03367-9781462311187.Pdf

IMFstaffpapers Special Issue Transition Economies: How Much Progress?

Ratna Sahay, Oleh Havrylyshyn, and Rick Haas Co-Editors

Sean M. Culhane and Marina Primorac Assistant Editors

Rafael Romeu ResearchAssistant

Rosalind Oliver Administrative Coordinator

IMF StaffPapers seeks to publish high quality research produced by 1MF staffand invited guests on a variety of topics of interest to a broad audience, including academics and policy makers in IMF member countries. The papers selected for publication in the journal are sub ject to an extensive review process using both internal and external referees. JMF Staff Papers also welcomes outside comments, criticisms, and interesting replications of published work. The views presented in published papers are those of the authors and do not necessari ly reflect the position of the Executive Board or of the IMP.

Subscription: US$56.00 a volume or the approximate equivalent in the currencies of most countries. Three issues constitute a volume. Single copies may be purchased at $18. The indi vidual academic rate to fulJ-time professors and students of universities and colleges is $28 a volume. Subscriptions and orders should be sent to:

International Monetary Fund Publication Services 700 19thStreet, N.W. Washington, D.C. 20431, U.S.A.

Telephone: (202) 623-7430 Fax: (202) 623-7201 E-mail: [email protected] Internet: http://www.imf.org

©International Monetary Fund. Not for Redistribution lnternorionol Monetary Fund IMFstaffpapers

Volume48 Transition Economies:

HowMuchProgress?2001

©International Monetary Fund. Not for Redistribution EDITOR'S NOTE

The Editor invites from contributors outside the IMF brief comments (not more than 1,000 words) on published articles in IMF Staff Papers. These comments should be addressed to the Editor, who will forward them to the author of the original article for reply. Both the comments and the reply will be considered for publication. The data underlying articles published in IMF Staff Papers are available on the journal's website (http://www.imf.org/staffpapers). Readers are invited to use these data to expand on the material in the articles, and the journal will consider publishing such work.

This serial publication is catalogued asfollows:

International Monetary Fund IMF staff papers- International Monetary Fund. v. 1- Feb. J 950- [Washington] International Monetary Fund.

v. tables. diagrs. 26 em.

Three no. a year. 1950-1977: four no. a year. 1978-

Indexes: Vols. 1-27. 1950-80. I v. ISSN I 020-7635 = IMF staff papers -International Monetary Fund. I. Foreign exchange-Periodicals. 2. Commerce-Periodicals. 3. Cwrency question-Periodicals.

HG3810.15 332.082 53-35483

©International Monetary Fund. Not for Redistribution International Monetary Fund

. " contents 2001

Ten Years of Transition: Looking Back and Looking Forward Stanley Fischer • 1

The Transition in Central and Eastern Europe: The Experience of Two Resident Representatives Mark Allen and Rick Haas • 9

Ten Years After ... Transition and Economics

Gerard Roland • 29

Recovery and Growth in Transition: A Decade of Evidence

Oleh Havrylyshyn • 53

Escaping the Under-Reform Trap

Anders Aslund, Peter Boone, and Simon Johnson • 88

What Moves Capital to Transition Economies? Pietro Garibaldi. Nada Mora, Ratna Sahay, and Jeromin Zettelmeyer • 1 09

The Gains from Privatization in Transition Economies: Is "Change of Ownership" Enough?

Clifford Zinnes, Yair Eliot. and Jeffrey Sachs • 146

Federalism With and Without Political Centralization: China Versus Russia

Olivier Blanchard and Andrei Shleifer • 171

Falling Tax Compliance and the Rise of the Virtual Budget in Russia Brian Aitken • 180

©International Monetary Fund. Not for Redistribution lhispage intentionallyleft hlank

Letter From the Editor

In the first article in this special issue Stanley Fischer provides a unique insider's overview of both the papers in the issue and the world-changing events that led the papers to be written. The work repot1ed here was done by well-known academics and excellent economists who were on the ground during what may be the greatest economic evenl of the last century. I will let the papers and authors speak for themselves.

The editorial burden of d1is issue of IMF StaffPapers has been carried entirely by Ratna Sahay, Oleh Havrylyshyn, and Richard Haas. I thank them for making this issue possible. t

This is our second type of special issue. The first is the IMF Annual Research Conference issue that we now publish regularly. In addition to the highly policy relevant contributions of the experts in the transition economies field, the current special issue represents our effort to bring to our readers the working experience of lMF staff. It is the first time we have tried it, and I think it is a big success. Oleh Havrylyshyn and Richard Haas brought the idea to me in the early days of my time as editor. Working with them has been a pleasure and a great learning experience. I look forward to staff proposing similar projects in the future.

Robert P. Aood Editor, IMF StaffPapers

©International Monetary Fund. Not for Redistribution lhispage intentionallyleft hlank

Ten Years of Transition: Looking Back and Looking Forward

STANLEY FISCHER

he rise and fall of central planning will be one of the great themes of twentieth century economic history. At its peak, almost half the world's population lived Tin economies that described themselves as being centrally planned, and not even the most hostile critics of the system predicted bow quickly it would begin to unravel in the late 1 980s. But unravel it did, and over the succeeding decade dozens of countries began the often painful transition to a predominantly market economy. Many have traveled very far down that road-so much so that it is increasingly inappropriate, at the beginning of the twenty-firstcentury, to still label them "tran sition" economies. Cuba and North Korea are the only significant economies that still profess allegiance to state ownership and central planning. It is a good time, therefore, even to look back at what we have learned from the transition experi ence and to look forward to the challenges that now face these economies. This issue of IMF StaffPapers contains eight papers on different aspects of the first 10 years of transition. As a collection of papers, it cannot provide a complete overview of what happened in the fust decade, nor of the key issues that arose, nor of the issues that will need to be addressed in future.1 But it does provide insights

'Several conferences on the first decade of transition were held in the last few years. A Decade in Transition: Achievements and Challenges, ed. by Oleh Havrylyshyn and Saleh M. Nsouli (Washington: IMF. 2001) contains the proceedings of an lMF conference held February 1-3, 1999. The American Economic Association organize{! a panel discussion on "A Decade of Transition'' at its January 2000 Annual Meeting. And in Poland CASE organized a conference on "Ten Years After: Transition and Growth in Post-Communist Countries," on October 15-16, 1999. The proceedings of this conference were published in Transition and Growth i11 Post-Communist Countries: The Ten-Year Experience, ed. by

Lucjan T. Orlowski (Northampton, Massachusetts: Edward Elgar, 2001). See also the World Bank report, Transition-The First Ten Years: Analysis and Lessons for Eastern Europe and the Soviet Union (Washington: World Bank, 200 I).

©International Monetary Fund. Not for Redistribution Stanley Fische

into many of the key issues. The fascinating paper by Mark Allen and Rick Haas, who were IMP resident representatives in transition economies, describes the mindset of people in economies setting out on the path of transition; it would serve well as a standard reference on the market economy for an introductory course in economics. Rather than attempt to summarize the other seven papers, I will refer to them as appropriate in the remainder of this introduction.

I. Performance During Transition

Although important reforms began in the 1960s in Hungary and in the l980s in China, Poland's economic reform program in 1989 was the first attempt at a comprehensive rapid move to a market economy. No one believed that the process would be painless. It was anticipated that output and employment would fall initially as macroeconomic stabilization took place, and as structural reforms moved resources from unproductive to productive uses. But the adjustment turned out to be much more painful than expected. In the transition economies of Central and Eastern Europe, the Baltics, Russia, and the other countries of the former Soviet Union, output fell by more than 40 percent on average. There are well-known reasons to believe these data exaggerate the real output loss, but there can be no doubt that transition outside east Asia was accom panied by severe dislocations, large redistributions of income, and severe income losses by many people. Both the depth of the initial recession and the strength of the subsequent recovery-it generally took about two years for growth to follow stabilization-varied widely from country to country. Only a handful have seen output recover to pretransition levels. In contrast, the transition economies of east Asia managed to maintain strong growth throughout the transition period, aUowing them to enjoy big cumulative gains in output. Econometric analyses of the factors determining economic performance in the transition economies of Europe and the former Soviet Union are summarized in the paper by Oleh Havrylyshyn. This paper, which makes use of the measures of structural reform developed by the EBRD, conftrms some striking lessons. Growth was affected by the initial conditions affectingthe economy, including for instance how long it had been in the Soviet bloc and its dependence on previous trade patterns. But while initial conditions were important in determining the size of the initial recession, they had less influence on the subsequent recovery path. Here and this is the key point-the strength of reform efforts was the dominant factor. Indeed, the evidence confums that the basic strategy advocated by market-oriented proponents of reforma decade ago was correct, namely, that both stabilization and structural reforms-particularly price liberalization and the small-scale privati zation of state enterprises-contribute significantly to growth. The fa ster these refo rms are implemented-and the more consistently they are pursued-the quicker the economy pulls out of the initial recession and the stronger the subse quent upturn in growth. In addition to falls in output, the early phases of transition were marked by big price increases in most countries, the results of price liberalization in the presence

©International Monetary Fund. Not for Redistribution TEN YEARS OF TRANSITION: LOOKING BACK AND LOOKING FORWARD

of inherited monetary overhangs. Falling output and fiscal restructuring widened budget deficits, which in tum were monetized, fueling inflation. The resulting macroeconomic instability was particularly virulent and persistent in the CIS countries. With the help of IMP-supported programs, in most-but not all-cases using exchange rate nominal anchors, most transition economies had achieved reasonable price stability by the mid-1990s. However, there were setbacks in several countries later in the decade, most of them associated with the Russian crisis in 1998. The Russian crisis was a result of external shocks, capital flight, a weak financial system, failures to undertake struc tural reforms, and an extremely weak fiscal situation, which had rapidly produced an unmanageable debt servicing burden-all this in the context of a crawling peg exchange regime that had performed well in stabilizing inflation, but that the authorities were loath to give up in time. The Russian fiscal crisis was associated with the seemingly intractable nonpayments problem that beset both the fiscal accounts and other sectors of the economy, an issue taken up by Brian Aitken in his paper. The fiscal crisis was also associated with the loss of revenues by the center to the regions, a process that forcefully raised the issue of fiscal feder alism-which is the subject of the paper by Blanchard and Shleifer, and is of great importance also in China and other large countries. The Russian crisis in 1998 hit the other CIS countries very hard. But across the transition economies as a whole, the situation-with the help of rapid growth in Russia-is now clearly improving. In the past two years, growth in the Central and East European, Baltic, and CIS transition countries has rebounded to 5.5 percent and inflation has more than halved to less than 20 percent, with most coun tries enjoying single-digit inflation. So in terms of macroeconomic stabilization, much has been achieved. But experience has differed widely from country to country-depending in large part on policy effort. The same is true of structural reform. Progress has generaJly been most impressive in central Europe and the Balkans, and most disappointing in the CIS and east Asia. It has also varied widely from one policy area to another. Reform is most advanced in the privatization of small-scale enterprises, the elimination of price controls, and the liberalization of foreign trade and exchange. But less bas been achieved in the regulation and supervision of financial sectors, enterprise restructuring, and the reform of governance in the public and private sectors. The paper by Zinnes, Eilat, and Sachs provides evidence on the importance of building institutions for privatization to succeed. In countries where certain key structural reforms were delayed-notably the elimination of soft budget constraints and the Liberalization of trade and prices inflation has tended to be higher and more protracted and the recovery in output has been less robust. The most successful strategy seems to have involved rapid privatization of small-scale enterprises, followed by more gradual privatization of large-scale enterprises-once commercial legislation has been put in place to ensure sound corporate governance structures and adequate competition (or regu lation in the case of natural monopolies). As shown by Garibaldi, Mora, Sahay, and Zettelmeyer, the more successful reformers have further strengthened their position by gaining rapid access to

©International Monetary Fund. Not for Redistribution Stanley Fischer

foreign capital markets and direct investment, thereby reducing their reliance on financing from governments and international fmancial institutions. At the other extreme, the problems of the less successful reformers-notably Russia-were exacerbated by capital flight. Much of the evidence in Havrylyshyn's paper and the discussion so far bears on the choice between gradualism and shock therapy, which received a great deal of attention early in the transition process. It was Poland that most clearly elected to do as much as possible as rapidly as possible-though even in that case, priva tization Jagged a great deal-and until growth began in Poland in about 1993, critics of shock therapy seemed to have the upper hand. Later it became clear both from case studies and from econometric evidence that the countries that had moved faster started to grow sooner. It also became clear that different aspects of reform had to be carried out at different speeds-for instance, that while stabi lization could be effected relatively quickly, the creation of the institutions of a market economy, including important aspects of corporate governance, would take much longer. As Yegor Gaidar makes clear in his memoirs,2 in practice the choice between gradualism and shock therapy was very much one of political economy. Careful sequencing, of the type that leads Gerard Roland in his paper in this issue to recommend the gradualist East Asian approach, assumes a degree of political control of the reform process that was simply not available in many of the transi tion economies, especially in the CIS countries. One frequent criticism of the reform strategy in the transition economies-and of Western support for it-is that too little thought was given to institution building, for example, of sound legal and regulatory systems and well-functioning tax administrations. The problem was not one of attention or thought, but of execution. The importance of these elements was recognized from the beginning, in the joint study of the Soviet economy published by the leading international financial institutions in 1991,3 and a massive technical assistance effort was mounted by both the international institutions and on a government-to governmentbasis to try to build the necessary institutions. But such changes were too slow to go into effect. No doubt this was partly because the difficulties involved in implementing institutional reforms were underestimated. There was a lack of experienced personnel and the relevant administrative capacity in many countries was limited-and the resistance of existing bureaucracies was very effective. Much still remains to be done in completing the necessary institutional reforms.

II. Why Is Asia Different?

Before turning to policy challenges for the future, it is worth considering further why the experience of the East Asian transition economies-in particular

2Yegor Gaidar, Days of Defeat and Victory, trans. by Jane Ann Miller (Seattle, Washington: University of Washington Press, 1999). 3A Study of the Soviet Economy (Paris: IMF, World Bank, OECD, EBRD, 1991).

©International Monetary Fund. Not for Redistribution TEN YEARS OF TRANSITION: LOOKING BACK AND LOOKING FORWARD

China-has been so different from that of the countries in the former Soviet bloc. Most indicators suggest that progress on structural reform in East Asia has been relatively modest, yet output performance has been far superior to even the best reformers in Europe and the CIS. Some observers and analysts-including, as previously noted, Gerard Roland-have argued that the Chinese experience shows that a gradualist approach to reform is better than a rapid one. Others have argued that China's success has come about despite gradualism rather than because of it, and that the strong output performance there reflects favorable initial conditions and political stability. It is certainly true that one of the most conspicuous differences between the transition process in East Asia and elsewhere is that in the former it was under taken by incumbent governments, rather than coinciding with a period of political upheaval. The reform process in China began in the late 1970s, with very rapid changes in the agricultural sector. The present objective of developing a socialist market economy was not formalJy adopted until 1993. Reforms have been gradual but far-reaching. They have remained on track in part because they were initiated and implemented in a stable political climate, a luxury not on offer elsewhere. The East Asian transition economies were also able to increase output by drawing on a large pool of surplus labor in the agricultural sector. By permitting greater private sector activity in agriculture and relaxing entry into industry in rural areas. it was possible to increase output substantially without employment falling sharply. In contrast, in most other transition economies, adjustment required large-scale job shedding from big state-owned enterprises. But China's success has been secured at a price. Reform of large state enter prises has been relatively slow, contributing to a serious worsening of their finan cial position. This in turn has resulted in a substantial accumulation of nonperforming loans in the state-owned commercial banks. The authorities have made significant progress in privatizing small enterprises and reducing the level of nonperforming loans in state banks, but much remains to be done to restructure large enterprises and to develop a sound and market-oriented commercial banking system. Accession to the World Trade Organization (WTO) means that these prob lems will have to be tackled with increasing vigor in the future. This in turn will impose fiscal burdens, underlining the need to strengthen the country's budgetary position.

Ill. The Challenges Ahead

What about the challenges facing the other transition economies? Consider first the relatively successful reformers on the European Union accession track; then Russia and Ukraine; and finally the other countries of the CIS. The EU bas accepted 10 transition economies as full candidates for acces sion-Bulgaria, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, the Slovak Republic, and Slovenia. Many of them have been relatively successful reformers, which many observers attribute to their underlying historical affinity to Western Europe. With central planning introduced to these countries only in the late 1940s, there was still considerable understanding of how a market

©International Monetary Fund. Not for Redistribution Stanley Fischer

economy works. The political upheavals of the late 1980s were seen as an oppor tunity to return to a Western European modeL With a national consensus in favor of rejoining Europe, the incentives provided by the EU accession process created a powerful external anchor that has helped keep reform on track, and should continue to do so. Economic growth has resumed in most of the accession economies, while inflation is moderate or declining. However, relatively large budget and current account deficits remain a source of concern in some economies. On the structural side, for the most part these countries have successfully liberalized trade and foreign exchange systems and privatized many small- and large-scale public enter prises. There has also been significant progress in liberalizing prices. But there has been less progress on competition policy, governance, enterprise restructuring, and the development of financial institutions. In 1999 the European Commission said that 6 of the 10 accession economies in effect had functioning market economies. Lithuania and the Slovak Republic were "close" and Bulgaria had made good progress from a poor starting point, but the situation in Romania remained "very worrying." Even the best performing accession economies have policy challenges to confront if they are to make the most of EU membership and minimize its risks. For example, all applicants still need to see significant adjustments in their indus trial and occupational structures, trade patterns, and financial sectors. These will occur more smoothly and at lower cost if measures are taken to increase the flex ibility of product and labor markets. More also needs to be done to strengthen the institutions that support market activity: in particular, concerns about judicial systems, banking system weaknesses, and corruption need to be addressed. Outside the EU accession group, the remaining transition economies face an even more difficult set of challenges. Although almost all have achieved a reason able degree of macroeconomic stability, rudimentary financial systems complicate the task of setting monetary policy, while high government spending and ineffi cient tax collection put upward pressure on budget deficits. Fiscal problems are exacerbated by the nontransparent nature of some spending, notably implicit subsidies to existing enterprises. And structural reforms lag far behind those in the EU accession countries Among these countries, attention naturally focuses most intensely on Russia and Ukraine. Political factors, including the capture of important elements of the reform process by opponents of reform, bulk large in determining the future of reform. In these countries, there is not much doubt about what needs to be done, and in each country, policymakers frequently proclaim their intention to reform. With growth having returned to both countries in 2000, reform should have become politically easier to implement. The tax reform passed in Russia in 2000 is a major step forward; so would be the Gref reform program if it were imple mented. Similarly, the Ukrainian program supported by the IMF at the end o£2000 provides a clear road ahead. Whether those reforms are pursued remains to be seen. It is in the interests of the Russian and Ukrainian peoples that those reforms be pursued. And given the importance of those two economies-particularly the

©International Monetary Fund. Not for Redistribution TEN YEARS OF TRANSITION: LOOKING BACK AND LOOKING FORWARD

Russian economy-in the CIS, reform and the beginning of sustained growth in Russia and Ukraine would also help the other countries in the CIS, in most of which economic performance has been very poor. The precise mix of reforms required in the CIS countries-as well as in the east Asian transition countries--differs from country to country, bur there are common challenges in three areas.

