_____. 2000. “Progress Toward the Unificationof Europe.” Paper for the World Free ofPoverty series, Washington, D.C.

_____. 2001. “Bulgaria—The Dual Challenge ofTransition and Accession.” Washington, D.C.

The Politics of Economic Reform in TransitionEconomies

Balcerowicz, Leszek. 1994. “UnderstandingPostcommunist Transitions.” Journal ofDemocracy 5(4): 75–89.

_____. 1995. Socialism, Capitalism, Transforma-tion. Budapest: Central European Univer-sity Press.

Bunce, Valerie. 1999. “The Political Economyof Post Communism.” Slavic Review 58(4):756–93.

Ekiert, Grzegorz, and Jan Kubik. 1998. “Con-tentious Politics in New Democracies: Hun-gary, the former East Germany, Poland, andSlovakia.” World Politics 50(4): 547–81.

Elster, Jon, Claus Offe, Ulrich K. Preuss, FrankBoeuher, Ulrike Goetting, and Friedbert W.Rueb. 1998. Institutional Design in Post-communist Societies: Rebuilding the Shipat Sea. Cambridge, MA: Cambridge Uni-versity Press.

Fish, M. Stephen. 1998. “The Determinants ofEconomic Reform in the Post-CommunistWorld.” East European Politics and Soci-eties 12(1): 31–78.

Greskovits, Bela. 1998. The Political Economyof Protest and Patience: East Europeanand Latin American Transformations

Compared. Budapest: Central EuropeanUniversity Press.

Hellman, Joel S. 1998. “Winners Take All: ThePolitics of Partial Reform in PostcommunistTransitions.” World Politics 50(2): 203–34.

Offe, Claus. 1991. “Capitalism by DemocraticDesign? Democratic Theory Facing theTriple Transition in East Central Europe.”Social Research 58(4): 893–902.

Przeworski, Adam. 1991. Democracy and theMarket: Political and Economic Reforms inEastern Europe and Latin America. Cam-bridge, MA: Cambridge University Press.

Roland, Gerard. 2000. Transition and Econom-ics: Politics, Markets, and Firms. Cam-bridge, MA: MIT Press.

Shleifer, Andrei, and Daniel Treisman. 2000.Without A Map: Political Tactics and Eco-nomic Reform in Russia. Cambridge, MA:MIT Press.

Corruption, State Capture, and the Rule of Law

Frye, Timothy. 1997. “Contracting in theShadow of the State: Private ArbitrationCommissions in Russia.” In Jeffrey D.Sachs and Katharina Pistor, eds., The Ruleof Law and Economic Reform in Russia.New York, NY: Westview Press.

Frye, Timothy, and Andrei Shleifer. 1997. “TheInvisible Hand and the Grabbing Hand.” TheAmerican Economic Review 87(2): 354–8.

Hay, Jonathan R. and Andrei Shleifer. 1998.“Private Enforcement of Public Laws: ATheory of Legal Reform.” American Eco-nomic Review 88(2): 398–403.

Hendley, Kathryn, Peter Murrell, and RandiRyterman. 2000. “Law, Relationships, andPrivate Enforcement: Transactional Strat-egies of Russian Enterprises.” Europe-AsiaStudies 52(4): 627–56.

Hellman, Joel S., Geraint Jones, and DanielKaufmann. 2000. “Seize the State, Seizethe Day: State Capture, Corruption, and


124

Influence in Transition.” Policy ResearchWorking Paper No. 2444, World Bank,Washington, D.C.

Johnson, Simon, Daniel Kaufman, and AndreiShleifer. 1997. “The Unofficial Economyin Transition.” Brookings Papers on Eco-nomic Activity 27(2): 159–239.

Johnson, Simon, Daniel Kaufmann, and PabloZoido-Lobaton. 1998. “Government inTransition: Regulatory Discretion and theUnofficial Economy.” American EconomicReview 88(2): 387–92.

Johnson, Simon, John McMillan, and C. Woo-druff. 2000. “Entrepreneurs and the Or-dering of Institutional Reform—Poland,Slovakia, Romania, Russia, and UkraineCompared.” Economics of Transition8(1): 1–36.

La Porta, Rafael, Andrei Shleifer, Robert W.Vishny, and Florencio Lopez-de-Silanes.1998. “Law and Finance.” Journal of Po-litical Economy 106(6): 1113–55.

Sachs, Jeffrey D., and Katharina Pistor, eds. 1997.The Rule of Law and Economic Reformin Russia. Boulder, CO: Westview Press.

Shleifer, Andrei, and Robert Vishny. 1998. TheGrabbing Hand: Government Pathologiesand Their Cures. Cambridge, MA: HarvardUniversity Press.