• Enterprise re structuring: Inefficiencies and Joss-making activities in large scale enterprises remain a serious drag on economic performance. ln Russia and other CIS economies, the problem lies mainly in newly privatized firms, where there bas been little incentive to restructure; the ability to amass arrears and take advantage of subsidized energy has maintained soft budget constraints. In China and Vietnam, the problem lies more with state-owned enterprises, where borrowing from state-owned banks is financing over capacity and debt accumulation. In both cases, hard budget constraints need to be combined with stronger social safety nets to ease the social costs of adjustment.

• Financial sector re form: In Russia and other CIS economies, the banking system remains small and underdeveloped. Action needs to be taken to deal with those banks that remain insolvent, as welJ as to develop the sector by stimulating wider private ownership and greater competition. Better regula tion and supervision will be required to guard against abuses, including insider lending. Reforms of the legal and regulatory environment are also necessary, but these will inev.itably take some time. In China, the banking system is much bigger and deals with a large propor tion of domestic savings, but channels much of it to state-owned enterprises. Nonperforming loans are being transferred from the four state banks to asset management companies, which will strengthen their financial position. But it will be important to ensure that a fresh bad loan problem does not emerge. This means developing a commercial credit culture in the banks and strength ening regulation and supervision. The prospect of WTO accession underlines the need for action, as it will give foreign banks full national treatment by 2005. • Transforming the ro le of the state: More needs to be done to establish and enforce the rules of the game for a market economy, especially in the CIS and East Asia. This means strengthening tax systems and tax administration; improving fiscal transparency and government accounting; reformiJ1g rela tionships between central and local government; and strengthening the legal system and systems of corporate governance. It will also require continuing reform of expenditure policies, phasing out subsidies, and strengthening social spending progran1s. In many transition economies, progress in these areas is being held up by d1e resistance of vested interests-a point emphasized for the cases of Russia and Ukraine by Aslund, Boone, and Johnson. Overcoming this problem is no easy task. Pressure to do so could come from a variety of sources: strong political lead ership, democratic pressures, pressure from a growing middle class, or pressure from foreigners seeking to compete in domestic markets. The international finan-

©International Monetary Fund. Not for Redistribution Stanley Fischer

cial institutions, representing the international community, also have an important role to play in the advice they give and the reforms they support. Very few countries remain to set out on the path of transition. Yugoslavia is starting now; North Korea and Cuba are likely to follow eventually. The transition strategy for those countries will draw on the lessons learned from the experience of the 26 countries that started the process between 1989 and 1992. Already now. after a decade, many of these countries can no longer be regarded as transition economies, and within the next decade, most of the remaining transition economy members of the 1989-92 group will also have graduated. Increasingly, the prob lems they face will be the same as those confronting other developing and indus trialized economies. And the history of the transition economies will be studied for what it teaches us about economic reform in general, as well as about the history of a particular, important, group of countries.

The Transition in Central and Eastern Europe: The Experience of Two Resident Representatives

MARK ALLEN AND RICK HAAS*

This essay reflects the thoughts and experience of two staff members. Mark Allen has been a resident representative in both Poland (1990-93) and Hunga1y ( 1996-1998), two countries that have moved well along the road of transforma tion. Rick Haas was the resident representative in Belarus (1996-98), a country where much remains to be done. In some respects they had similar experiences, in others very dissimilar ones. In what follows they describe the circumstances they t found themselves in, the advice they gave, the lessons they learned, and they recount some of their experiences as resident representatives. [JEL A13, E20, E50, E65, F15]

rom the earliest days of the IMF, staff members have been assigned to live and work in designated member countries, typically those with an active financial program.F In March 1990 a resident representative post was established in Warsaw, Poland, the first in a transforming economy. The breakup of the Soviet Union significantly expanded the resident representative program; ultimately virtually every transforming economy had a resident representative assigned to it. At most posts (Moscow and Kiev, and for a time, Warsaw, being the excep tions), a single staff member is assigned, typically for two years, assisted by two

*Mark Allen was educated at Cambridge, Yale, and the Karl Marx Higher Economics Institute in Sofia. Bulgaria. He is currently a Deputy Director in the Policy Development and Review Depanment of the IMF. Rick Haas has degrees from Kalamazoo College and Duke University. He is an Advisor in tbe European 11 Department of the 1MF and Mission Chief for Moldova. They would like to thank Rolando Ossowski and JuHa Lyskova for making chaLlenging assignments enjoyable and for providing much help. insight, and, above all, friendship.

©International Monetary Fund. Not for Redistribution Mark Allen and Rick Haas

or three local people. Resident representatives are charged with helping senior local officials formulate and implement policy, providing technical assistance at the working level, gathering data and information for the IMF, and, more gener ally, explaining the work of the fMF to the people of the country to which they have been posted. Resident representatives in transforming economies have had a unique oppor tunity to witness and participate in one of the most interesting and challenging events of the economics profession in the past 50 years: the transformation of centrally planned economies into market based systems. The job is intellectually fascinating, frequently extremely rewarding, occasionally frustrating, but never boring. lt is relatively easy to list the long list of actions needed to transform an economy, and when approached in this manner the task does not seem over whelming, just very difficult. But, when one realizes that what is really being attempted is getting an entire society to change the way it thinks and acts, in a very fundamental way, the task can seem daunting. Yesterday's vices-individual entrepreneurial activity, for example-are today's virtues-something to be rewarded, not punished. When approached from this direction, it is clear that time and patience are required.

I. The Setting We Found Ourselves in

The history of the breakdown of central planning is well known (see Campbell, 1992). Nevertheless, a brief overview is useful to set the stage on which IMF resident representatives found themselves. The old system was not so much a planned as a large administered economy. Enterprises perfom1ed tasks one year because they had performed them the previous year. Coordination between enterprises was largely done through administrative channels. Planners, rather than consumers, were sovereign in the system, and this meant that the po.litical authorities and the associated bureau cracy had complete legitimacy to intervene as they wished to achieve the outcomes they wanted. The price system was loosely based on domestic cost structures and operated more as an accounting mechanism than as a guide to resource allocation. Enterprises paid little attention to profitability as an ultimate objective, but instead concentrated on maintaining or increasing output. The system was characterized by a chronic hunger for imports and foreign exchange. While, in its early days, the system of central planning delivered rapid growth, stagnation set in throughout the region during the 1970s, which led to a general disillusionment about the ability of central planning to deliver prosperity to the population. From the late 1950s, economists throughout the region had tinkered with the planned economy in an attempt to revive growth, but nothing they did restored the growth rates of the early Communist period. Among the results was that, when communism collapsed, there was little popular or intellectual support for the planned economy per se.

©International Monetary Fund. Not for Redistribution THE TRANSITION IN CENTRAL AND EASTERN EUROPE

II. Legacy of the Past

One of the most challenging-and important-things a resident representative bas to do is to understand the mind-set of the local authorities. Jf this is not done, the two parties will simply talk past each other. Linguistic differences are a minor-and easily solved-aspect of this problem; the major hurdles come from the legacy of the old regime. Some of the difficulties are generalized and often include a suspicion and mistrust of foreigners. While this is frequently based on incomplete or inaccurate information (something that can be corrected), it often takes time to establish a good working relationship with officials of the host government. Some of the other difficulties are fairly specific. ln the former Soviet Union, the general suspicion and distrust of foreigners complicated attitudes toward the IMF resident representative. In Central Europe, where foreign influence was more welcome, the translation of IMF resident repre sentative as Fund rezydent occasionally engendered suspicions that the feared KGB residency had been replaced by an equally sinister IMF office. It was an article of faith (shared with many other civil services) that the state has the best interests of the population at heart and should be proactive in defending them. At the core of the matter is a deep-seated belief that the state can always improve on market solutions because truly free prices are the reflection of greed and, as such, are suspect prima fa cie. The concept of the "invisible hand" is unknown, difficult to communicate, and even more difficult for many officials to embrace: it was completely contrary to their instincts. The person who did most to explain the "invisible hand" to both the Poles and other East Europeans may well have been Jeffrey Sachs. He was particularly good at talking to the politicians in Poland and giving them confidence that the market would supply things when the bureaucrats left it alone. He gave them confidence that free prices would give someone somewhere an incentive to find or produce the goods and sell them on the street. He also instilled confidence that ordinary people would have the wherewithal to buy these goods. He just had a knack of explaining the elementary functioning of markets in a simple and convincing way. Some of this is discussed in bjs book, Po land's Jump to a Market Economy. The principal problem was convincing the authorities that markets were a good way to organize economic activity. This was relatively easy in places where the positive results were very visible, such as Central Europe, and in places where there was a deep dislike of the old system, such as the Baltics, and more difficult where the old system was most entrenched. In Belarus, Turkmenistan, and Uzbekistan, it was often relatively easy to get lip service to market ideas but diffi cult to get substantive policies implemented. Where western economists see the orderly function of markets solving economic problems, many Local officials see instability, disorder, and even chaos. And it scares them. Often enough they revert instinctively to direct controls. A Large premium is placed on stability, not in any equilibrium sense of the word, but in the sense of not changing. Indeed, the desirability of equilibrium in markets is often not fuiJy appreciated by officials who have known nothing but queues and shortages but for whom the need for social-and therefore economic-stability is self-evident.

©International Monetary Fund. Not for Redistribution Mark Allen and Rick Haas

The importance of stability surfaces quickly when questions of unemployment arise. One of the boasts of the old regime was that every citizen had a guaranteed job, and they did. Maybe not productive, maybe not value adding, but they had a job. ("The state pretends to pay us and we pretend to work" is a phrase heard early and often.) No doubt guided by a strong desire for social and political stability, the ft rst question many officials ask about any new policy is, "Will it increase unem ployment?" If the answer is yes, even if only temporarily, tile discussion is over. Of course this is not unique to policymakers in transitional economies; officials throughout the world should-and do-ask this question. But the preoccupation with it is unique and detrimental because the essence of transformation requires precisely that many workers in inefficientjobs relocate to new, different jobs. This necessarily implies some, a.lbeit transitory, unemployment. Old ways of thinking also affect policy through a strong desire for equal income distribution, even if such equality decreases incentives to save and invest. This is not unrelated to the desire for social stability. Simply put, the tendency is to devote a lot of effort into making sure the economic pie is sliced up evenly, even if it means that the pie will be smaJJer and many will have less to consume in abso lute terms. An intuitive understanding of markets is often lacking. So is a global (that is, a macro) view of the economy. Tn the old system, the practice was to divide up aU economic problems into a large number of small, individual tasks. Almost no one could visualize the entire system, nor did they need to. A common experience of resident representatives is to be presented with a list of problems when, in fact, the list is not a list of problems at aU but rather a list of undesirable consequences of a single policy. Interrelationships between the exchange rate, exports, imports, and the price level is a case in point. The method of compartmentalizing problems and setting individual tasks has bad another legacy. It has firmly established the view that, if the desired results are not achieved, it is the result of individual failure. It is rarely seen as a policy faiJure or a systemic flaw. In this view, people and personalities are important; policies and principles are not, which is almost exactly the inverse of the resident repre sentative's view. The old system-and the versions of it that still persist-was, in fa ct, flawed; by its very construction it was inconsistent. The individual tasks were invariably overambitious and, indeed, frequently unachievable, which meant, of course, that the plan could not be met in its entirety. There are two very different-and not necessarily incompatible-views on why this happened. The positive view is that setting high targets guaranteed maximum effort; even if the tasks were not totally fulfilled, more would be done than would have been the case with more modest goals. The negative view is that setting unachievable targets means that the author ities will never lack for a reason why the complete plan failed. "K'to vinovat?" (Who is to blame?) is a phrase learned early in the former Soviet Union. Another legacy of the old system was the view toward speculators. Speculation was a criminal activity and punjsbed as such. It requires a Damascene conversion for an official to see currency speculators as Mandeville's bees in a social garden of Eden and not just as Mafia thugs. The notion that speculation can

©International Monetary Fund. Not for Redistribution THE TRANSITION IN CENTRAL AND EASTERN EUROPE

be a stabilizing force and should be tolerated, even encouraged, was one of tbe more difficult concepts to get across. Frequent reference is made to the "mentality of the people," especially in the Commonwealth of Independent States (CIS) countries where no living memory of a market economy remains. By this it is usually meant that, despite the clear need for reforms, they must go slowly, because people-not the policy makers-lack the understanding to move quickly. Our experience was generally the reverse. Anyone who has ever seen how quickly and well a few people can make a freed up plot of land produce vegetables, or how quickly something can be constructed in the right conditions, knows that incentives matter. The mentality at fault is usually that of the officials, who are either overly fearful of the unknown or find the rent-seeking opportunities presented by slow or incomplete reforms too tempting. It was not always clear what the underlying motivation was. Corruption and ignorance are often observationally equivalent: are tax officials and enterprise managers behaving the way they do because they are getting substantial side payments for protecting rent seekers and criminals? Or, do they simply not under stand what needs to be done and why? Of these, ignorance is obviously a lot easier to deal with. Both of us mounted elaborate education campaigns on various levels. First, and foremost, with the authorities, both on the technical level and at the policy level. Second, with the population, through the media and by giving lectures and teaching at the local universities. Corruption and rent seeking, more generally, have to be dealt with by focusing on incentives, good governance, trans parency, and the legal system. In brief, there is frequently a reluctance to believe fuiJy that free markets are a better way of organizing economic activity than the old system.

Our Interlocutors' State of Knowledge of Market Economics

In Poland, and even more so in Belarus, there bad been little formal training in market economics, since it was not a discipline with particular value in the planned economy. The one exception was probably those trained in the foreign trade schools, who had been taught enough economics to allow them to support the marketing activities of foreign trade corporations. This situation meant that relatively few of those who found themselves dealing with the emerging market economy in 1990 had had training in economics as we know it in the West. And. of those few, only a handful really had confidence that a wide range of market solutions was possible. Poland was fortunate that many of that handful were in positions of responsibility. Paradoxically, and even tragically, a number of those who had devoted much of their life to studying market economics, and to promulgating it within the communist system, turned out to be the most strongly opposed to the stabilization and reform program of the Polish government and supported by the IMF. These economists, who had risked being dissidents under the Communist system thanks to their espousal of Western economics, found themselves totally overtaken by events when communism collapsed. Their focus had often been on how to deliver

©International Monetary Fund. Not for Redistribution Mark Allen and Rick Haas

the Third Way, a social democratic approach, that would use a combination of market and intervention to achieve the goals of socialism. in terms of a just and prosperous society. Their ability to fo llow the development of economic thought il1 the West had been severely limited by the sporadic nature of the contact, and for many, the vaunted ability of Keynesian demand management to deliver growth and full employment represented the last word. The more fundamentalist turn that economics had taken in the 1980s and the lessons of heterodox stabilization programs were not part of their experience. They consequently expressed some of the most articulate, if misguided, opposition to the so-called policy of "shock therapy." In Hungary, the situation was a little different. At the start of the transition, Hungary bad more sophisticated economists than anywhere else ill the Soviet bloc. Of course the bureaucracy was full of people who had been trained in the planning tradition, but there was no shortage of economists who understood market economics very well. Here the problem was more that they had spent much of their careers reforming the Hungarian economy. The New Economic Mechanism, incorporating substantial market elements had been introduced in Hungary in 1968, and in the fo llowing years economists in Hungary bad been involved to a man in trying to make this system work more efficiently within the established constraints. Thus, when the transition occurred, most Hungarian economists were generally convinced that they already bad a market economy and that the transi tion would be quite smooth. It took some years to realize how far the sociaUst market economy of the late 1980s was from a true modern market economy and the size of the adjustment effort required to put it on a sustained growth path.

Ill. The Policy Advice We Gave

Standard IMF programs focus on macroeconomic stabilization issues helping a country get its spending and financing needs to realistic and sustainable levels. The transitional economies were no different in needing to do this, but the magnitude of the tasks was arguably larger than the typical case. But it was also clear from the outset that responsible monetary, fiscal, and exchange rate policies were just a part of the required actions in the case of transforming economies where much of what bad to be done illvolved structural change at the most funda mental level. Macro stabilization was necessary, but not sufficient, to achieve tran sition. Appropriate policy advice for transitional economies has been dealt with extensively elsewhere; what fo llows are merely some idiosyncratic observations.'

Conventional Stabilization Issues

The breakdown of the old regime meant the almost immediate breakdown of the revenue collection process, but not a concurrent reduction in expenditures, with the state continuing to provide most of tbe social services it had always provided. This was exacerbated by the state's assumption of the enterprises' social

ISee Bruno, 1992.

©International Monetary Fund. Not for Redistribution THE TRANSITION IN CENTRAL AND EASTERN EUROPE

expenditures made necessary to allow enterprises to concentrate on their core responsibilities. This meant that budget deficits expl.oded. Foreign borrow ing went up, as did money and credit, and an increase in arrears of aU sorts fo llowed: tax arrears; budget arrears; wage and pension arrears; and inter-enterprise arrears. The govemment sector was simply too large and underfinanced for a market-based system. This was often viewed with more concem by IMF staff than by the authori ties. The single lesson of modern economic history taught in the old system was Roosevelt's New Deal and how a large state had spent itself into prosperity. If a local counterpart only knew one thing about western economics-and they all seemed to-it was this oversimplified half-truth of Keynesian economics. The old system relied on a monobank that was both a central bank and a commercial bank. But it was never central to economic policy, rather it largely functioned as an after the-fact bookkeeper. Setting up a true central bank was important. It was equally important that it be an independent central bank. This, in many cases, turned out to be more djfficult than might be thought, because it involved a separation of powers, an alien concept in many transitional countries. A difficult problem encountered early on was to decide how credjt should be allocated. Relying on markets totally and completely could not be done. First, credit analysis was an unknown practice. No one was equipped ro assess· the crerutworthiness of the various fi rms. Information was lacking as well, and where it existed, it could be misleading. Firms that might well be profitable if prices were liberalized might not be when they were not, as was often the case early on. Nor was there any way to evaluate the debts that had sprung up from the absence of a hard budget constraint and how these affected a firm's balance sheet. Another problem arose when credit auctions were used. Auctions in normal circumstances can be an extremely efficient way of allocating credit-or anything else-but in a system where many of the agents are insolvent, there is a very real adverse selection problem. Bankrupt firms, or firms that are confident they will be bailed out if they lose money, have an incentive to bid very high rates for funds. This will squeeze out commercially viable firms subject to hard budget constraints, and the likelihood is that credit extended to these shaky firms will never be repaid. There were few monetarists in transitional economies, at least in the begin ning. The common view was that the decline in activity was the result of insuffi cient money and credit. Indeed, the argument was heard that a decrease in the real money supply (as was commonly observed in inflationary times) was the result of monetary policy being too tight. The idea that prices were related to money was often not intuitive. Rather, the price fo rmation process was viewed as a complex process that involved many technical elements and many factors outside the authorities' control, especially external factors. In fact, tight monetary policy was precisely designed to make the traditional cost-plus-markup pricing policy impossible. This meant that the same situation IMF economists would see as inflationary would often be viewed as too tight a monetary policy setting by the local authori ties. They frequently argued that the "monetary coefficient" (that is, the reciprocal

©International Monetary Fund. Not for Redistribution Mark Allen and Rick Haas

of velocity) was too low, and the level of activity required more-not less-money to support it. Convincing them that increases in money led to increases in prices and that inflation would lead people to hold less money was not easy in the more recalcitrant countries. It is easy to see how it is counterintuitive to many to argue that monetary tightness will lead to an increase in the ratio of money to output. and there was always a Local variant of the real bills doctrine-namely, credit expan sion is not inflationary if it is issued for productive purposes, typically construc tion or agriculture. Even when the money-price nexus was acknowledged, there was no reason to believe that a common policy prescription would emerge. Indeed, the Phillips curve was a concept that was easily grasped by many local authorities. For example. they often had little trouble in deciding where they wanted to be on the curve-the point that minimized unemployment. lnflationary considerations were secondary. The idea that price stability was necessary was difficult to convey, and Fischer, Sahay, and V egh ( 1996) was useful in doing this, especially after it was translated into Russian. Economics professors have always known that foreign exchange topics are among the most challenging to teach. This is certainly true for resident represen tatives and their students, local officials. The relationship between money creation and exchange rate movements was often not properly understood. Generally, the response-almost invariably termed "scientific"-was to argue that the exchange rate should be adj usted by precisely the same amount as inflation. This is, of course. a pw·chasing power parity view. The idea that real exchange rates needed to change in some circumstances was particularly hard to communicate, largely because of a view in wltich price changes were felt to have little effect. The IMF has shown a fairly ecumenical approach to exchange rate systems. It has supported programs in which the rate bas been rigidly fixed and programs where it has been free to float. There are technical arguments on eit11er side. The inescapable fact is that. if the fiscal and monetary fundamentals are not ri ght, it doesn't make any difference which system is chosen; neither will work. On the other hand. if the fundamentals are in order. either system will work reasonably well. Often t11e autllorities were loatll to let rates float, even when reserves were at alarmingly low levels and there was little choice. Market solutions were rarely trusted, but administrative controls in the form of surrender requirements and capital controls were. One view was that a liberalized exchange rate regime could be established only after the central bank was able to hold sufficient foreign exchange reserves (obtained from the IMF) to stabilize the rate long enough to aLLow export receipts to materialize that could be used to repay the IMF. Conspiracy theories of the foreign exchange market were always popular. The argument was that tile nominal rate should be stable between the country and its major trading partners. If it wasn 'l, it was because markets were being corrupted by banks and criminal elements that had a vested interest in destabilizing the rate. A major concern in many countries was the need to set the exchange rate to accommodate energy policy-in particular, the need to keep energy (in local currency) cheap. This was the resuJt of compartmentalizing problems and not viewing the economy as a whole.