Volkov, Vladimir. 1999. “Violent Entrepreneur-ship in Post-Communist Russia.” Europe-Asia Studies 51(5): 741–54.

World Bank. 2000. Anticorruption in Transition:A Contribution to the Policy Debate.Washington, D.C.

Classifying Political Regimes in Transition

Kitschelt, Herbert, Zdenka Mansfeldova,Radoslaw Markowski, and Gabor Toka.1999. Post-Communist Party Systems:Competition, Representation, and Inter-

Party Cooperation. Cambridge, MA: Cam-bridge University Press.

Linz, Juan J., and Alfred C. Stepan. 1996. Prob-lems of Democratic Transition and Con-solidation: Southern Europe, SouthAmerica, and Post-Communist Europe,Part IV. Baltimore, MD: Johns HopkinsUniversity Press.

Roeder, Philip G. 1994. “Varieties of Post-So-viet Authoritarian Regimes.” Post-SovietAffairs 10(1): 61–101.

Governance

Elster, Jon, Claus Offe, and Ulrich K. Preuss.1998. Institutional Design in Post-Com-munist Societies. New York, NY: Cam-bridge University Press.

Offe, Claus. 1991. “Capitalism by DemocraticDesign? Democratic Theory Facing theTriple Transition in East Central Europe.”Social Research 58(4): 865–92.

Przeworski, Adam, and others. 1995. Sustain-able Democracy. Cambridge, MA: Cam-bridge University Press.

Nationalism, Ethnic Conflict, and Secession

Fearon, James D., and David D. Laitin. 1996. “Ex-plaining Interethnic Cooperation.” Ameri-can Political Science Review 90(4): 715–35.

Laitin, David D. 1998. Identity in Formation.Ithaca, NY: Cornell University Press.

Offe, Claus. 1997. “Ethnic Politics in East Eu-ropean Transitions.” Varieties of Transi-tions: The East European and East Ger-man Experience. Cambridge, MA: MITPress.

Vujacic, Veljko. 1996. “Historical Legacies, Na-tionalist Mobilization, and Political Out-comes in Russia and Serbia: A WeberianView.” Theory and Society 25(6): 763–801.

125

The word “processed” describes informally reproduced works that may not be commonly availablethrough library systems.

References

Alesina, Alberto, and Howard Rosenthal. 1995.Partisan Politics, Divided Government,and the Economy. New York, NY: Cam-bridge University Press.

Alesina, Alberto, Nouriel Roubini, and GeraldCohen. 1997. Political Cycles and theMacroeconomy. Cambridge, MA: MITPress.

Aslund, Anders. 2001. “The Myth of OutputCollapse after Communism.” CarnegieEndowment Working Paper No. 18,Carnegie Endowment for InternationalPeace, Washington, D.C.

Aslund, Anders, Peter Boone, and SimonJohnson. 1996. “How to Stabilize: Lessonsfrom Post-Communist Countries.”Brookings Papers on Economic Activity26(1): 217–311.

Balcerowicz, Leszek. 1995. Socialism, Capital-ism, Transformation. Budapest: CentralEuropean University Press.

Berg, Andrew, Eduardo Borensztein, RatnaSahay, and Jeromin Zettelmeyer. 1999.“The Evolution of Output in TransitionEconomies: Explaining the Differences.”IMF Working Paper No. WP/99/73, In-ternational Monetary Fund, Washington,D.C.

Campos, Nauro F., and Fabrizio Coricelli. 2000.“Growth in Transition: What We Know,What We Don’t, and What We Should.”Centre for Economic Policy Research,London.

Carlin, Wendy, Steven Fries, Mark Schaffer, andPaul Seabright. 2001. “Competition andEnterprise Performance in TransitionEconomies: Evidence from a Cross-Country Survey.” EBRD Working PaperNo. 63, European Bank for Reconstructionand Development, London.

Castanheira, Micael, and Vladimir Popov. 1999.“Framework Paper on the PoliticalEconomy of Growth in Russia and


126

Central and Eastern European Countries.”Paper for the Global Development Net-work conference, World Bank, Washing-ton, D.C. November.

Caves, R. 1998. “Industrial Organization andNew Findings on the Turnover and Mobil-ity of Firms.” Journal of Economic Litera-ture 36(4): 1947–82.

de Melo, Martha, and Alan Gelb. 1996. “A Com-parative Analysis of Twenty TransitionEconomies in Europe and Asia” Post-SovietGeography and Economics 37(5): 265–85.

de Melo, Martha, Cevdet Denizer, and AlanGelb. 1996. “From Plan to Market: Pat-terns of Transition.” World Bank, Wash-ington, D.C.

de Melo, Martha, Cevdet Denizer, Alan Gelb,and Stoyan Tenev. 1997. “Circumstancesand Choice: The Role of Initial Conditionsand Policies in Transition Economies.”Policy Research Working Paper No. 1866,World Bank, Washington, D.C.

Dethier, Jean-Jacques, Hafez Ghanem, and EddaZoli. 1999. “Does Democracy Facilitate theEconomic Transition?” World Bank, Wash-ington, D.C. Processed.

Djankov, Simeon, and Peter Murrell. 2000. TheDeterminants of Enterprise Restructuringin Transition: An Assessment of the Evi-dence. Washington, D.C.: World Bank.

Easterly, William. 2000. “It’s Been a Rough Ride,Toto, but I think We’re Back in Kansas.”World Bank, Washington, D.C. Processed.

Eichengreen, Barry, and Christof Ruehl. 2000.“Financial Sector Developments in Transi-tion Economies.” University of Californiaat Berkeley. Processed.

EBRD (European Bank for Reconstruction andDevelopment). 1999. Transition Report1999. London.

_____. 2000. Transition Report 2000. London.

Eurostat. 1998. New Enterprises in CEC in1997. Studies and Research, theme 4.

Available at http://www.europa.eu.int/comm/eurostat.

Fernandez, Raquel, and Dani Rodrik. 1991. “Re-sistance to Reform: Status Quo Bias in thePresence of Individual-Specific Uncer-tainty.” American Economic Review 81(5):1146–55.

Fischer, Stanley, and Alan Gelb. 1991. “The Pro-cess of Socialist Economic Transforma-tion.” Journal of Economic Perpectives5(4): 91–106.

Fischer, Stanley, Ratna Sahay, and Carlos A.Veight. 1998. “How Far is Eastern Europefrom Brussels?” IMF Working Paper No.WP/98/53, International Monetary Fund,Washington, D.C.

Frye, Timothy, and Joel Hellman. 2001. “PoliticalData-Base of Post-Communist Countries.”World Bank, Washington, D.C. Processed.

Freedom House. Various years. Freedom in theWorld: The Annual Survey of PoliticalRights and Civil Liberties. New York.Available at www.freedomhouse.org.

Gaddy, Clifford, and Barry Ickes. 1999. Russia’sVirtual Economy. Washington, D.C.:Brookings Institution.

Haggard, Stephan, and Steven Webb. 1993.“What Do We Know about the PoliticalEconomy of Economic Policy Reform?”World Bank Policy Research Observer8(2): 143–67.

Havrylysyhyn, Oleh, and Ron van Rooden.1999. “Institutions Matter in Transition,but So Do Policies.” Fifth Dubrovnik Con-ference on Transition, Dubrovnic. June.

Hellman, Joel. 1998. “Winners Take All: ThePolitics of Partial Reform.” World Politics50(2): 203.

Hellman, Joel S., Geraint Jones, and DanielKaufmann. 2000. “Seize the State, Seize theDay: An Empirical Analysis of State Cap-ture and Corruption in Transition Econo-mies.” Policy Research Working Paper No.2444, World Bank, Washington D.C.

References

127

Hellman, Joel S., Geraint Jones, DanielKaufmann, and Mark Schankerman. 2000.“Measuring Governance Corruption andState Capture: How Firms and BureaucratsShape the Business Environment in Tran-sition Economies.” Policy Research Work-ing Paper No. 2312, World Bank, Wash-ington D.C.

Hernandez-Cata, Ernesto. 1997. “Growth andLiberalization during the Transition fromPlan to Market.” IMF Staff Papers 44(4):405–29.

Heybey, Berta, and Peter Murrell. 1999. “TheRelationship between Economic Growthand the Speed of Liberalization duringTransition.” Journal of Policy Reform 3(2):121–37.

Hirschman, Albert O. 1963. Journeys TowardsProgress. New York, NY: The TwentiethCentury Fund.

IMF (International Monetary Fund), WorldBank, OECD (Organisation for EconomicCo-operation and Development), andEBRD (European Bank for Reconstructionand Development). 1991. A Study of theSoviet Economy, vol. 3. Washington, D.C.

Jensen, Michael C., and William H. Meckling.1976. “Theory of the Firm: ManagerialBehavior, Agency Costs, and OwnershipStructure.” Journal of Financial Econom-ics 3(4): 305–60.

Johnson, Simon, and Andrei Shleifer. 2001.“Privatization and Corporate Gover-nance.” Paper prepared for the 12th An-nual East Asian Seminar on Economics,June 28–30, National Bureau of EconomicResearch, Cambridge, MA.

Johnson, Simon, Daniel Kaufmann, and AndreiShleifer. 1997. “The Unofficial Economyin Transition.” Brookings Papers on Eco-nomic Activity 27(2): 159–239.