©International Monetary Fund. Not for Redistribution THE TRANSITION IN CENTRAL AND EASTERN EUROPE

Structural Issues

Stabilization of the economy and liberalization of transactions are at the center of the IMF's policy role throughout the world. The distinguishing feature of work in the transition economies, however, has been the vastness of the structural change that was needed. Work on structural issues occupied a major part of the resident representatives' time. Some of this work fell clearly into the mainstream of the IMP's activities. Thus much time was devoted to fiscal reform issues, covering the consolidation of accounts, the reform of spending systems, and the establishment of modern tax structures. The struggle was often one of replacing the discretionary systems preferred by the traditional bureaucracy and its clients with transparent, predictable, and nonnegotiable systems. Another area of central JMF concern was the banking system, replacing the monobank with a narrowly defined independent central bank and a number of properly supervised commercial banks. This also proved a difficult area, as the managements of the newly created. or newly sepa rated, commercial banks were faster to exploit their new freedoms than the author ities were to appreciate the need for close supervision, not to mention make such supervision effective. Thus, most transition economies fo llowed the path beaten by so many market economies where the need for good supervision was onJy really appreciated after clocking up losses of 5-10 percent of GOP in bad loans. As resident representatives, trying to help countries through the totality of the economic transition process, we were also involved with trying to ensure that the full agenda of structural reform measures was implemented. Central to this was privatization, an area where the IMF's own expertise was limited. Our approach was usually that the faster enterprises could be privatized, the better, and the more privatization brought in foreign capital, the faster the modernization of the economy would occur. There was considerable local opposition to this approach. The first argument was that, since foreigners were flush with money while resi dents were impoverished, the entire economy would fall into foreign hands. Thus schemes for privatization needed to be devised that would prevent tbis outcome. The second argument was that the nation's patrimony, created by the huge sacri fices of previous generations, would be sold for a song in its current state. Thus, what was needed was an extensive industrial policy designed to upgrade industry to a level where it could be sold for something closer to its value. As resident representatives, our voices were not very influential in these debates. We were quite hostHe to an expanded industrial policy on the grounds that it ran the risk of being a backdoor return to central planning, and we were skeptical that such a policy would not be dominated by special interests. In the event, interventionist industrial policy did not take off, but this probably owed as much to the absence of fiscal resources to finance such a policy as to our advice. The structural reform agenda extended to other areas, which, while clearly essential, fe ll outside our expertise. Land privatization and laws on the use and possession of collateral were clearly vital for a properly functioning market economy, but their technical (and political) ramifications fell outside our orbits. Some observations on the housing and mortgage finance markets are given below.

©International Monetary Fund. Not for Redistribution Mark Allen and Rick Haas

Contract enforcement and predictable dispute settlement systems were recognized as being central to the functioning of a market economy, but not areas where IMP economists had relevant ski1ls.

IV. Emergence of Markets: Successes and Failures

Growth of Markets in Warsaw

One of the most rewarding aspects of the resident representative job in Poland at the start of the transition was to observe markets springing up and evolving. The day to day changes allowed one to observe adjustment in action. As macro economists, we are used to seeing the results of our recommendations reflected slowly, if at all, in daily street scenes. What was so satisfying as resident repre sentatives was to see change happening so quickly and positively, and to see it so clearly as a result of the policies that we were helping the government implement. Conversely, the frustration of the job in the slow-reforming countries was to witness, on a daily basis, the immobi1ity of patterns that could easily be changed. At the end of World War II, on the site of the Warsaw ghetto, destroyed in the fighting of 1943, the Poles (on Soviet instructions) built the Palace of Science and Culture, Warsaw's Stalinist Gothic masterpiece. The building is surrounded by open space, over part of which May Day marches were held. On this ground, a fascinating market sprang up. It started with people selling items purchased in Berlin supermarkets, largely foodstuffs that had been novelties or scarce in previous years. Bananas, a considerable rarity in the past, were sold one by one in this market. The car trunk trade was joined by people selling their surplus assets. Pensioners sold clothes and other odd items either draped over their arms or from a plastic sheet spread on the ground. Then people began to bring tables and chairs to sell from, and the fir st booths were erected. Other enterprising types set up hot kielbasa and mustard stands or sold bread products. Someone had the idea of importing brightly colored ready-made sales booths and renting them out. This gave the market a growing air of permanence. The next stage reflected the devel opments in countries further east. Initially, Poles would go east and purchase goods on whose resale they could tum a profit. These included tools and industrial parts, precision drawing equipment, and toys. These catered to the impoverished side of the Polish market that could afford to buy cheap Russian goods that had previously been scarce, but could not afford the relatively more expensive, higher quality western products. The whole market was a visible exercise in the power of arbitrage, as the differences in the relative price structures and relative scarcities prevailing in Germany, Poland, and Russia became equalized.2

2While Lhi.s was not visible to the local resident representative, the street markets in Warsaw had their counterparts in Lbe large number of Poles and other Central and East Europeans who appeared on the streets of Berlin, Vienna, and other cities, selling stuff they had brought from home. and using the proceeds to buy up goods to take back home.

©International Monetary Fund. Not for Redistribution THE TRANSITION IN CENTRAL AND EASTERN EUROPE

Some Poles quickly acquired sufficient capital by this trade to move up the retail market. The government quickly (by rnid-1990) ensured that the Warsaw municipality and state enterprises leased out the retail space they owned, and some of these traders, who had started by filling their car trunks in Berlin, were able to come in from the cold and set up their shops. The market then gradually acquired a more eastern flavor, as Russians moved in selling cheap clothing and tools. It became a daily occupation to wander around the market to see who was selling what. Perhaps the most remarkable sight we saw was a group of Mongolians selling brand-new Japanese pumps with the instructions in Mongolian. How the economics of transporting pumps from Ulan Bator to Warsaw worked out was hard to fathom. Then the authorities started to put up signs in Russian to tell people where to park and where not to spit, and the market became more and more settled. One area was cleared after a year or so, and large hangars erected where better security could be maintained and shoppers would be protected from the elements-such an anarchic untidy market in the center of Warsaw did not agree with many officials' view of what a modern European capital should look like. Regulations on street trading were tightened, and the market outside the Palace of Science and Culture was closed. In its place, a new market, the "Russian market" was opened in a la.rge stadium on the other side of the Vistula river. This allowed better policing of "undesirable elements," without stopping trade, but it lacked some of the vibrancy of the open market in the center of town. While the street market was being removed from the sight of those to whom it gave aesthetic offence, a remarkable transformation was visible in Warsaw's retail space. Warsaw had a reputation as a gray and drab city. (See O'Rourke, 1988). The courage and heroism of the Warsaw Ghetto uprising in 1943, the Warsaw uprising of 1944, and the vindictive and wanton destruction by the retreating Germans had left over three-fourths of Warsaw in rubble. Huge efforts in the immediate postwar years to house the population had created a city of generally uninspired concrete apartment blocks. Not only was the whole effect profoundly depressing, but also it was hard to imagine bow anything could be made of this legacy. That, however, was a failure of our imagination. When in command of their own destinies and out to make something of themselves, people come up sponta neously with remarkable solutions. Often, people who bad accumulated some capital and business connections from the early street trade rented retail space that the local authorities had made available or that state enterprises no longer needed . The dingy and cavernousspaces on the street level were opened up, given a coat of paint, lights were installed everywhere, and the windows filled with promotional material and goods. Advertising sprung up to cover the vast slabs of gray concrete on the sides of buildings. Car dealerships and innumerable groceries, clothing shops, and the like opened up, and the city was transfonned. Within three years, the drabness was gone, replaced not, perhaps, with beauty but with excitement and vibrancy, and this was followed by the construction of the first shopping malls. Poland and Hungary thus joined the group of countries where shopping could become a major leisure pastime. While the change was overwhelmingly favorable,

©International Monetary Fund. Not for Redistribution Mark Allen and Rick Haas

there were some causes for nostalgia. In particular, the number of bookshops, subsidized under the old reg ime and a source of cultural pride. could no longer exist. One indicator that we waited for with mixed feelings was the closure of the shops open in every socialist capital selling the books and cultural products of other socialist countries. While there had, perhaps, never been much demand for Russian books in Warsaw, it occasioned some regret to see the Soviet Bookstore closed, even though this was inevitable and logical given the prime piece of commercial real estate it occupied in Warsaw. The resident representatives also regretted the transformation of the Economist's Cafe next door and under the Polish Economic Association offices into a shop selling expensive perfumes.

Housing and Mortgage Market

One missed opportunity was the failure to develop a housing market. Even where, as in Poland and Hungary, the belief that an acceptable equilibrium would develop in most markets if appropriately freed, no one had the confidence to apply this to the market in that basic commodity, housing. The population, particularly the urban population. was largely housed in apartments owned by the state or by enterprises. Since rents were very low, the demand for apartments grea.tly exceeded the supply. and at the same time, complicated rules governed who was permitted to occupy what sort of an apartment. In these rules, possession played a very important role, and so young fa milies were stuck Jiving with their parents waiting for elderly relatives occupying other apartments to die. While the stock of apru1ments was privatized and the apartments sold to residents for relatively low prices, the very low level of incomes in the early transition period, the absence of a mortgage market, and the fact that people could continue Lo rent and buy their apartments only gradually, meant that for most people there was very little change in their perception of the housing situation. People, especially those dependent on incomes from the state or state industry, were unable to buy outright, and if they did, they would not immediately plan to move, so the supply of housing coming on to market was very small. Thus the price for an apa11ment looked horrifically high to the average apartment dweller, and this in turn made him less likely to lhink about moving. As a result, there was a highly suboptimal allocation of housing: many relatively large apartments were occupied by old couples who had originally been allocated the space when they had children living with them, while in other relatively small apartments three generations would be stuck together. To an economist who genuinely believed in the ability of markets to find acceptable equilibria, the solution would have been to get more apartments onto the market. This could have been done by waiving the purchase price and just giving apartments and full title to res idents. Then an old couple on a low income occupying a prime location in town might realize d1ey were sitting in a rather valu able asset. Someone earning a reasonable salru·y, but with wife ru1d children living cheek by jowl with his parents in two or three rooms, might be prepared to pay the older couple to move out. The latter would then have realized some of the capital they were sitting on. could move to more modest accommodation if they wru1ted. and have a pile of cash left over to be used to augment their income.

©International Monetary Fund. Not for Redistribution THE TRANSITION IN CENTRAL AND EASTERN EUROPE

The number of apartments in dilapidated buildings was also a problem, since the communal (utility) fees charged for maintenance and services had been kept at very low levels. Given the severity of the adjustment, it was hard to persuade a majority of tenants to spend more on renovation, but a renovation would have increased the value of the property severalfold. Indeed, cases started to appear where strong-ann tactics were used to encourage the older and poorer tenants to move so that the remainder, or someone entrepreneurially mmded, could get in and improve the property, with the rum of makjng a packet on the resale. Here the market fr ulure was a mixture of the fa Hure to give clear property title to residents, the unwillingness or inability of banks to develop a small loans business using the value of the property as collateral, and outright criminal behavior. Similar problems occurred in the individual housing sector. Many people both urban and rural had dreamed of building their own houses, either in the suburbs or in the country. Under communism, house-builcling had the nature of a savings account. The country was littered with half-finished homes. The problem then was not that cash was scarce, but that building materials were in short supply. As soon as someone could lay his hands on a paLlet of bricks or a load of cement, they would be snapped up and a bit more of the foundation or some of the next story would be built. The construction of houses in this fashion could last for years, �nd for much of the period the investment was unusable since the half-finished house was completely uninhabitable. This illusion of construction activity often caused first-time visitors from the West to believe they were witnessing a construction boom. As resident representative, this problem hit home as the stock of suitable houses for an economist and his family, used to living in d1e subw·bs of Washington, DC, was very limited. (While the reward of helping to make history might have been enough to compensate living very poorly for many well motivated resident representatives, it was not clear that spouses and children would get the samr psychic rewards.) The few suitable houses had usually been constructed and completed by someone who had had access. in the old days, to reasonable amounts of foreign exchange, often through working in the West. With the influx of well-heeled foreigners and the expansion of dle cliplomatic commu nity that followed the transition. the prices of such places were bid up to a level several multiples of the prices of similar accommodation in Wa shington. It was very frustrating to walk the streets armed with the IMF's administrative guideline on house rental and see so much unfinished construction. Lf only there had been a mortgage market, the owners of these places could fm isb them within weeks or months and start to realize a large rental income, sell them, or move in themselves from cramped accommodation elsewhere. The obstacle to this was not so much the high rates of interest prevailing in the early days of the transition, but the Jack of appropriate legislation on collateral. The authorities shied away from allowing eviction in the event of fa ilure to pay a mortgage. which meant that even if the bank could treat dle house as collateral. it could not effectively fo reclose on the mortgage. The inexperience of the banks and the inability of dleir advisors to devise schemes for mortgages under conclitions of high inflation also played a role.

©International Monetary Fund. Not for Redistribution Mark Allen and Rick Haas

In all these matters, a misperception of the economk consequences lay at the failure of policy. Legislators had a fear that simple freeing of the housing market and the passage of proper regulations on the control of collateral would result in a general rise in housing prices that would leave a good part of the electorate desti tute and unable to afford housing. Arguments that this would not happen-that since the immediate effect would just be to reallocate the existing stock, the free market would lead to a Pareto optimal reallocation of the existing stock-were not believed. Nor was it believed that, fairly soon, the supply of new housing would increase and the cost would fall. What was particularly fru strating in this situation was that the true liberalization of the market would have countered the sense of impoverishment felt by the majority of the population fo llowing the liberalization of prices in general. Had the housing market been liberalized quickly, there would have been a countervailing wealth effect from the realization that people possessed a valuable asset in the form of the real estate they were sitting in. It is not only in transition economies, however, that people have difficulty in appreciating the advantages of a free market in housing, judging by the frequent use of rent controls in the West. While the spread of street markets was the first visible sign of the penetration of market relations, capitalism requires far more than this. The transition frQm socialism to capitalism requires the accumulation of capital and large-scale production and distribution based on market principles and relations. While this further extension of market relations has been slow to come about in Belarus, and has been fraught with difficulties in several countries of the former Soviet Union, it happened relatively smoothly in Poland and Hungary. lt was possible to gauge the development from day to day by observations in the street. In the fir st days of the transition, an amusing occupation was to note the different reasons why shops, at which one needed to buy something, were closed at hours when they might have been expected to be open. Under the sellers' market prevailing before the transition, shopkeepers could open and close at their conve nience, knowing that customers needed them and would wait. Thus there was no question about opening for anything other than the standard working hours. Maybe a department store would be open in the evening one day a week and most shops were open from 8:00a.m. to noon on Saturday, but no longer. People were expected to take time off work to do their shopping, but even during these working hours it was normal to find shops closed. The commonest cardboard sign to tind in the window was "Break," another was "Vacation," and a third, reliably seen at least once a month, was "Taking Inventory." Another frequent sign (and the bane of those looking for gasoline) was ''Taking Deliveries." Then there were the more exotic, such as "No Electricity," or "Pipe Burst." At one point, we collected a list of about ten conditions that one might expect to find a shop in. As the transition progressed, and in particular as shops started to face compe tition from street traders in a buyers' market, these signs started to disappear, shops started opening for longer periods, and shopkeepers began to organize themselves in ways to ensure that they did not have to close in prime shopping time. About a year and a half into the transition in Poland. we realized how fundamental the change had been when the tiny comer store near our house was not only open on

©International Monetary Fund. Not for Redistribution THE TRANSITION IN CENTRAL AND EASTERN EUROPE

Sunday, but was taking deliveries from its suppliers and serving its customers at the same time.

End of the Black Market in Poland

The black market, especially in foreign exchange, was a feature of life under socialism. In any city, travelers wouJd be accosted by people offering to sell them local currency at black market exchange rates, more or less openly, depending on the degree of repression. In Poland, this was particularly the case, since the zloty bad long been among the most risible of socialist currencies. The first move in the transition process in Poland was to eliminate current account restrictions and move the exchange rate to a competitive level where il could be supported. This had been a central element in the fL rst stand-by arrange ment negotiated for Poland in late 1989, and in the fL rst months of 1990, there was clear evidence that the policy was working. On every street corner, small foreign exchange dealers (kantors) opened up for business, openly and legally buying and selling foreign currency for zlotys at very narrow rates. As a resident in Poland, the resident representative could always take a US$100 bill into any kantor and get the zlotys needed for daily activities, and it looked as though Poles could get.any amount of dollars too. So why didn't the suspicious looking characters offering an exceptionally favorable rate of exchange disappear? Economic logic suggested that when the resident representative could openly buy and sell a dollar for Zl 9,500, and so apparently could anyone else, there was no way that someone could offer Zl 15,000 for a dollar and make money on the deal. But, perhaps this was some form of Potemkin village, and there were really exchange restrictions that we were not being told about. Perhaps the freedom to buy and sell at the kantor was more circumscribed than we thought. Perhaps the restrictions on capital outflows were preventing Poles engaging in the capital flight that they desperately wanted to do, and somehow this demand was being satisfied through shady men who were prepared to pay a premium to get their hands on our dollars. Unlikely, but as representative of the institution that is supposed to know everything about countries' exchange restrictions, il was the resident representative's job to find out. So it was decided to use $20 of the resident representative's own money, not knowing how to account for it in the Resident Representative Office accounts, to explore this black market, even if it meant doing something illegal. The next day, as the resident representative walked past a big department store in the center of town, someone approached him and surreptitiously indicated that he would buy dollars at Zl 15,000. [ agreed, and he ushered me into a comer of the department store with a great show of making sure that we were unobserved. Since the exchange of currency was not illegal, as far as was known, the precautions seemed unnecessary. Under the cover of his raincoat, he then counted through a large wad of Zl l ,000 bills which he set aside. Then the $20 bill was exchanged for this wad of zlotys again making sure that no one was watching, and the "black marketeer" disappeared rapidly, as if afraid of being stopped by a newly materialized policeman. The resident representative looked at his wad of zlotys

©International Monetary Fund. Not for Redistribution Mark Allen and Rick Haas

and discovered that only the top one was a Zl 1,000 bill, and the rest were virtu ally valueless Zl 5 bills. He could thus happily report that he was the victim of a confidence trick, and that the liberalization of Poland's exchange system was as perfect as we had been led to believe.

V. What We Learned as Economists

The first is that incentives are important and markets really do work to solve the allocation problem. This is something we learned inte!Jectually from Samuelson's Principles of Economics, but the experience of life in the transition economies gave us daily, living proof it was true. It took considerable courage for the politicians in the more westernized transition economies to make the jump to trusting that spontaneous forms of supply and entrepreneurship would spring up to replace the old systems of administrative coordination. It may be that they had few alternatives, but they did take a huge responsibility when they committed themselves to "leap iJ1to a market economy." It is also understandable if other politicians fu rther east did not have the courage to make this leap. In the early phases of the transition, the responsible authorities faced the question, ''Precisely how will such or such needed good be delivered to the people who need it if we scrap the old system?" And the answer that, "Somehow, it would be worth someone's while to seiJ the good to those who needed it," was not always convincing ex ante.3 To people used to administering a system of production and supply, it was bard to give up the control. The second lesson was that there usually is an equilib1ium solution, no matter how unlikely it may seem at fL rsL While many western economists were extremely optimistic at the start of the transition that a set of relative prices would quickly be found to allow the productive assets of the former socialist countries to be put to work, and that growth would follow very soon. some of us were more pessimistic. For fifty years and more, these countries had been pouring resources into ill advised investment projects that had resulted in a proliferation of industrial behe moths, using obsolete technology to produce substandard products; in short, junk producing junk. Was there really a set of prices at which any of this could be viable? Or was the entire investment of the socialist period worth merely its scrap value, and w.ould the entire industrial plant of the nation need to be shut down, throwing urban millions out of work? It turned out, however, that a good part of this unpromising legacy could be used. The transition process imported a set of relative prices from the We st. The exchange rate played a crucial role in ensuring that some part of production could remain profitable, as did the low dollar wages at the start of the transition. In addi tion, crucial breathing space was given by implicit subsidies in the form of the erosion of state-owned capital and the credil from the barlking system that ulti mately went bad. But central to the process was the economics of desperation and the resulting entrepreneurship.

JThis gave rise to a new version of lhe old Polish joke. ·'In the new Poland. how many Poles does it take to change a light bulb?" '"None. The market does it."

©International Monetary Fund. Not for Redistribution THE TRANSITION IN CENTRAL AND EASTERN EUROPE

It took time. maybe a year or two in Poland, for those entmsted with managing enterprises to realize that they were running out of options. Fiscal subsidies were Umited, attempts to establish an industrial policy had been beaten back by the reformers in the Ministry of Finance, and the days when banks would provide credit to cover the service of old debts stopped. The tax authorities were insisting that taxes be paid, and the workers were resisting financing theirenterprises them selves through unpaid wages. The budget constraint had become hard.4 Management bad run out of the sort of options that required no effort on their part other than lobbying. These desperate circumstances forced a change in attitude and the resort to entrepreneurial behavior patterns largely unknown before. They started to look for the ways they could economize through small production line changes, inventory management, and small-scale investments. They piled into cars and scoured Germany for outlets for their products. They started to look for customers in Poland itself, and to control their accounts receivable. In this way, the inherited means of production were salvaged and a new and acceptable equiHb rium reached. ln this connection, another Jesson was that the switch from the sellers' market of socialism to the buyers' market of capitalism entails profound psychological change. All psychological change is very hard to bring about, for people will resist changes in their way of thinking. The experience of Poland and Hungary showed that the psychological change can be made if the government is credible in its hardening of the budget constraint. There will be considerable resistance, however, to the change and very widespread calls for loosening the constraints. If these siren songs are resisted, which can be a challenge for an emerging democracy, the prospects for renewed growth are good. If it is impossible to resist them or if the government is not serious about tightening the constraint, then the recovery is likely to be much slower, as was the case further east. Another lesson concerned structural reform. From the outset, it was recog nized that structural reform was central to the transformation process. but it was initially thought that such reform would largely happen spontaneously once stabi Hzation had been achieved. Subsequently a fa irly extensive structural reform agenda was set, but in the former Soviet Union it ran into vested interests that were more deeply entrenched than in Central Europe. Structural issues are important and figure prominently in the process; macroeconomic stabilization by itself is not sufficient to ensure transformation. A major lesson for economists that have had the experience of living with tran sition is that attention must be paid to establishing the institutional underpinnings of the market. The task of creating market economies has reminded us that markets are embedded in a set of institutions and behavior patterns that economists have normally taken for granted. A legal system compatible with market behavior is crit ical. Bankruptcy and collateral laws are needed, as is a dispute resolution mecha nism to resolve contractual differences. Economists in general have paid no more attention to the institutional structure in which markets work than fish pay to the water they swim in. But once the first step of the transition was completed-that of

4See Komai, 1980.

©International Monetary Fund. Not for Redistribution Mark Allen and Rick Haas

scrapping controls, opening markets and liberalizing prices-the problems that immediately surfaced were those of inadequate or nonexisting institutions. In particular, a strong, independent central bank is important. ln those countries where the transition was more successful, the inherited institutional structure or people's memory of pre-socialist structures were in better shape than in those countries further east. Thus the tendency for the latter to reach back for controls rather than introducing the needed institutional change is understandable. We also learned that transformation has noneconomic consequences. In partic ular, it leads to the col1apse of traditional values. The old systems in Central and Eastern Europe and in the Soviet Union were based on political repression and brutality, but nevertheless, certain values were promoted. The resistance of parts of the elite to change ret1ected their adherence to a set of admirable values that the transition process undermined. The first of these values was a sense of social justice. No matter how poorly the system actually delivered social justice, it was a broadly accepted goal that society should care for all, and that universal education and health care should be provided. The system did deliver on job protection, and it did provide a minimal level of universal health care and education. Those who thought about these things-and that included many of the officials with whom we carne into contact with on a daily basis-hoped for a new system that would deliver these social benefits, at least at the level the old system aspired to do. The undelivered promises for social welfare were among the strongest causes of the loss of legiti macy of the old system. The transition, however-for a variety of reasons, including the need for transparent and disciplined fiscal processes-initially made it more difficult to deliver on these social goals. Indeed, the papers were full of stories of the worsening conditions of pensioners, hospitals unable to finance the care they wished to give, and underfunded schools. Related to this sense of social justice was a sense that society should reward people according to their merit. The hierarchy of pay scales under socialism, the availability of rewards, and the associated prestige given to different jobs corre sponded grosso modo to society's conception of the appropriate way of rewarding people. Of course, there were exceptions, in particular the privileges appropriated by the aparatchiks and those gained through overt corruption, and these excep tions again fueled the opposition to communism, but otherwise the social order seemed intact. With the transition, however, the wrong people, in the public's view, seemed to be getting rich. The skills of the black marketeer, associated with criminality in the past, became the way to thrive in the new economy. Operators using sharp practices began to become very rich as they exploited the opportunities for arbi trage. With the new legal framework unclear, these individuals were often none too scrupulous about bow they enforced contracts, and, along with some from the nomenklatura, smartly appropriated state property. The border between legitimate economic activity and criminal behavior was very unclear at the start of the tran sition, and remains unclear in some parts of the former Soviet Union. It is not hard in these circumstances to see why decent, hard-working people felt that the new system was undermining their values.

©International Monetary Fund. Not for Redistribution THE TRANSITION IN CENTRAL AND EASTERN EUROPE

Finally, the transition brought the immediate triumph of western matelial values and undermined the position of the intelligentsia in these countries. While the prevai ling ideology of the socialist system was an illiberal Marxist-Leninist one, it claimed to be descended from the finest traditions of the enlightenment. Similarly. the national cultural u·aditions were fostered to give added legitimacy to the socialist regime. Thus, for whatever reason, these countries devoted consider able resources to the promotion of the best Western cultural traditions. At least since the end of d1e Stalinist repressions, the intelligentsia, a term that encom passes the liberal professions, journalists, artists, etc., was able to live in a humanist tradition, while paying no more than lip-service to the official ideology. Many of these people were happy to see communism collapse, but then found that the level of material support available for cultural life had collapsed along with it. Books were published not when the publishers, editors, and authors considered them important when viewed through their own cultural prism, but when there was a large enough market. So the bookstal1s were flooded with lurid copies of transla tions of Western best sellers of doubtful literary merit. To d1ose for whom high culture was a central and meaningful element in their lives, the new system brought cultural vandalism. Such people, influential in the media, and often in the bureau cracy, came quickly to see the systemic change as a very mixed blessing.

A Final Thought

In writing this essay, it became clear that experiences of the former Soviet Union countries differed in some important respects from those of the transition economies in Central Europe, which have been more successful in the transition process. While most of the negative phenomena that occurred in the former Soviet Union were present, at least in ernbryonjc form, in Central and Eastern Europe, the elites in the latter countries had a better idea of the sort of economy they wanted to create. so there was less work to do to explain how a capitalist economy worked and more of a commitment to reform. This also helps explain, in part, why the IMF's advice worked better in Central Europe than further east. The reason for a good prut of the differing experience, however, lies in the political process. It proved relatively easy in Poland and Hungary to establish a normally functioning democracy in which checks and balances among different branches of government could function properly in a civil society. In Russia, Belarus, and the neighboring republics, the democratic political framework proved much more difficult to establish. Thus, we would be remiss if we did not highlight the importance of the political system as one of the necessary preconditions for a properly functioning market economy.

REFERENCES

Bruno. Michael, 1992, "Stabilization and Reform in Eastern Europe: A Preliminary Evaluation." Staff Papers, International Monetary Fund, Vol. 39 (December), pp. 741-77.

Campbell, Robert Wellington, 1992, The Fa ilure of Soviet Economic Planning: Sysrem. Performance, Reform (Bloomington, Indiana: Indiana University Press).

©International Monetary Fund. Not for Redistribution Mark Allen and Rick Haas

Fischer, Stanley, Ratna Sahay, and Carlos A. Vegh, 1996, "Stabilization and Growth in Transition Economies: The Early Experience," Journal of Economic Perspectives, Vol. I 0. No. 2, pp. 45-66.

Komai, Janos, 1980, Economics of Shortages (Amsterdam and New York: North-Holland).

O'Rourke, P.J. 1998, Holidays in Hell (New York: Atlantic Monthly Press).

Sachs. Jeffrey, 1993, Poland's Jump to the Market Economy (Cambridge. Massachusetts: MIT Press).

Ten Years After ...Transition and Economics

GERARD ROLAND*

This paper attempts to portray a synthesis of what has been Learned in the past 10 years with regard to the transition process. It contrasts the mainstream "Washington consensus" view of transition with the "evolutionm:v institution alist" perspective. It argues that the Latter gives a more adequate and complete picture both of the transition processes and of economic systems and is of better help to prevent serious transition fa ilures. [JEL PJO, PI6, P4 1, P5 1, P52]

ore than 10 years after the fall of the Berlin Wall and the beginning of tran sition in former socialist countries, it is useful to try to gather some of the Mlessons that have been learned with respect to our vision of the transition process and also with respect to how transition processes have affected our way of thinking about capitalism as an economic system. Transition issues have appeared initially quite controversial. There have been controversies on the speed of reforms, priva tization methods, the role and organization of government, the kind of fi nancial system needed, etc. While these controversies often have been seen as ideological, they also reflect to a large extent the initial ignorance and unpreparedness of the economics profession with respect to the large-scale institutional changes implied by the transition from socialism to capitalism.

When the Berlin Wall fe ll unexpectedly, there was indeed no preexisting theory of transition. We had no preexisting theory on the effects of political constraints on transition strategies, the effects of liberalization in socialist economies with no preexisting markets, how to privatize socialist enterprises given

*Gerard Roland is Professor of Economics at the Unjversity of Califomia at Berkeley. at ULB and CERGE-EI. He is director of the CEPR research program on transition economics and a fellow of the William Davidson Institute. The author thanks Ratna Sahay and an anonymous referee for helpful comments.

©International Monetary Fund. Not for Redistribution Gerard Roland

the legacies left by socialism, and how to harden budget constraints and achieve efficient restructuring. Not surprisingly, transition brought many unexpected surprises:

• The huge output fall after price liberalization, which was not predicted by the economics profession at large nor by scholars of economic systems.

• The continuous economic decline in Russia and other countries of the former Soviet Union and the divergence in economic performance between Central Emope and most countries of the fa nner Soviet Union. • The extent of insider privatization. • The observation of restructuring in many state-owned enterprises despite initial fears of widespread asset-stripping before pJivatization of the enterprise.

• The extent of the development of the Mafia phenomenon. Where we expected the emergence of markets. often the Mafia emerged instead. • The breakup of countries (Yugoslavia, Czechoslovakia, USSR, and seces sionist tendencies inside Russia). • Last and not least the huge economic success of Chinese reforms, which led to sustained growth and prosperity over a period of 20 years. The economics profession was thus very ignorant when transition started. What have we learned in the past ten years? I should start by stating that we remain relatively ignorant and that there remain important disagreements among economists doing research on transition. Ten years can be a long time in the life of people, but in research, it is a short time span. Compared to other economic events and issues, transition is also especially complex. Understanding transition requires a comprehensive understanding of the constitutive elements of capitalism and of their interactions. Moreover. it requires also an understanding of the dynamics of large-scale institutional change, an even more difficull obj ective. Nevertheless, important progress has been made by the economic research on transition. As a result, there also has been an impo1tant convergence on many issues. Not only have visions of transition evolved substantially in the past ten years, but research on transition also has contributed to certain shifts of emphasis in the way economists look at economic systems. The various surprises of transi tion have further contributedto a change of focus that has been taking place in the past two decades in the vision of economics. They have very much reinforced the institutionalist perspective, emphasizing the importance of the various institutions underpinning a successful capitalist economy. Successful institutions of capitalism are already present in advanced economies, and we tend to take them for granted when reasoning on economies in transition or on developing economies where such institutions are absent. If anything, the experience of transition shows that policies of liberalization, stabilization, and privatization that are not grounded in adequate institutions may not deliver successful outcomes. Much of this change of focus toward the institutionalist perspective already had been taking place with the development of contract the01y, political economy, law and economics, regulation theory, corporate finance, and other areas in applied economic theory. The experi ence from transition, however, has contributed to accelerate various changes of focus in the way we think about economics. Thus, there is a shift of emphasis from

©International Monetary Fund. Not for Redistribution TEN YEARS AFTER ...TRANSITION AND ECONOMICS

markets and price theory to contracting and to the legal, sociaL and political envi ronment of contracting. Transition has not only helped to reinforce this change of focus in economic thinking, it has also renewed interest in thinking about the inter play and complementarities between the various constitutive institutions of capi talism. Finally, transition has forced us to think about institutions not in a static way but in a dynamic way: how momentum for reform is created. how institutions can evolve but also how momentum can be lost, and how one can get stuck in inef fi cient institutions. 1n this sense, transition has reinforced what I would call the evolutionary-institutionalist perspective, insisting both on the institutional envi ronment of agents at any moment in time but also on its evolution.

I. The Washington Consensus Ve rsus the Evolutionary-Institutionalist Perspective

To organize the discussion, it is useful to present a simplified and schematic presentation of the differences between the two main visions of transition that have crystallized since the beginning of the transition process and that shaped both policy recommendations and research programs. The first vision of transition is generally referred to as the "Washington Consensus."' I will call the second one the "evolutionary-institutionalist perspec tive:' One could label the former the big bang or the shock therapy view and the latter the gradualist or incrementalist view. I prefer not to do so because these labels are too narrow in their focus since they mainly emphasize the speed of reforms, whereas, as we will see below, there are many more dimensions than speed that are involved in both visions. This broadening of the focus since the beginning of tran sition is the object of a consensus among researchers working on transition on both sides. The presentation that follows is certainly schematic and incomplete. Most researchers would not fully recogn ize their own vision of tran sition in either category, and there also has been an important degree of conver gence over time in many of the dimensions that I will evoke. Even though the "Washington Consensus" is usually associated with the views of the IMF and World Bank, the description below should in no way be seen as a description of their official views. Views within the IMF and the World Bank, moreover, have undergone change over time and should not be seen as monolithic. The purpose of this presentation is thus not to draw pictures of two camps or to draw a "straw man" but to crystallize ways of thinking about transition in a comprehensive way. In all the dimensions of transition treated below, debates have been taking place in the past ten years reflecting these two visions in one way or another. These debates, and more importantly the transition experience itself, have Jed to a process of learning. As of today, many fervent advocates of the Wash ington

IThe term "Washington Consensus" was initially coined by John Williamson in 1990 and did not refer at all to transition. Since then. as observed by Williamson (2000) himself. the tenn has taken a life of its own. It has been used in panicular to label the thinking behind lMF and World Bank onhodoxy and has aroused controversy not only in the context of transition but also in the context of structural adjustment programs. l will say nothing about the latter and will focus only on transition aspects.

©International Monetary Fund. Not for Redistribution Gerard Roland

consensus would only partially recognize their views in what follows because of the results of this process of learning. Also, the presentation of the evolutionary institutionalist perspective may be seen as biased because the vision presented here already reflects the learning process that has taken place. The vision of those who, ten years ago, were already critical of the Washington Consensus, was certainly not as systematic as it is today. Presenting the Washington Consensus mainly as it was expressed ten years ago and the evolutionary-institutionalist perspective as it exists now may seem biased if one thinks in terms of analyzing a battle between camps, but it is very useful if the main objective is to understand what we have learned from the transition experience itself. The Washington consensus view on transition was clearly dominant in the beginning of the transition process in Central and Eastern Europe. It has shaped policy recommendations and dominated thinking in international financial organi zations and has been supported and endorsed by famous economists from the best universities in the world. It has influenced to a large degree economic policy in most transition economies with the very large exception of China, which has fo llowed its own transition path in a very pragmatic way.2 Intellectually, it is rooted in a combination of (l) standard neo-classical price theory; (2) standard macroeconomics and experience of stabilization policy: and (3) a broad body of knowledge in comparative economic systems emphasizing both the complemen tarity of the constitutive institutions of economic systems and the disappointing experiences with partial reform in Central and Eastern Europe prior to the fall of communism. The evolutionary-institutionalist perspective has had more support in academic circles than in international policy circles. It was clearly a minority view in the beginning of transition but has gained more and more support over time in light of the transition experience. Intellectually, it is rooted in (1) the institution alist perspective given by modern microeconomic theory and shaped by the devel opment of non cooperative game theory; (2) the evolutionary approach to economics (see, for example, Murrell, 1992); and (3) a philosophical skepticism, influenced by Hayek and Popper, with a strong emphasis on our relative ignorance of economk and social systems and their transformation, and an emphasis on the uncertainty associated with societal engineering and a strong aversion to any kind of Bolshevist-style campaigning in large-scale institutional transformation. Ta ble 1 gives a sununary and simplified presentation of these two opposed visions (see also typologies in Murrell, 1992; and Stiglitz, 2000). Going from the "forest" to the "tree," we look successively at the attitude toward (I) the political economy of reform and reform strategies; (2) allocative changes; and (3) gover nance changes. The former relates to the sectoral reallocations following liberaliza tion and their macroeconomic consequences, and the latter relates to changes at the micro level and in particular at the level of enterprises. Some of the differences listed in the table are less important and less acute, while others are more important.