Johnson, Simon, John McMillan, and Christo-pher Woodruff. 2000. “Entrepreneurs andthe Ordering of Institutional Reform—Po-land, Slovakia, Romania, Russia, and

Ukraine Compared.” Economics of Tran-sition 8(1): 1–36.

Johnson, Simon, Peter Boone, Alasdair Breach,and Eric Friedman. 2000. “CorporateGovernance in the Asian Financial Crisis.”Journal of Financial Economics 58(1–2):141–86.

Johnson, Simon, Rafael La Porta, FlorencioLopez-de-Silanes, and Andrei Shleifer.2000. “Tunneling.” Harvard Institute ofEconomic Research Discussion Paper No.1887, Harvard Institute of Economic Re-search, Cambridge, MA.

Klugman, Jeni. 1999. “Financing and Gover-nance of Education in Central Asia.”MOCT–MOST: Economic Policy in Tran-sitional Economies 9(4): 423–42.

Kornai, János. 1986. “The Soft Budget Con-straint.” Kyklos 39(1): 3–30.

Levine, Ross, and Sara Zervos. 1998. “StockMarket, Banks, and Economic Growth.”American Economic Review 88(3): 537–58.

Maddison, Angus. 1982. Phases of CapitalistDevelopment. Oxford, U.K.: Oxford Uni-versity Press.

Martin, Ricardo. 2000. “Revisiting RegressionAnalysis: Initial Conditions, Policies, andGrowth.” Background paper, World Bank,Washington D.C.

Orenstein, Mitchell. 2001. Out of the Red: Build-ing Capitalism and Democracy in Post-Communist Europe. Ann Arbor, MI: Uni-versity of Michigan Press.

Pinto, Brian, Vladimir Drebentsov, andAlexander Morozov. 2000. “Give Macro-economic Stability and Growth in Russiaa Chance: Harden Budgets by EliminatingNon-Payments.” Economics of Transition8(2): 297–324.

Popov, Vladimir. 1998. “Will Russia Achieve FastEconomic Growth?” Communist Econo-mies and Economic Transformation 10(4):421–35.


128

_____. 1999. “Shock Therapy versus Gradual-ism: The End of the Debate.” CarletonUniversity, Ottawa. Processed.

Przeworski, Adam. 1991. Democracy and theMarket: Political and Economic Reformsin Eastern Europe and Latin America. NewYork, NY: Cambridge University Press.

Roubini, Nouriel, and Jeffrey Sachs. 1989. “Po-litical and Economic Determinants of Bud-get Deficits in the Industrial Democra-cies.” European Economic Review 33(5):903–33.

Ruehl, Christof, and Viatcheslav Vinogradov.2001. “A Simple Model of the Transitionfrom Central Planning.” Center for Eco-nomic Research and Graduate Educationof Charles University and the EconomicInstitute of the Academy of Sciences of theCzech Republic, Prague. Processed.

Selowsky, Marcelo, and Ricardo Martin. 1998.“Policy Performance and Output Growthin the Transition Economies.” AmericanEconomic Review–Papers and Proceedings87(2): 350–53.

Shleifer, Andrei, and Robert Vishny. 1997. “ASurvey of Corporate Governance.” Jour-nal of Finance. 52(2): 737–83.

Shleifer, Andrei, and D. Wolfenson. 2000. “In-vestor Protection and Equity Markets.”

Harvard University and University ofMichigan. Processed.

Tarr, David. 1994. “The Terms of Trade Effectsof Moving to World Prices on Countriesof the FSU.” Journal of Comparative Eco-nomics 18(1): 1–24.

UNICEF–ICDC (United Nations Children’s Fund–International Child Development Centre).1999. “TransMONEE.” Database.

World Bank. 1998. World Development Report1998/99: Knowledge for Development.Washington, D.C.

_____. 1999a. Czech Republic: Capital MarketReview. Washington, D.C.

_____. 1999b. “Russia: Regional Education Study.”Human Development Sector Unit, Europeand Central Asia Region, Washington, D.C.

_____. 2000a. Poverty Reduction and the WorldBank: Progress in Fiscal 1999. Washing-ton, D.C.

_____. 2000b. Making Transition Work for Ev-eryone: Poverty and Inequality in Europeand Central Asia. Washington D.C.

_____. 2000c. Anticorruption in Transition: AContribution to the Policy Debate. Wash-ington, D.C.

_____. 2001. World Development Indicators,2001. Washington, D.C.

Documents

Transition The First Ten Years - ekonomiks.weebly.comekonomiks.weebly.com/uploads/5/5/1/5/5515573/world_bank_2002.pdf · Transition The First Ten Years ... Learning from China?