2Qbviously, it has not influenced those govemmenrs that were mainly opposed to reform and whose strategy was to block rransition as much as possible. ln contrast, Chinese reformers have sought to advance the reform process and to bypass conservative opposition from within the Communist party. They have been successful in doing so and have, in the process, developed an original transition path.

©International Monetary Fund. Not for Redistribution TEN YEARS AFTER ...TRANSITION AND ECONOMICS

A fundamental difference in both visions is the attitude toward uncertainty with respect to the outcome of reforms. The Washington consensus emphasizes that reform will deliver sure efficiency gains. It contains a strong faith in societal engineering. The idea behind this is that the economics of reform is well under stood. Since the initial situation is characterized by fundamental inefficiencies and since economic theory predicts that transition will deliver sure efficiency gains, then these reforms should be implemented with the faith that the efficiency gains will be reaped. We know that capitalism, as experienced in the United States or Europe, has proved more successful, so it is simply a matter of copying the better models. In contrast, the evolutionary-institutionalist perspective emphasizes the aggregate uncertainty of transition outcomes. Aggregate uncettainty means that aggregate, economy-wide reform outcomes may range from very positive to very negative. In other words, economy-wide success is by no means guaranteed. We do not know in advance whether the outcome will be closer to the West German miracle or the We imar republic, not to speak of the former Yugoslavia. To be sure, the point is not whether we know what the successful models of capitalism are, but what is the combination of ingredients that has delivered the observed successes and how difficult it is to export those ingredients. Even in trying to copy the better models of capitalism, things may thus go wrong. Our understanding of these large-scale changes is still rudimentary and nothing guar antees that there will not be huge unexpected and undesired outcomes. There are a huge number of coordination problems, moreover, to be solved among economic agents. Among the huge multiplicity of equilibria implied by these coordination problems, we do not know in advance which one wm be selected and why. For these different reasons, there is likely to be important aggregate Lmcertainty over the outcomes of reform. This uncertainty will be reflected in both popular attitudes towards reform and in decision making by policymakers. This important difference in starting points has implications OD reform stJ·ate gies. For the Washington Consensus view, the political economy emphasis is to use early windows of opp01tunity or periods of "exceptional politics" to push reforms through as fast as possible and to create irreversibility. lD the evolutionary-institutionalist perspective, the latter strategy may be dangerous and lock whole countJies in situations of inefficient economic outcomes that are hard to reverse because of the ineversibilities created. (Such outcomes may break social cohesion and generate important political instability.) The emphasis is rather on ensuring a continuous and growing supp01t for reforms among the popu lation. This implies a more gradual and experimental approach to reforms, relying on the flexibility of experimentatioD, with an adequate sequencing of reforms, to possibly reverse reforms that do not work and try other ones. The Washington Consensus tends to reject iD general any partial reform. The idea is that any partial reform will create rents for given groups that will be threat ened by further reforms. Partial reforms, therefore, create constituencies that will tend to oppose further reform, whereas this will not be the case with a compre hensive iDtroductioD of refmms. The evolutionary-institutionalist view is less pessimistic about partial reform-all depends on the sequenciDg of reforms. While some partial reforms may indeed stall the reform process and even lead to unnec-

©International Monetary Fund. Not for Redistribution Gerard Roland

Table 1. A Simplified Presentation of the Two DifferentVisions of Transition

Washington Consensus Evolutionary-lnstirulionalist View Perspective

J. Political economy of reforms andreform strategies

Attitude toward uncenainly Insistence on sure efficiency gains; Insistence on aggregate fait h in societal engineering unce!'lainty: skepticism toward societal engineering

Political economy empha:.i> Use window of oppoi'IUnity to Ensure continuou• and create irreversibility growing suppol'l for reforms

View of partial refonns Create rents that block fUI'Iher Depends on sequencing: can refonn progress either create momentum or �tall reform process

View of refolD! complementanties Ofabsolute importance. Very important but Necess ity to jump-stan the comprehen.�iveness of initial market economy by simultaneous refoons not necessary, introduction of all main refonns provided initial reforms can create momemum for furtherreforms. Transitional institutions can develop and evolve gradually toward more perfect institutions

Main support group for refoiDls Owners of privatized enterprises Middle class and newprivate sectOr

Focus of refonns Libemlization, stabilization. Create in�titutional privatization underpinnings of markets to encoumge strong entrepreneurial entry

Attitude toward in stitutional change Emphasis on adoption of laws Comprehensive: legal and financial change. law enforcement, refonn of o.rganization Of govcmmenL development tlf self-enforcing social norms

Auirude toward initial condition.� Create "clean slate" conditions Usc existing institutions to by bre;�king existing Communist prevem economic di sruption state structure and social unrest while developing new institutions

2. AJiocative changes

Main view of markets and Markets will develop Impo11anceof insntutional liberalization spontaneously provided underpinnings needed to government does not intervene: enhance market growth: supply and demand as focuJ. uf miuimum legal and analysis contracting eovirorunent, law enforcement, political stabilily. buililing of business networks and long tennparmcn.b ips; contracting agen� and instilUlionaltheir environment asunit of analysis

Mainattitude toward ineflicient Aggressive closing down Containment and politically state·owned enterprises feasible downsLdog. Rely on evolutionary development of private sector to shrink state sector

Main view of government Weaken it as much as possible to Role of government in law prevc:nt intervention in market& enforcement and insecuring propeny rigbl!>

©International Monetary Fund. Not for Redistribution TEN YEARS AFTER ...TRANSITION AND ECONOMICS

Table 1. (concluded)

3. Governance cbanges

Focus of privatization Fast transfer of ownership to Emphasis on organic private hands via mas.� development of private sector. privatization to break government Emphasis on sales to outsiders power and jump-start market to achieve efficient tran�fer of economy. Faith in market to ownership from the stan ensure efticicnt resale

M11in cmphll:li� of govcmmr;n\ Mn'm cmphn�··� b �lu'm'!Jng the RcfUIIll in the UtgunitaiiUII ur reform si1.eof government government so as to align as much as possible the interests of govemmcn1 bureaucrats with the development of market�

Hardening budget constraintS Exogenous policy choice that Endogenous outcome of depends on political will in stltutional changes

essary reversal, starting with other partial reforms may create momentum for further reform. This will be the case especially when reform complementarities are important. Models of reform sequencing by Dewatripont and Roland (1995, 1997) show that complementarities can be exploited by reformers when choosing the order of reforms. Indeed, by starting with more popular reforms, it is possible to build constituencies and political support for further, initially less popular reforms. The idea is that positive uncertainty resolution enhances voters· perception of positive reform outcomes. In that case, when reforms are complementary, the only choices are to go forward or go backward. Voters will thus be more prepared to move forward in order to keep the gains from initial reforms after partial uncertainty resolution about those reforms. Reforms should thus be ordered in a way to start with reforms more likely to deliver good news for voters. Popular reforms should precede less popular reforms. On the contrary, starting with reforms more likely to hurt a majority of the population risks a backlash, since bad news on initial reforms will build support for reform reversal. For the Washington Consensus, complementarities in reform are of absolute importance and are an overriding argument for a big bang approach whereby all reforms are introduced in a simultaneous and comprehensive way.3 For the evolutionary-institutionalist perspective, complementarities, while clearly impor tant, are not an overriding argument as long as sequencing of reforms can be used to create momentum for further reforms. This difference between both visions also has an important implication with respect to institutional reform and the introduc tion of new institutions.4 The Washington Consensus insists on fast introduction of "best practice" institutions. This is not the case with the evolutionary-

)N ote that if the latter view is taken seriously, then partial reform cannot stall the retorm process over lime since complementarities make partial reform inherently unstable leaving as the only choice to move forward or backward in the reform process. It is thus i nconsistent to insist both on the importance of complementarities and of the danger of partial reforms stalling the transition process. 4We will come back in more detail below to the differences with respect to institutional reforms.

©International Monetary Fund. Not for Redistribution Gerard Roland institutionalist perspective. Introduction of "best practice" institutions may not be possible for political and social reasons, and may not be necessary. The evolu tionary perspective implies that transitional institutions can develop that are adequate to the initial conditions but that can gradually evolve toward more perfect institutions. Here also, flexibility is important to prevent a lock in in inefficient institutions that are hard to change. A more minor point, but nevertheless with possible wide-ranging practical consequences, is a difference in emphasis on the main support groups for reform. The Washington Consensus emphasizes mostly the support of owners of privatized enterprises. The idea is that fast mass privatization creates constituencies among insiders and among those who benefit most from mass privatization to block any reversal. That is one of the reasons for pushing for rapid privatization. Since the owners of privatized enterprises are a minority and are not likely to be median voters in elections, the focus is mainly on the creation of powerful lobbies for capi talism via mass privatization. The emphasis is clearly on interest group politics as opposed to electoral politics (on this distinction, see Persson and Tabellini, 2000). The evolutionary-institutionalist perspective pays more attention to the attitudes of voters and to electoral politics. The purpose is more to rely on a broad group of small entrepreneurs and the middle class that emerges from the entry of new enter prises both in urban areas and in the countryside. The idea is that the middle class always plays an important role in democracies since voters of that group are more likely to be pivotal in elections. This also explains a greater care for social cohe sion and a fear of excessive inequalities, which are likely to create more pressures for redistribution and may generate political instability. Let us now turn to issues related to the substance of reforms and their focus. Here also, there are nonnegligible differences in vision. First of all, the Washington consensus view on transition is based on liberalization, stabilization, and privatization. At the start of this new millennium, few serious economists dispute the need to liberalize, stabilize, and privatize. There is not much disagree ment either about the advantages of using shock therapy as a stabilization method when it is politically feasible. For the evolutionary-institutionalist perspective, the emphasis is more on the creation of the adequate institutional underpinnings of markets to encourage a vigorous process of entry, competition, and exit. The idea is that liberalization, privatization, and even stabilization will not necessarily deliver the desired outcomes in the absence of such institutional underpinnings and may lead to unpleasant surprises. To avoid any misunderstandings, it would be completely wrong to state that the Washington consensus has ignored institutional reform while the evolutionary institutionalist perspective has not. While the latter insists more on institutions than the former, the main differences are elsewhere. The Washington consensus emphasizes mainly the introduction of laws: adequate laws to secure private prop erty, rights of shareholders, creditors, and so forth. The evolutionary-institution alist perspective takes a more comprehensive view toward institutional conditions. These include not only legal and financial change but also comprehensive condi tions of law enforcement, including reform of the organization of government and the development of self-enforcing social norms that foster entrepreneurship, trust,

©International Monetary Fund. Not for Redistribution TEN YEARS AFTER ...TRANSITION AND ECONOMICS

respect of legality, and commitment. The evolutionary-institutionalist perspective has sometimes been derided by statements according to which it recommends that perfect institutions should be introduced first before implementing any liberaliza tion or privatization. As stated above, this is not the case. The evolutionary attitude toward institutions implies that minimum institutions underpinning market devel opment must be present from the beginning because these define the rules of the game and thus place restrictions on undesired kinds of individual behavior and because they reduce uncertainty. Adequate institutions must develop via trial and error and must evolve over time into more perfect institutions that cannot be intro duced overnight. Related to the latter question is the attitude toward the initial conditions of reform. The Washington Consensus emphasizes the need to wipe the slate clean by breaking as fast and as thoroughly as possible the existing Communist state structure. The logic is one of "cavalry attack" to break any possible resistance and sabotage to reforms by the conservative communists of the former nomenklatura. The rhetoric uses revolutionary metaphors, and comparisons are often made between post-Communist transition and the French revolution (see, for example, Sachs, Woo, and Yang, 1999). This emphasis is very strong. Early in transition, most experts on socialist economies were pushed aside and declared obsolete by the shock therapists. Many of these new transition experts stated repeatedly that knowledge of the former system is a liability and ignorance an asset in under standing transition. Tbis "clean slate" view has found its most accomplished implementation in East Germany where the old system was thoroughly broken up at grand speed while West German experts came to build their own institutions on the ashes of the old system. In contrast, the evolutionary-institutionalist perspec tive sbies away from a revolutionary approach to transition. It emphasizes the need to use the existing institutions to prevent economic disruption and social unrest while developing new institutions. We have here again an area where both sides put a strong emphasis on insti tutions, but the Washington Consensus insists on an uncompromising approach and the absolute need to thoroughly destroy past institutions and directly put in place the best imaginable institutions. This difference is related to the distinct views about institutions. The main emphasis of the Washington Consensus on the introduction of laws is coherent with a "clean slate" approach and a will to intro duce best practice institutions independently of initial conditions. On the other hand, the more comprehensive approach of the evolutionary-institutionalist perspective explains the greater skepticism toward import of institutions from outside and the greater insistence on trying to guide an evolutionary and flexible change of institutions in accordance with the initial conditions. On allocative changes-that is, the sectorial reallocations related to price Liberalization and their macroeconomic consequences-both views differ on their main vision of markets and liberalization. The Washington Consensus emphasizes that markets will develop spontaneously, provided there is price flexibility and the government does not intervene in markets. Supply and demand are the main focus of analysis, and the main implicit theoretical tools guiding this vision are price theory and general equilibrium theory. The evolutionary-institutionalist perspec-

©International Monetary Fund. Not for Redistribution Gerard Roland

tive relies much less on standard market analysis and emphasizes the institutional underpinnings of markets and the effect they may have on the speed of growth of markets and entrepreneurial activity. Contracts and the relations between contracting agents are the focus of analysis. Therefore, a strong emphasis is put on the general environment of contracting: the minimum legal environment, security of property rights and law enforcement, political stability, the development of business and market networks facilitating search, the development of specific investments in long term business relationships, and so forth. A less important difference, though not without consequence, is related to atti tudes toward inefficient state-owned enterprises. The Washington Consensus developed, early in transition, a hostile view toward state-owned enterprises, emphasizing the need for fast privatization to prevent asset stripping and the need to quickly close down loss-making plants and firms. Again, it was in East Germany that this approach was implemented most thoroughly. The evolutionary institutionalist perspective is less aggressive and takes an attitude of containment, gradual downsizing, and of hardening of budget constraints, taking into account political constraints. The emphasis is more on developing a strong new private sector to attract workers away from the state sector and to let the latter shrink grad ually over time. The main view of government in transition developed by the Washington consensus view is that of the necessity of weakening government as much as possible to "de-politicize" the economy and to prevent intervention in markets. The evolutionary-instin1tionalist perspective emphasizes the importance of government in enforcing the law and the security of property rights. ln particular, adequate government infrastructure (police, courts) is needed to ensure that the rules of the market game are followed. Among others, it is important to fight orga nized crime and racketeering. It is also important to enforce an adequate competi tion policy to prevent monopolization. Turning now to governance changes, the two visions differ in their focus of privatization policy. The Washington Consensus emphasizes the need for a fast transfer of ownership to private hands via mass privatization to break government power and to jumpstart the market economy. Speed is of the essence. The idea is that any privatization is always better than maintaining government ownership so that the benefits of fast privatization outweigh the costs in terms of possible misal location of assets to private individuals and groups. There is also a strong emphasis on developing stock markets so that efficient resale of assets can take place after privatization. In contrast, the evolutionary-institutionalist perspective puts , in general, less emphasis on the importance of fast privatization of large state-owned enterprises. There is a broader view of privatization not onJy of existing state-owned enterprises but of privatization of the economy with, as main element, the organic devetopment of the private sector. In terms of pr\vatization of large enterprises, the emphasis is on competitive sales to outsiders to ensure effi cient transfer of ownership from the start. There is a great skepticism with respect to the possibility of efficient resales given the necessarily rudimentary develop ment of financial institutions and markets at the beginning of transition. The latter must necessarily evolve over time and cannot be jump-started.

©International Monetary Fund. Not for Redistribution TEN YEARS AFTER ...TRANSITION AND ECONOMICS

With respect to reform of the organization of government, the Washington consensus view attaches less importance to this question and emphasizes mrunly the need to shrink the size of government. The evolutionary-institutionalist perspective goes further than the simple dimension of government size and emphasizes changes in incentives of government bureaucrats. Since government agencies and officials can predate markets and the private sector, and since they can be captured by interest groups such as monopolies and Mafia, it is important to implement reforms in the organization of government so as to align, as much as possible, the interests of government bureaucrats with the development of markets. The idea is that markets and the private sector cannot develop in an envi ronment of government hostility. Adequate refonn of government administration. therefore, is needed to create more congruent interests between the private sector and government bureaucrats. In terms of hardening budget constraints, a dimension that has appeared more and more important with the transition experience, there are also conceptual differences. For the Washington consensus view, hardening budget constraints is mainly an exogenous policy choice that depends on the political will of policy makers. For the evolutionary-institutionalist perspective, soft budget constrrunts are related to a commitment problem. Because of thls commitment problem, exhortations to harden budget constrrunts may not be credible. Hardening budget constraints must be an endogenous outcome of institutional changes designed to create credibility for hardening. While we have emphasized the differences between the two visions-the Washington consensus view and the evolutionary-institutionalist perspective-it is important to note that these are not visions that are diametrically opposed. Both aim at introducing a successful market economy based on private ownership. Nevettbeless, these differences in vision reflect differences in the approach to economic analysis and lead to differences in policy emphasis in several important dimensions.

II. Broad Lessons from the Transition Experience So Far

Ta king a broad view of the transition experienc..:e, it would seem at first view that the following general assessment can be made. Central European countries that started transition early, that are growing agrun, and that now face the prospect of entry in the Ew·opean Union, can be broadly seen as a success of the Washington consensus, whereas Russia, with its dismal economic performances throughout the 1990s, can be basically seen as a failure of that view. In contrast, the success of Chinese transition cannot be attributed to the Washington consensus view but can be seen as a confirmation of the evolutionary-institutionalist perspective. Such a broad assessment is not the object of consensus in the profession but would seem to be an a priori no-nonsense characterization. Clearly, however, we cannot content ourselves with such broad generalizations. A more detruled and precise approach is required. We must look at the detruls of the various dimensions of transition and see in which areas research has led to consensus views, in which

©International Monetary Fund. Not for Redistribution Gerard Roland

areas controversies remain important, and in which areas research has been rela tively neglected.

Reform Strategies and Political Constraints.

In terms of the attitude toward uncertainty. I would claim that the facts have vindicated the evolutionary-institutionalist perspective and proved the Washington consensus view to be wrong. Indeed, refom1s have not uniformly delivered good outcomes or economy-wide efficiency gains. A simple look at the differences in performance between the transition countries shows wide variation in the evolu tion of output, between the strong growth of the Chinese economy, the U-shaped output evolution in Central Europe, and the continuous decline in many former Commonwealth of Independent States (CIS) countries (see Figure I).

180

160 China 1983 = 100

140

1.20 >< <> "0 ..5 100 c.. 0 0 80

••• •••• ••• •• • • Russia 1991 = 100 . .. .. 60 ...... 40

0 2 3 4 5 6 7 8 9

Nwnber of Years After Liberalization

One can always come up with ex post explanations for the observed evolutions in the various countries and trace them to wrong or incomplete policies, lack of comprehensiveness of reforms, and so forth. In practice, there have indeed been many such cases. It is not acceptable, however, to invoke this argument to dismiss these empirical observations. From the ex post point of view, it is nearly a tautology to explain failures in petformance by wrong or incomplete policies. Policies are endogenous, however, and depend on political constraints. It is thus a bit vain to lament incompleteness of reforms without taking into account existing political constraints. More important, the policies strongly endorsed by the advo cates of the Washington consensus led in several dimensions to important surprises and unexpected outcomes. This is the case for the important output falJ after liberalization. It was not predicted.

©International Monetary Fund. Not for Redistribution TEN YEARS AFTER... TRANSITION AND ECONOMICS

Another unexpected outcome is the asset-stripping that followed mass priva tization in Russia and the Czech Republic. Also on the downside, the development of the Mafia, the strong increase in the size of the hidden economy in former CIS countries, and the resistance of large Russian enterprises to tax collection were not predicted either. On the upside, the development of To wnship and Village Enterprises (TVEs) in China were also an unexpected corollary of decollectiviza tion. To be sure, most of these events were not predicted either by the advocates of the evolutionary-institutionalist perspective. Most can and will be explained ex post. Nevertheless, the importance of these large unexpected outcomes shows the relevance of the emphasis on aggregate uncertainty. Finally, it is important to note that an important convergence has taken place among researchers working on tran sition, certainly in the academic community, on the importance of aggregate uncertainty. In terms of the political economy emphasis, the picture is more mixed. The Central European experience broadly appears to vindicate the Washington consensus view. The experience in China broadly appears to vindicate the evolu tionary-institutionalist perspective. In a way, this should not be seen as a surprise. While there is clearly a difference of emphasis between the two visions, the theo ri es that have been developed (see, for example, Dewatripont and Roland, 1997; Roland, 2000) can make sense of both observations on the basis of a basic trade off involved in the political economy of reforms. In other words, a strategy of sequencing of reforms can relax the ex ante political constraints and gradually build constituencies for further reform and thus enhance political feasibility, while big bang can ensure more irreversibility of reforms in cases where there is a window of opportunity, that is, where ex ante political constraints are less impor tant. This can be seen as fitting both with the Chinese and with the Central European experience. Such results are derived, however, within a framework where aggregate uncertainty is assumed. From that point of view, the Russian experience shows the downside of the Washington Consensus view because of its neglect of aggregate uncertainty. A window of opportunity was used to implement mass privatization and this has been done so as to create irreversibility of reform, but the relative irreversibility thus created has locked the Russian economy in an inefficient situation where interest groups, who gained most from mass privatiza tion (the famous oligarchs), have become so powerful as to block further reform such as tax reform, government reform, stronger law enforcement, and stronger security of property rights (see, for example, Polishchuk, 1999; and Sonin, 1999). This turn of events in Russia also tends to suggest that the poUtical economy view, relying mainly on the support of owners of privatized enterprises, was seri ously one-sided. Not only did this lead to capture and relative lock in, but also the fact that a great majority of the population has suffered from the transition so far, and has resented the strong concentration of wealth created by a privatization process viewed as illegitimate and corrupt, is worrying both for poUtical and democratic stabiHty and for the continuation of reforms. On the effects of partial reform, I would claim that the Chinese experience vindicates the evolutionary-institutionalist perspective that sequencing of reforms does not necessarily staU refom1 progress but can be used to create momentum for

©International Monetary Fund. Not for Redistribution Gerard Roland

further reform. The successes of decollectivization created momentum for reform in the state sector (see, for example, Naughton, 1995; Lin, 1992; and Qian, 1999). While the Chinese reform process has had its ups and downs-its periods of fast progress and periods of stagnation in reform-the general momentum has continued until today. In line with the Dewatripont-Roland (1995) model, the tran sition process started with reforms like decollectivization that delivered enormous welfare gains and created constituencies for more difficult reforms like restruc turing in the state sector. Tnsufficient attention, however, has been given so far to thjs issue of partial reform and the conditions under which it creates momentum or, on the contrary, creates vested interests that block further reform. The existing theories of the polit ical economy of reform currently tend to be too abstract to treat this problem in a sufficiently satisfactory way. One needs political-economic models with more flesh to better understand the economic and political dynamics of sequencing. Recent empirical research on China by Morduch and Sicular (1999) already gives evidence that rents to bureaucrats increase with the pace of reforms, while the benefits of reform are shared with the population at large. The relevance of the Chinese gradualist experience is often dismissed because of the dictatorial character of its regime. It is interesting to note, however, that, despite the political regime, painful reforms have not been brutally imposed on the population. On the contrary, both the sequencing and the design of reforms have been tailored so as to benefit a majority and hurt only a minority. The choice of dual-track price liberalization was in fact even designed to be Pareto-improving and to protect existing rents (see Lau, Qian, and Roland. 2000). It is not clear whether the Chinese reform process would have been politically infeasible if China had been a democracy and whether any democratic system could have sustained such a process. If anything, the absence of democracy has made it more difficult to enforce private property rights and the rule of law and to encourage the development of a newly developing private sector. Recent research (Che and Qian, 1998) suggests that the development of township and village enterprises was a spontaneous response to the specific Chinese institutional situation with the absence of the rule of law and sufficient safeguards against predatory government behavior. (More on this below.) One lesson that emerges from the transition experience is that the political constraints to reform have been less strong in Central Europe, especially in the countries close to Germany, compared with the fo rmer Soviet Union. (Yegor Gaidar did not have the support that Leszek Balcerowicz or Vaclav Klaus bad in Poland and the Czech Republic.) Even when former Communist coalitions came to power in Central Europe, they did not question the direction of reforms, only its speed and its redistributive aspects. Why this striking difference in the importance of political constraints to reform? This is clearly an area where further research is needed. In Roland ( 1997), I suggested that the geopolitical factor has played an impor tant role, a factor that was underestimated in the beginning of transition, certainly by myself. Economists trying to understand transition have generally had the vision of transition as a shift toward democracy and the market. lf we take some

©International Monetary Fund. Not for Redistribution TEN YEARS AFTER ...TRANSITION AND ECONOMICS

historical distance, we can see that transition also represents a very important geopolitical move, that is, the shift of Central Ew-ope and the Baltic states to the We st. To populations in those countries, the single most important factor about transition is the change from the status of a satellite country of the Soviet empire to that of a cmmtry belonging to the western bloc or even the European Union. Transition represents a unique historical opportunity for several nations to get strongly anchored in Western Europe. Not only is this "anchoring" to the European Union desired by nations of Central Europe, it also focuses expectations and gives credibility to the political and economic process of transition. Entry in the European Union implies adopting the political and economic system of the West. The potential reward of belonging to the club of Western nations makes it more worthwhile to undergo the cost of transition and thus to accept the latter more easily. Moreover, the geopolitical factor increases the perceived cost of a policy reversal since it implies the risk of being left out of the western club, a perspective that many in Central Europe would view as disastrous. This geopolitical factor may be strong enough, as a focal point, to explain why countries from Central Europe did not suffer from the type of government collapse, anarchy, and general diff-usion of criminality, inside and outside govern ment, that Russia has been facing (see Roland and Verdier, 1999a and l999b). Ability to enforce the law and to protect property rights seems to be a first-order effect in explaining why Central Ew-ope recovered from their output fa ll while Russia and other countries, not facing the prospect of entry to the European Union. have experienced a continuous decline of output (see also Johnson, McMillan, and Woodruff, 1999a, 1999b). The geopolitical factor in Central European countries was reinforced by a "transition tournament" between the Czech, Hungarian, and Polish governments where each pretended to be the most advanced transition country in the hope of attracting the bulk of fo reign direct investment to the region. The incentives related to the "prize" of such a tournament are strong enough for countries of that size to create credibility for economic transformation. Countries that entered the transi tion race later, like Bulgaria and Romania, have little hope of catching up on the more advanced countries, or even of pretending to do so, and thus have less possi bilities of attracting foreign direct investment. To understand the strength of the geopolitical factor, compare the situation of Central European countries to that of Russia, where this geopolitical factor is absent. By contrast to the former, where transition is seen as a liberation from the Soviet empire and access to the Western club of nations, transition is viewed in Russia as a traumatic experience by large parts of the population. Indeed, transi tion represents the loss, not only of the Soviet empire, but also of territories, such as the Ukraine and the Baltic states, whkh once belonged to Tsarist Russia. Tlus Loss does not only mean a wound to Russian nationalist pride, but it implies uncer tainty for the families of those who have relatives among the millions of Russians Living in the former Soviet republics and who became "immigrants" in former Soviet territories, often with the status of "second-class" citizen. The trauma of the loss of superpower status, similar in a way to the trauma of Gem1any after World War I, could be, to a certain extent, compensated for by economic gains from tran-

4.3

©International Monetary Fund. Not for Redistribution Gerard Roland

sttwn. Unfortunately, this has not thus far materialized for the maJOflty of Russians. No entry of Russia in the European Union is expected nor especially desired. The size of the country implies that the impact of fo reign direct invest ment is likely to be more diluted, thereby reducing the incentives to participate in "transition tournaments." It is thus no wonder that resistance to transition proved much harder in the former Soviet Union, as witnessed by the greater difficulties in hardening budget constraints of enterprises or of adopting stabilization measures. Nor is it clear that no major policy reversal will take place or is expected to take place. If we believe the geopolitical factor played a major role in Central Europe, then it would be seriously flawed to compare transition in Central Europe and Russia without taking this factor into account. To understand the effect of political constraints on the transition process, it is better to look at the experience of large countries that must achieve transition by their own efforts without counting too much on outside help. The comparison between Russia and China is of relevance there, taking into account differences between those two big countries (political regime, level of development, etc.). If the hypothesis of the geopolitical factor explaiJlS the stronger resistance to refom1 or smaller support in Russia and former CIS countries, it also implies that, everything else equal, the cost of reversal of transition policies is much higher in Central European policies than in Russia and the fo rmer CIS countries. In that sense, the perceived sense of haste that existed in Central Europe to implement reforms fast to achieve irreversibWty may have been exaggerated. This last observation is potentially important. Indeed, the political economy theories of transition usually assume that the long-term economic outcome of reforms are independent of the speed of their introduction. ln reality, the speed of liberalization and privatization does have important efficiency and distributive implications. More important, the outcomes of liberalization and privatization teach us something about the two perspectives.

Liberalization and the Output Fall

With hindsight. I would claim that the outcome of liberalization vindicates the evolutionary-institutionalist perspective and that, on the basis of the transition experience, this is the consensus view that has been developing in the academic community. The important output fall that occurred after price liberalization in Central and Eastern Europe was not predicted. Standard textbook economics based on supply and demand at best would have predicted a low supply response to liberalization, but not a negative one. The debate on the output fall was in the beginning inspired mainly by the Washington consensus centered around macro economic policies, asking questions such as whether or not stabilization had been too harsh. When Russia liberalized but failed to stabilize and nevertheless experi enced an output fall, it was clear that new answers were needed. Two transition-specific answers have come up so far and have not yet been refuted. One is the traditional double-marginalization idea from the industrial orgarlization literature. To the extent that central planning created monopoly-like

©International Monetary Fund. Not for Redistribution TEN YEARS AFTER ...TRANSITION AND ECONOMICS

structures without real substitutes across firms. and to the extent that import competition does not play that role, liberalization induces a cascade of price increases and output contraction along supply chains (Li, 1999). The other, newer idea, is that of disorganization. That view takes seriously the idea that markets have not yet been created when liberalization occurs. Due to bargaining ineffi ciencies or to a combination of investment specificity and search frictions related to the prior absence of markets, existing output chains may suffer from acute disruption where the efficiency gains reaped by the producers. who exit the chain. do not compensate the disruption losses for the other producers in the output chain (Blanchard and Kremer, 1997; Roland and Verdier, 1999b). The ensuing macro economic fall in GDP and welfare losses can be very important. These models are de facto inspired by the evolutionary-institutionalist perspective. The relevance of the evolutionary-institutionalist perspective appears even more if we address the question of whether the output fall was an inevitable by product of liberalization. Here, the Chinese experience is again helpful because it gives a negative answer to that question. A transition-specific institution has been created to prevent the output fall associated with liberalization: it is the dual-track liberalization. The dual-track liberalization has several interesting properties. Prices are liberalized at the margin so that the market information obtained from price liberalization is the same as what would be obtained under full price liberal ization. In the absence of preexisting markets, the most interesting properties of the dual track are (I) that it allows, by construction, the achievement of Pareto improving gains from liberalization, which is interesting from the political economy point of view because it is a way of overcoming potential resistance to price liberalization due to its distributive effects; and (2) also by construction, it prevents the output fall by maintaining past contractual obligations from the plan. It is in a way surprising to observe not only that the dual-track approach had not been proposed by academic economists in the context of transition (in Eastern Europe at least) but also that it took several years before economists started to understand the advantages of the dual-track system (see Byrd, 1991; Sicular, 1988; Lau, Qian, and Roland, 1997 and 2000; Li, 1999). While the dual track has worked well in China, further research is needed to understand some important aspects of the dual track, such as how it can be enforced credibly and prevent the ratchet effect. This may also help us understand why dual-track liberalization was not applied in Eastern Europe and the countries of the former Soviet Union. If we ignore politics, it is possible to argue that the dual-track liberalization could have been applied in the context of the Council for Mutual Economic Assistance (CMEA) trade. The CMEA breakdown (as well as the breakup of the Soviet Union) has been considered the single most important explanatory factor for the general fall of output in the region (see. for example, Rodrik, 1994), but it has generally been perceived as an exogenous shock. The breakdown of CMEA, however, was not exogenous. It was decided in early 1990, when the Czechoslovak and the Polish governments insisted on regaining their freedom of export with respect to CMEA agreements. The Soviets at that time responded by insisting that imports from the Soviet Union would from 1991 onward be paid at world prices and in bard currency. From the economic point of view, it would have been better,

©International Monetary Fund. Not for Redistribution Gerard Roland

and possible, to implement a version of the dual-track system between former CMEA countries to avoid the strong trade disruption of 1991. However, here again politics and the geopolitical factor have played an impor tan t role. Individual Central European countries wanted to leave the Soviet bloc as quickly as possible and to be the fi rst to knock at the door of the European Union. The CMEA breakdown was thus an economic consequence of the political will prevailing in Central European countries to leave the Soviet bloc. One may argue that the dual-track approach could not have been in1plemented in the former Soviet Union and Russia because of the government collapse that followed the implosion of communism af ter the failed putsch of 1991. As shown by Roland and Ve rdier (1999a). however, the dual-trac k system itself can be seen as an instrument to prevent government collapse in a credible way. Again, careful research is needed to understand whether conditions existed to implement and enforce the dual-track system itself in a credible way.

Government Collapse

Another area where the evolutionary-institutionalist perspective is vindicated and where a consensus view has been developing fast within the academic community, aJ beit with strong differences in emphasis, relates to the effects of governmentcollapse, mainly in many CIS countries. Governmentcollapse was not an important concern in the Washington consensus view. The concern was mainly to cut down the size of government and to "get the state out of the economy." Government collapse would have been seen as a second-best option where markets would emerge at full speed but there may be too few public goods-a lesser evil compared to the evils of communism. The emergence of organized crime, its predatory racketeering ac tivities, and Mafia-related internal corruption of government have had deleterious effects on private sector growth. Again, this evolution was not predicted and was strongly underestimated. As of today, the Mafia is still seen by many as a suboptimal contract enforcement ag ency. Also, the more comprehensive view of institutions of the evolutionary institutionalist perspective seems to be vindicated given the experience of govern ment collapse. While it is al ways possible to claim for each transition country that there were shortcomings in the legal framework. in Russia, one cannot claim th at legal reform was neglected. Many important laws were adopted, often with the help of prominent scholars-be they corporate 1aws, commercial laws, or financial regulations. Nevertheless, law enforcement is a real problem and confidence in courts is lower in Russia than in Central Europe (see, for example, Black, Kraakman, and Tarassova, 2000; Johnson, McMillan, and Wo odruff, 1999b). The adequate social norms for a market economy have not yet emerged, and the level of bu siness trust remains low.

Privatization

The experience with privatization also tends to vindicate the evolutionary institutionalist perspective against the Wash ington consensus view. The view,

©International Monetary Fund. Not for Redistribution TEN YEARS AFTER ...TRANSITION AND ECONOMICS according to which speed was of the essence in privatization in order to stop asset stripping by incumbent managers, has been refuted by reality as many managers in state-owned enterprises did show early signs of restructuring. The doomsday predictions on the consequences of too low a speed did not materialize in coun tries like Poland and Hungary where privatization took place in a gradual way. The prediction of generalized asset stripping was wrong. lt made sense for good managers to try to attract private investors and to engage in early defensive restruc turing for that purpose. Also, both theory and evidence show the importance of using privatization to achieve efficient matching of managers and assets (see, for example Bolton, and Roland, 1992; Barberis and others, 1996). Insider privatiza tion did not achieve that objective. Insider privatization did not in general lead to significant improvement in performance (see, for example, Frydman and others, 2000), whereas the Washington consensus view claimed that any privatization was always better than no privatization.s Moreover, vested interests created by insider privatization may have made ulterior privatization more difficult to achieve (Aghion and Blanchard, 1998) and may have also maintained or aggravated the soft budget constraint syndrome in firms privatized to insiders (Debande and Friebel, 1995). This soft budget constraint syndrome usually takes the form of tax arrears (Schaffer, 1998). Further research should determine the extent to which the combination of free distribution ofassets and tax erosion in Russia has reinforced tendencies to government coUapse. In China, we have seen the emergence of a very original institution: the TYEs owned by township and village governments. The Washington consensus view cannot make sense of TVEs, except to claim that they are pseudo-private enter prises, which we know is not the case. Theories developed in the transition litera ture suggest that TYEs operate under conditions of hard budget constraints, whkb are very important for efficiency (Qian and Roland, 1996 and 1998), and also that, because of the public goods they contribute to create in their activity, they are more protected from government predation than private firms in China (Cbe and Qian, 1998).

Organization of Government The Chinese experience, where privatization remained taboo until recently, also shows the importance of reform in the organization of government with the decentralization of government and the development of different forms of compe tition between local governments that can be made to work in favor of market development. Also, the fiscal federalism arrangements, whi.ch make local govern ments residual claimants on any increase in the tax base, creates partial alignment of interests of bureaucrats with the development of markets and entrepreneurship (on all this, see Qian, 1999). In Russia, reform of the organization of government was relatively neglected, as the main focus of reform was implementation of mass privatization.

SThis does not mean that no fonn of privatization had a positive effect. For recent studies, see Havrylyshyn and McGettigan (2000), and Djankov and Murrell (2000).

©International Monetary Fund. Not for Redistribution Gerard Roland

With hindsight, reform of government with the purpose of building stable constituencies for reform should have been a priority of reform sequencing in Russia. The unexpected collapse of communism in 1991 created a formidable opportunity to create a democratic constitution with appropriate separation of powers and checks and balances between branches of government as well as between the center and provinces. Legitimate and accountable institutions coming out of a constitutional process wouJd have helped to build more solid constituen cies for refonn and create a government structure with legitimate authority. There is no clear view among researchers about what should be the best constitution for Russia. Since legitimacy is fundamental, the appropriate constitution should not necessarily result from a blueprint but must be the result of a serious consensus building deliberative process, as was the case with the successful democratic tran sition in Spain. As of today, government reform is still a priority in Russia, but the initial aspirations to democracy have been strongly dampened by the turmoil of transition under Ye ltsin.

Soft Budget Constraints A final difference we will mention between the Washington consensus view and the evolutionary-institutionalist perspective concerns the question of soft budget constraints. Here the dominant thought was that hardening budget constraints was only a matter of exogenous choice by policymakers without thinking about the institutional factors or the factors in the environment of firms that contributed to credibly harden the budget constraint of firms: privatization, demonopolization, government reform, banking reform, and so forth. The subsis tence of soft budget constraints in different forms, even in the advanced transition countries, shows that hardening budget constraints is not just a matter of political will but a matter of devising institutional mechanisms that create credibility for hardening. While there has been quite a body of theoretical work on soft budget constraints, empirical work on the issue is only beginning and should be a priority in research. Despite the lack of direct empirical tests, the general evidence on enterprise behavior suggests a difference in enterprise behavior with a rapid hardening of budget constraints in the accession countries and persistence of soft budget constraints in most other countries. Here again, the geopolitical factor may have played an important role in shaping expectations early on in accession countries. Further research should give us a clearer picture of the hardening of budget constraints.

Ill. Conclusion The broad assessment developed above tends to show, on the basis of the tran sition experience, that the evolutionary-institutionalist perspective is more complete and adequate than the Washington consensus view. There is an increasing consensus among professional economists that the "Washington consensus" view with its so-called trinity of transition (liberalization, stabilization,

©International Monetary Fund. Not for Redistribution TEN YEARS AFTER ...TRANSITION AND ECONOMICS

and privatization) is a misguided recipe for a successful transition. While profes sional economists do not deny the need to liberalize, stabilize, and privatize, they increasingly recognize that these policies cannot achieve their goals without the existence of institutional underpinnings of capitalism appropriate to the specific conditions of each country. In practice, there has been a growing convergence within the academic community, if not a consensus, toward the evolutionary institutionajjst perspective.6 While the above discussion has attempted to propose a synthesis of the main lessons of transition, an important word of caution is necessary, in line with the evolutionary-institutionalist perspective. While general policy conclusions can be drawn from theory and empirical research on transition-the point is of course valid beyond transition-there is still a long distance between general policy conclusions and direct polky recommendations. While we can draw general lessons from the economic failures of transition in Russia, in the 1990s for example, it is still quite a different matter to draw a complete and convincing counterfactual of how things could have evolved. given the conditions of deci sionmaking. Finally, I would like to draw attention to two areas that, with hindsight, have appeared important for transition, but oo which relatively little research has been done. The first one relates to inequality of wealth and income. While the increase in inequality of income and wealth was predicted, the patterns of increasing inequality are quite different across countries (Milanovic, 1998; Garner and Terrell, 1998; Keane and Prasad, 2000). What are the political and economic effects of this increase in inequality? While this is a general question for economics (Persson and Tabellini, 1994), it is particularly relevant in the case or transition. In particular, one would like to go beyond the median voter model and understand better the political channels through which ao increase in inequality affects political decision making: the relative role of electoral politics and special interest politics, the policy and political coalition formation process, and so forrh. The second theme is that of social behavior, social norms, and social capital. The question of social norms is a general and important question for social science, and this is an area that has been under-researched by economists at least during the transition process. An obvious route of investigation that has not been used in formal analysis of transition processes so far, at least to my knowledge. is the use of evolutionary game models. Within the evolutionary-institutionalist perspective, this seems a natural route to take, given the usefulness of evolutionary game theory in selecting equilibria.

6Jt is interesting to note from that point of view that at the Fifth Nobel Symposium in Economic� devoted to the economics of transition, om of six sessions. only one was devoted to macroeconomic dc\cl opments while there were fivesessions on institutions. three of which were devoted to the organization of government and two on contracts.

©International Monetary Fund. Not for Redistribution Gerard Roland

REFERENCES

Aghion, Phi llipe., and Olivier Blanchard. 1998. "On Privatization Methods in Eastern Europe and Their fmplications," Economics a/ Tr ansition, Vol. 6 (May), pp. 87-99.

Barberis, Nicholas, M. Boycko, A. Shleifer. and N. Tsukanova, 1996, ·'How Does Privatization Work? Evidence from the Russian Shops." Journal of Po litical Economy, Vol. I 04, (August), pp. 764-90.

Black, B., R. K.raakman, and A. Tarassova, 2000, "Russian Privatization and Corporate Governance: What Went Wrong?" Stanfo rd LAw Review, Vo l. 52, No. 6, pp. 1731-808.

Blanchard, Olivier. and Michael Kremer, 1997. "Disorganization;· Quarterly Journal of Economics, Vo l. 112 (November). pp. 1091-126.

Bolton, Patrie, and Gerard Roland, 1992. ''Privatization in Central and Eastern Europe," Economic Policy: A European Fo rum, Vol. 7 (October), pp. 27 1-309.

Byrd, William A., 1991, The Market Mechanism and Economic Reforms in China (Armonk, New York: M.E. Sharpe).

Che, Jiahua, and Yingyi Qian. 1998. "Insecure Property Rights and Government Ownership of Firms," Quarterly Journal of Economics, Vo l. L 13 (May), pp. 467-96.

Debande, Olivier, and Guido Friebel, 1995, "Privatization, Employment. and Managerial Decision-Taking" (unpublished; Bmssels: European Center for Advanced Research in Economics and Statistics, Universite Libre de BmxeJJes).

Dewatripont, Mathias, and Gerard Roland, 1995, "The Design of Reform Packages Under Uncertainty," American Economic Review, Vol. 85 (December), pp. 1207-23.

---, 1997, "Transition as a Process of Large-Scale Institutional Change," Chapter 2 in Advances in Economic Theory, ed. by David Kreps and Kenneth Wallis (New York: Cambridge University Press). pp. 240-78.

Djankov, Simeon, and Peter Murrell, 2000, "The Determinants of Enterprise RestructUJing in Transition: An Assessment of the Evidence" (Washington: World Bank).

Frydman. Roman, Cheryl Gray, Marek Hessel, and Andrzey Rapaczynski, 2000, "When Does Privatization Work? The Impact of Private Ownership on Corporate Performance in the Transition Economies," Quarterly Journal of Economics. Vo l. 114 (November), pp. I 153-92.

Gamer, Thesia, and Katherine Terrell, I 998, "A Gini Decomposition of Inequality in the Czech and Slovak Republics During the Tra nsition," Economics of Transition, Vol. 6 (May), pp. 23-46.

Havrylyshyn, Oleh, and Donal McGettigan, 2000, "Privatization in Transition Countries: A Sampling of the Literature;· Post-Soviet Affa irs, Vol. 16 (July-September), pp. 257-86.

Johnson. Simon, John McMillan, and Christopher Woodmff, 1999a, ''Contract Enforcement in Transition;· paper presented at the Fifth Nobel Symposium in Economics: The Economics of Transition, Stockholm, Sweden; also published as CEPR Discussion Paper No. 2081 (London: Centre for Economic Policy and Research).

--, 1999b, "Property Rights, Finance, and Entrepreneurship'' (unpublished; Cambridge, Massachusetts: Economics Department. MIT).

Keane, Michael P., and Eswar Prasad. 2000. ·'Inequality, Transfers, and Growth: New Evidence from the Economic Transition in Poland," IMF Working Paper 00/117 (Washington: International Monetary Fund).

Lau, Lawrence, Yingyi Qian, and Gerard Roland, 1997. "Pareto-Improv ing Economic Refom1s Through Dual-Track Liberalization;· Economics Letters, No. 55 (August), pp. 285-92.

©International Monetary Fund. Not for Redistribution TEN YEARS AFTER ...TRANSITION AND ECONOMICS

--, 2000, "Reform Without Losers: An Interpretation of China's Dual-Track Approach to Reforms." Journal of Political Economy. Vol. 108 (February). pp. 12�3. Li. Wei, 1999, "A Tale of Two Reforms," Rand Journal of Economics, Vo l. 30 (Spring). pp. 120-36. Lin, Justin Yifu, 1992, "Rural Reforms and Agricultural Growth in China," American Economic Review, Vol. 82 (March), pp. 34-51. Milanovic, Branko, 1998, Income Inequality and Poverty During rhe Transition from Planned to Market Economy, World Bank Regional and Sectoral Studies (Washington, World Bank). Morduch, Jonathan, and Terry Sicular, 1999, "Politics, Growth. and Inequality in Rural China: Does It Pay To Join the Party?" Harvard Institute for International Development, Discussion Paper No. 1832 (Cambridge, Massachusetts: HliD) Murrell, Peter, 1992, "Evolution in Economics and in the Economic Reform of the Centrally Planned Economies;' in The Emergence of Market Economies in Eastern Europe, ed. by Christpher Clague and Gordon Raisser (Cambridge, Massachusetts: Blackwell), pp. 35-53. Naughton, Barry, 1995, Growing Out of the Plan: Chinese Economic Refo rm, 1978-1993 (New York: Cambridge University Press). Persson, Torsten, and Guido Tabellini, 1994, "Is Inequality Harmful for Growth?" American Economic Review, Vol. 84 (June), pp. 600-2 1.

--, 2000, Political Economics: Exp laining Economic Policy (Cambridge, Massachusetts: MlT Press). Poli shchuk, Leonid, 1999, "Distribution of Assets and Credibility of Propetty Rights" (unpub lished; College Park, Maryland and Moscow, Russia: University of Maryland and New Economic School). Qian, Yingyi. 1999, "The Institutional Foundations of China's Market Transition," in Proceedings of rhe World Bank's Annual Conference on Developmellf Economics 1999, ed. by Boris Pleskovic and John Stiglitz (Washington: World Bank), pp. 289-3 10.

---. and Gerard Roland, 1996, "The Soft Budget Constraint in China," Japan and the World Economy, Vo l. 8, No. 2, pp. 207-23. --, 1998, "Federalism and the Soft Budget Constraint," American Economic Review, Vol. 88 (December). pp. 1143-62. Rodrik, Dani, 1994, "Foreign Trade in Eastern Europe's Transition: Early Results," in The Tr ansition in Eastern Europe, ed. by Olivier Blanchard, Kenneth Froot, and Jeffrey Sachs (Chkago, Illinois: NBER and Chicago University Press). pp. 319-56. Roland, Gerard, 1997, "Political Constraints and tbe Transition Experience" in Lessons from the Economic Transition: Central and Eastern Europe in the 1990s, ed. by Salvatore Zecchini (Boston: Kluwer), pp. I 69-88. --, and Thierry Ve rdier, 1999a, '·Law Enforcement and Transition" (unpublished; Brussels, Belgium: European Center for Advanced Research in Economics and Statistics, Universite Libre de Bruxelles, and Departement et Laboratoire d'Economie Theorique et Appliquee). --. 1999b. "Transition and the Output Fall,'' Economics of Transition, Vol. 7 (May). pp. 1-28. ---, 2000, Transition and Economics: Politics, Markets, and Fi rms (Cambridge. Massachusetts: MIT Press).

Sachs, Jeffrey, Wing Tye Woo. and Xiaokai Yang, J 999, "Economic Reforms and Constitutional Transition," Annals of Economics and Finance, Vol I (November), pp. 435-9 1.

©International Monetary Fund. Not for Redistribution Gerard Roland

Schaffer, Mark, 1998, ··oo Firms in Transition Economies Have Soft Budget Constraints? A Reconsideration of Concepts and Evidence," Journal of Comparative Economics. Vol. 26 (March), pp. 80-103.

Sicular, Thieny, 1988, "Plan and Market in China's Agricultural Commerce," Journal of Political Economy, Vol. 96 (April), pp. 283-307.

Sonin, Constantin, 1999, "Inequality, Property Rights Protection, and Economic Growth in Transition Economics: Theory and Economic Evidence," CEPR Discussion Paper No. 2300 (London: Centre for Economic Policy Research).

Stiglitz, Joseph, 2000, "Development Thinking at the Millenium: Keynote Address;· in Annual World Bank Conference on Development Economics 1999, ed. by Boris Pleskovic and John Stiglitz (Washington: World Bank).

Williamson. John, 1990. "What Washington Means by Policy Reform." in Larin American Adjusrmem: How Much Has Happened? ed.by John Willian1son (Washington: Institute for international Economics).

---, 2000, ''Whar Should the World Bank Think About the Washington Consensus?" World Bank Research ObserveJ; Vol. 15 (August), pp. 251-64.

Recovery and Growth in Transition: A Decade of Evidence

OLEH HAVRYLYSHYN*

f This paper reviews a range of studies that examine diferences in growth perfor mance among transition countries. There is a consensus in the literature about the core elements of transition and the policies necessary fo r sustainable growth, although considerable differences remain about how to implement these policies and about their proper sequencing. The empirical work identifies stabilization and structural refo rms (e.g., market liberalization, private ownership) as important determinants of growth, but underlines the role of initial conditions and institu- ' tions. There is divergent evidence, however, on the importance of specific refo rms. Traditional fa ctor inputs have as yet no role in explaining growth. [JEL 040, P20]

he first comprehensive program of reforms in a centrally planned socialist economy is widely recognized to be Poland's late 1989 package under the new TSolidarity government. This dating permits analysts to think of a decade anniver sary already in 1999 and certainly in 2000, hence the ensuing spate of 1 0-year conferences and retrospectives. More important than the decadal marking point is the reality that enough time has passed to allow substantial quantitative and qual itative observations, to see the variation in progress of transition toward the market, to note country differences in economic performance indicators such as GDP growth, exports, foreign investment, and to identify broadly the more successful and less successful cases. This paper addresses the narrow issue of growth in GDP, by providing a review of a growing literature explaining the differ-

*Oleh liavrylyshyn is Deputy Director in the IMF's European ll Department. He thanks Ratna Sahay, Richard Haas, Bogdan Llssovolik, and Thomas Wolf for their useful comments; Bogdan Lissovolik for contributing considerable effort to preparation and verification of the tabular material; and Joan Campayne for an excellent job of text preparation.

©International Monetary Fund. Not for Redistribution Oleh Havrylyshyn

ences in growth performance among transition countries. The focus is on non Asian transition economies. in keeping with much of the literature's separate treat ment of China and others in Asia. Transformation from a centrally planned to a market economy is a multi faceted process of political, economic, social, and-not least important, as is increasingly recognized-institutional changes. Within the economic sphere, growth of aggregate output may be considered only one of many dimensions in the process. But such a naJTow view would understate considerably the role of recovery and growth in GOP. True, growth is only the final link in a long and complex chain of transformation that involves, in approximate sequence, policies to implement market mechanisms; structural changes across and within sectors and firms (in other words, reallocation of resources): improvements in efficiency and new investments; and finally, recovery of output and sustained growth. However, even in the broadest '·system paradigm" of transition (Kornai, 1998), growth of output and the attendant improvement in the well-being of the populace is, arguably, the key purpose of changing the system. At a minimum, strong sustained growth is a necessary outcome of a successful transition, and-given the lack of a yardstick for measuring progress in the ex ante policy and institutional movements toward the market-growth may serve as a very good ex post proxy for such progress. J

I. Analytical Framework for Growth in Transition

It continues to be popular to say that there is no theory to guide the practical process of transition, only theories of capitalism and socialism. This may still be true in the sense that a new consensus paradigm has not emerged from the vast literature on transition, though it is not clear how much a unified, cohesive theory is needed to understand the main developments.2 To the extent it is useful to have a compact rather than complex analytical framework, it is not that difficult to cobble together from selected key writings a workable "model" of transition or transformatjon. Kornai (1994), in describing the special circumstances of the "transformational" recession compared with a market economy recession, high lights two key changes that are needed: fo rcing a move from a sellers ' to a buyers' market (via price liberalization), and enforcing a hard budget constraint (via privatization and elimination of various government support mechanisms such as budget subsidies, directed low-cost credits, and tax exemptions). This provides the two principal incentives for profit-maximizing market behavior of all economic agents. Blanchard ( 1997) defines the core process of actual change as comprising two elements: reallocation of resources from old to new activities (via closures and bankruptcies combined with establishment of new enterprises); and restructuring

'The central role of growth rates as reflections of successful transition is implicit in several studies projecting long-term future growth potential. motivated by the question of how long it might take to catch up to EU country levels (e.g .. Fischer, Sahay. and Vegh.l998a: Sachs and Warner, 1996). 2Komai ( 1998) explores the possibility of what he prefers to call a "system paradigm": one key argu ment he makes is that transition by definition does not need a paradigm or theory-only the beginning and end-point systems do.

©International Monetary Fund. Not for Redistribution RECOVERY AND GROWTH IN TRANSITION: A DECADE OF EVIDENCE

within surviving firms (via labor rationalization, product line change, and new investment). These two changes, which are very reminiscent of the Schumpeterian concept of "creative destruction," should be stimulated by the new incentives.3 1n the end, the transformation moves the economy to a resource allocation state consistent with its comparative advru1tage. The key policy actions needed to put in place Kornai's new incentives are described in many works (including those by Kornai (1994) and Blanchard (1997)), and are exemplified by the list of Fischer and Gelb (1991): macroeconomic stabilization; price and mru·ket liberalization; liberalization of the exchange and trade system; privatization of state-owned firms; establishing a competitive environment with easy market entry and exit: and redefining the role of the state as the provider of macro stability, a stable legal framework, and enforceable property rights, and occasionally as a corrector of market imperfections. The above simplificationdoes not do justice to the large theoretical and policy literature, and the continuing shru-p debates about transition. One school of thought advocates following broadly the Washington consensus elements--essentially those listed above-and includes the view that more rapid and earlier implemen tation is generally better (in various shades of nuancing, with Sachs ( 1996) at the extreme end). A second school of thought argues-in theory if not in practice that transition can occur too quickly, which will cause more costly disruption than beneficial restructuring and therefore risk undermining the will to continue. Aghion and Blanchard (1993); Blanchard and Kremer (1997); and Roland and Verdier (1999) exemplify the theoretical work along these lines. Stiglitz ( 1999) with whom the former do not necessarily agree-argues ex post that excessive speed was indeed a problem in practice ru1d explains the dramatic fa ilures of priva tization and lack of recovery in many transition countries, including, for example, Russia. A third school of thought focusing on institutions argues, as Murrell ( 1992), that stabilization and liberalization are needed but will not have the intended results if the institutions of market operations are inadequately devel oped.4 But all of these debating schools appear to agree on what the transition process is, and their views are in general compatible with the Fischer-Gelb description of the core elements. Their differences are more about how to conduct poHcy to best achieve the results of transformation. Nevertheless, these debates are relevant and find reflection (mostly, ad hoc) in the specifications of econometric analysis of growth in transition reviewed later in this paper. From the core concept of transformation defined above there follow some implications for growth, which differentiate transition economies from market economies and provide the basis for the empirical analysis of determinants of recovery in transition. First, output will necessarily decline initially under the new buyer's market and hard budget constraints, since unsalable goods accumulate and

3The EBRD Tra nsition Report ( 1997) provides an excellent review of the conceptual framework in the "creative destrUction" spirit. as well as empirical analysis of structural changes in the transition so far. 4£t is frequently said by critics (e.g., Roland (2001)in this Special Issue of fMF Staff Papers) that the Washington Consensus erred greatly by omitting institutional development: the Fischer and Gelb (1991) article is but one of many that show that this was simply not U1le. Of course, it is still possible to argue that institutions, while mentioned in passing. were not given enough weight in practice.

©International Monetary Fund. Not for Redistribution Oleh Havrylyshyn

signal the need for cutbacks in production. FUither elimination of the wastage found under the old regime has to precede creation of the new regime, adding to the production cuts. Second, growth of the new regime will not occur until the new incentives are in place and made credible; that is, the sooner reforms achieve a hard budget and liberal price environment, the sooner reallocation and the restruc turing of old and creation of new production can begin. The lag can be attributed to the "disorganization" effect of Blanchard and Kremer ( 1997), as central plan ning mechanisms are not immediately replaced by market coordination mecha nisms. Third, in the early recovery period, a variety of effi ciency improvements are more likeLy to be useful than an expansion of either investment or labor factor inputs. With labor fully (but inefficiently) employed, and much capital stock again inefficient-accumulated from earlier Soviet investments, the room for factor-expansion growth is limited. Indeed, as Easterly and Fischer (1995) and De Broeck and Koen (2000b) suggest, growing ineffi ciency was one reason the Soviet system collapsed. Starting from a Soviet-period (inefficient and distorted) output level as shown in Figure Al in the Appendix, five types of mechanisms, many of which may be simultaneous or overlapping, can help increase output: recovery of underutilized capacity; elimination of egregious waste of labor, capital, and materials ex efficiency, improvements); efficiency gains from a more appropriate combination of capital and labor (factor efficiency); efficiency gains from resource reallocation toward goods in which a country has a comparative advantage or for which there is unsatisfiedconsumer demand; and output expansion via new net investment and employment increases.s It is notable that only the last item above involves factor expansion, while aU the others are some form of eficiencyf gains. The empirical evidence described below will confirm that the early period of transition recovery is largely based on efficiency gains and that investments-physical or human capital-are not the important determinant generally seen in the growth literature.

II. The Record on Growth

Output declined substantially from 1989, for 25 non-Asian transition coun tries.6 Recovery began only hesitantly in 1993 in two or three countries in Central Europe, then spread gradually and accelerated by the end of the decade (Table l). By 1995, virtually all Central European countries plus the Baltic countries were experiencing this recovery, while further east only 2 of 12 countries were begin-

5[[ is a simplification to say that all the efficiency improvements (all but the last item in the list) can come about without new net investments; what is meant here is that the investment required is often small. Also, such efficiency improvements can take place at the sector or finn level even if aggregate net invest ment in the economy is zero. since new gross investment is directed not to replace depreciated stocks in "old" industries but to expand it in the "new" ones. 6Data in this section use the official GDP measure, excluding what many have demonstrated is a large underground economy-see Johnson, Kaufman n. and Shleifer ( 1997)-and uncorrected for other possible errors-see Bloem, Cotterell, and Gigantes ( 1996). The reasons for this choice are simple: first, only GDP data are systematically available for a long period for all coumries; and second, different estimates of unofficial economy values give differing resulls. There is at least one study, Loungani and Sheets ( 1997),

©International Monetary Fund. Not for Redistribution RECOVERY AND GROWTH IN TRANSITION: A DECADE OF EVIDENCE

ning to grow (Armenia and Georgia after their civil conflicts came to an end). It was only by 1997 that a majority of the CIS7 countries saw positive growth. The Russian crisis of 1998 caused a brief slowdown in this trend for 1998-99, but by 2000 a strong recovery was evident in the whole region, particularly in the CIS. Thus, on balance, the first half of the 1990s was largely a period of decline, whHe the second half was marked by a gradual spread of recovery from the deep and long transitional depression, only briefly intenupted by the spillovers from the 1998 Russian crisis. It is generaUy agreed that the depth and length of the decline has been greatest for the CIS countries, somewhat less for the Baltics, and much less in Central and Southeast Europe. Fischer and Sahay (2000), for example, give respective decline values through 1998 of about 54 percent, 43 percent, and 28 percent. Growth rates since recovery began have been quite high in general at 4-6 percent, but, except for a couple of oil economies (Azerbaijan and Turkmenistan), this falls short of the levels in East Asia and China during their boom years. Indeed, the best non-oil performer, Poland, has approached 7 percent growth in only two of eight years of growth. At the other end of the range, Ukraine had continuous decline through 1999, and several countries bad an enatic pattern of growth followed by decl.ine Moldova, the Kyrgyz Republic, and Romania. Turkmenistan has only shown posi tive rates as a partial rebound after disastrous years of sharp decline. In summary, growth is spreading but there is a lot of variation across countries; it is too early to speak of sustained growth except in Central Europe and the Baltics; and even the better petformers still fall short of the growth levels seen earlier in Asia. Three broad categories of transition economies can be identified:8 about 15-16 countries with consistent growth; 3 countries with sharp reversals, and about 5-6 countries with almost continuous decline until 1999. The first group includes the Baltics and Central Ew-ope, in which the growth rates since recovery began have averaged approximately 4-5 percent-except for the brief but sharp dip in the Baltic countries after the Russian crisis. These cases form the strongest basis for the argument that better stabilization and structural refom1 policies produce better performance. Consistently positive growth is also observable in several CIS countries, but wilh rather cl.ifferent stories for each of tl1em. Annenia and Georgia had strong rebounds from civil war periods; Azerbaijan had an oil based expansion; growili in ilie Kyrgyz Republic is partly attributable to new gold production; and Belarus and Uzbekistan may have been special cases of delayed refom1s, delaying the post-Soviet decline-at ilie very least, iliey provide fodder for debate (see Zettelmeyer, 1998, on the Uzbekistan "puzzle"). The second

which finds Lhat using electricity consumpLion as a proxy for growth does not materially affect the results of growth empirics. The bias may be thought to mean Lhat less refom1ed economies with more under ground activity have higher growth than officially measured but. suprisingly. a recent paper by Aslund (200 I) that Lries to correct several measurement problems concludes thaL these corrections strengthen the positive correlation between growth and reforms. The issue nevertheless remains debatable. 7The CIS (Commonwealth of Independent States) comprises the following countries: the Republic of Armenia, the Azerbaijan Republic, Lhe Republic of Belarus, Georgia, the Republic of Kazakhstan, the Kyrgyz Republic, the Republic of Moldova, the Russia Federation, the Republic of Taj ikistan, Turkmenistan, Ukraine, and the Republic of Uzbekistan. BThe remainder of this section draws upon Havrylyshyn and others ( 1999).

©International Monetary Fund. Not for Redistribution Table 1. GOP Growth in Transition Economies by Country Group (Transition Time)

Nwnber Cumul:aJi1.'e of Years Decline A'uage Fir

Consistenl �

Cenl11ll Europe 1990-92 100.0 9-1.0 85.6 85.8 87.5 91.2 96.2 IO(tO •.• ... -6..0 -90 Q2 2.0 4.2 55 �-0 4.9 2.7 -14.4 �.2

Baltic countries 1992 100.0 74.0 �-' 61.8 63.6 66.1 71.6 7�.6 .•• -26.0-13.4 -35 2.8 �-0 8.3 .u -25.4 2.7 -38.2 4.6

�-3 CIS I� 100.0 7.;.3 63.8 56.6 �.4 56.8 60.8 64.::! .•. -25.1-]�.I -11.2 -3.9 7.0 5.7 -35.8 35 -45.6 5.8

CIS�d. Belanb and 0 1992 100.0 66.6 53.0 �5.9 44.7 47.4 51.0 53.6 ... -33.-1-20.4 -13.3-2.7 15 5.2 -46. 4 3.3 -553

53.9 . CIS 1992 100.0 83.2 15.5 61.3 56."' 53.0 5::!..3 . . •• -16.8 -9.2 -18.8-7.6 -1.9 -1.6 -1.4 . . . . . -17.7 5.3 -17.7 0.5

Source: Havryly�hyn and others (1999). Table 15. Annex V.

©International Monetary Fund. Not for Redistribution RECOVERY AND GROWTH IN TRANSITION: A DECADE OF EVIDENCE group-Albania, Bulgaria, and Romania-saw early recoveries with reversals to negative growth, apparently related to inadequate refom1 efforts; followed by policy corrections and a new recovery at least for the first two. Finally, the tblrd group, which encompasses much of the CIS including Russia and Ukraine, saw limited or no growth until 1999, well after the impact of the Russian crisis was overcome. The eventual strong growth of Kazakhstan and Russia was partly driven by favorable energy export prices, though the delayed cumulative effects of earlier reforms cannot be excluded as an explanation; the latter factor may have been even more important in Ukraine. As of 1999 (2000 growth rates are preliminary), only fow· countries had reat tained or surpassed their officially measured output levels in 1989.9 Each of these was in Central EUJ·ope, with real GOP in 1999 in Poland exceeding that of 1989 by 22 percent, Slovenia and Slovakia exceeded 1989 growth by about 5 percent, and Hungary just reached its 1989 level. At the other extreme, with measured output still less than one-half the immediate pretransition level, were five CIS countries. One of these, Georgia (47 percent of 1991 output), bas actually been growing for four years. The other four are all in the third grouping (little or no growth): Tajikistan and Turkmenistan (about 45 percent of the 199 I level), Moldova (40 percent), and Ukraine (41 percent). Wblle growth by 2000 was almost universal, there were clearly many differ ences in the record, and many different explanations. The general tendency that central European countries and the Baltics had better performance than countries of the CIS suggests the importance of better policy (stabilization and structural reforms) and different specific conditions such as wars, oil, or other natural resources, different starting points, or initial conditions. The very recent surge of growth in the CIS group is too short-lived to assess, but may reflect a rebound from very low levels of output reached after decline, perhaps analogous to the long-term convergence notion in growth empirics. Section III reviews more systematic econometric evidence, which has tried to elaborate and quantify the various determinants noted above.

Ill. Empirics of Growth in Transition The broader literature on transition has grown rapidly, and many recent arti cles address, at least peripherally, the issue of success, performance, and output growth. This smvey focuses more narrowly on studies that undertake a quantita tive analysis of growth determinants. Table 2 lists chronologically and summarizes 23 such studies, which form the core of this paper's analysis. As noted earlier, the focus is on non-Asian economies, that is, Central Europe, the Baltic countries, and

9'fhe possible bias in official statistics understating growth, as noted in footnote 6. is in this case in the opposite direction. The actual level of GOP before recovery is in most cases probably exaggerated. owing both to the probable underestimation by official statistics of the size of the unofficial economy during the transition and to the likelihood that initial GDP was in most cases overstated since it did not reflect the welfare losses from disequilibrium pricing and associated shortages and queues. The recent paper by Aslund (2001) explores this in some detail, looking at underground activity as well as price distortions in Soviet period. inadequate negative valuation of useless products, etc.

©International Monetary Fund. Not for Redistribution Oleh Havrylyshyn

the CIS.IO The coverage varies, but generally covers the period since 1989, and in one case through 1998; thus, the sample size is as tittle as 18-20 observations (cross-section regressions) to as many as 225 (panel of 25 countries, 9 years). Perhaps the ft rst comprehensive study to ask the question "what determines differences in growth rates across transition countries" was that of de Melo, Denizer, and Gelb (I 997). 11 The econometric specification there-with variables representing initial conditions, liberalization, or structural reform, financial stabi lization-largely set the tone for speciftcations and indeed the general conclusions in later studies. First, consider some aspects of coverage and methodology, summarized tele graphically in the last two columns of Table 2. Later studies naturally cover more years since 1990 and are, therefore, mostly panel regressions, in contrast to the early cross-country regressions. The difference in degrees of freedom is large; cross-section studies for this group of countries use average growth for several years and imply about 25 observations, often less, while analysis done in the late 1990s with 8-9 years of data and pooling can have as many as 200-225 observa tions. While one is tempted to heavily discount early and cross-section-only studies using OLS with nary room for a lag, a pleasant surprise pops up from this survey. The broad conclusions of the simpler studies-to wit, the standard fa ctor inputs are not important; stabilization is necessary: liberalization and structural refo rms strongly affect growth peiformance, and unfavorable initial conditions can hinder growth-are not dramatically altered in those studies with a large observation set ru1d much more sophisticated econometrics. The broad robustness of these conclusions to coverage and methodology (see the subsection below on methodological difficulties) may be a testament to the overwhelming importance of these determinant variables, or the potential for simple methodology wisely used to yield powerful results, or both. Nevertheless, it is only the more sophisti cated econometrics pennitted by a large sample that can isolate the finer points of interpretations-for example. how unfavorable initial conditions result in sharper output declines early on but become less important in determining growth over time; or the fact that discipline of reforms yields inlmediate gains in a newly private sector even if there is continued total decline in GOP; or the increasing importance of too slow institutional development. We now review both the key findings of this literature and some of these finer points.

Key Findings

Stabilization

The first and largely noncontroversial conclusion is that stabilization is a necessary condition for recovery of output. Virtually all of the studies included some measure of stabilization-fiscal deficit, inflation, and, in a handful of cases,

JOSome of the studies-Fischer, Sahay, and Ve gh ( 1996. 1998). and Berg and others ( 1999)-also include Mongolia. I IA first version of this was presented at the First Dubrovnik Conference on Tran sition, June 1995, and eventually published in the 1997 conference volume.

©International Monetary Fund. Not for Redistribution Ta ble 2. Summary of Empirical St udies Explaining Growth In Transition

Variables Used, By Type Method and ResuliS ;:u m .o\ulhor.; Initial Liberalizatioo or S1abilizationJ factOrs Instirutional Years Methodological Key findings oo growth 0 ('lear) conditions reform inflation (invesu:nenL ''llri.lbles peculiarities � m (lC) expons. ;:u labor) -< )> z DeMelo, • inoome proxy for • World Bank • inflation 1989-94 • separateequation • over-industrialimtion 0 Denizer. and 0\·er index cross-section for growth. moderately significant (j) Gelb (l997) industrializatioo inflatioo • libera lization promotes ;:u • \\-'31 • sw1tching growth 0 regJeSSions • stabilization isnee� � I

• reform Saclb(1996} EBRD 1989-95 • 0.025 • reforms ve!')significant z index cross--.eot.ioo ---j ;:u )>

inflation 1990--95 • z SelowsJ.) and • (differentiation • World Bank. • •lags liberalization bigbly (/) Martin ( 19971 CE FSU) index panel significant =i • "·ar • contemporaneous effec l 0 of reformnegative in FSU z )>

As.lund, • FSU dummy • World Bank • inflation 1989-95 •. • 1989-95 growth: FSU and 0 025 m Boone, and • \\a£ inde:{ cross-section •GDP growlh wardummies dominate 0 Johnson (96) adjusted 1o elCplanation )> 0 caprureunder· • 1995 grov.th: Stabilization m reporting of andreform most signifrcam 0 .., privateoutput • \V0£SeDS reform decline but m gi \'e:;earlier reco\'el)' < 0 m Fischer. • initialincome • World Bank • fiscal 1992-94 • faedeffeciS • oounll)• effeciS significant z Sabay. and • b"3de disruption index ddicit p:mel • fiscal deficitsignificant only 0 m vegh CI996J 1992 • e.xchange b} itself regime dummy • all othervariables also o- significant

©International Monetary Fund. Not for Redistribution o- IV Table 2. (contnued)i

Variables Used. By Type lethod andResulrs

ADI.hors lnilial Liberalization or Slabilizat:Kml Factors Institutional Years Melhodological Key findings on gT0\\1h (Year) conditions reform inflation (im·esiJDelll, •-ariables peculiarities (lq exports. labor)

Hernandez- • time lrend • World Bllllk • inflation 1990--95 • state/private • inflationand liberalization Cat! (1997) • regional dummies indelt panel sector model to most significant •war derive • war and wealth signilicanL • time under specilicanon but olher initial condilions communism less • national rescwce 0 .... 'Calth iD =r I Wo lf (l997} • industrial share • World Bank • inflation • inv. share 1991-95 • 2-stage LS, • stabilization necessary 0 < • distanceto market indelt • fiscal • expon panel iDSIIUment.ed on -< • lenglhof central • radical deficit growth sector enrollment.• degree of reforms -

DeMelo, • initial conditions •World Bank • political 1990--96 • principal • initial conditions most Denizer. cluster. macro index freedom panel components to imponant but m:gatiYe Gelb. and distonions andover- dem-ec�of influence decline..over time Tenev(1997} industrialization initial conditions • (Lagged)reforms still • ..re gional tensions� ' J-eqll3tiOO most imponantin growlh model: I iberalization. inflation. growth • lags and interaction

©International Monetary Fund. Not for Redistribution Brunetti, • 1992 income • trade share • inflation • Wo rld Bank 1993-95 • institutional • institutional variable Kisunko, and • 1992 secondary • government sun-ey data: cross- variable generally significant Weder ( 1997) enroUment COn!>lliilptiOn • predietabilit) section a\"el"ageof se••eral • almost oon.e of polic). m;o share • political stability nTedibilit}' initial conditions 0 - property rights of rules-, variablessignificant � - JUdiciary reliability • .erial testof • problemof insufficient data m ;o • COITUpOOD inslimtional with -< COOIJ'O) )> • ''politicalnghts z as i nstroment 0

• \V:IJ' • • • • G) Loungani and World Bank fiscal deficit 1991-94 inflation equation lhedeffects negati\'efor ;o SheetsC 1997) • yeardummies in dex • inflation panel includes central 1991-92 0 bank • fiscal negati\c and significant � independence • inflation afferu grov.'th and I • thed effect£ depends on central bank z independence -f ;o )> Fisc her. Sa.hay, • 1992 dummy • World Bank • fiscal 1992-95 • fixed effect..� • fixed effects highly z en andWgh indices: • exchange panel significant =i (l998) internal, regime • ti ght fiscal.exchange fixed 0 external. private rate. refo� allsignificant z envi ronment • internalliberalization not )> significant m0 0 Ha�-rylyshyn, • de Melo. Denizer • deMelo. Denizer• infiaJ:i.on • investment • EBRD legal 1990-97 • separate decline • ti.xed effecb significant )> 0 Izvorsld. and Gelb. and Tener Gclb. and Tener share panel (1990-93)and • initial conditions m van Rooden { 1997) clusters (1977) to 1994 • e.�port recovery 11994- importantearly 0 • in come • most T1 (1998) in itial EBRD after gnw.'th 97) stabilization. reforms m • degree of • components: • fixed efferu and imponant < industrializationand price liberalization, • liberalization effectat lags first. ma de"-iation from ease of entry, strong!) and later z • legal significant average. external. legal reforms m0 • shareof government

©International Monetary Fund. Not for Redistribution Ta ble 2. (continued)

Variables Used. B� Type Method and Results

Authors Initial Liberalization or Stabilization/ Fac!OrS Institutional Years MelhodoJogical Keyfindings on growlh (Year) conditions reform inflation (imoesiillenL, variables peculiarities (I C) exports, lnborl

Olristoffersen • war • World Bank • inflation •export 1990-97 • pro11y for e.�pons • inflation, reform significant and Doyle • dummies for 1992 index markets panel = giO"-W in • iofl31ion-output threshold (1998) growth tnal'kets I 3-15percent for the emire • sean::h for sample, butrna} be lower for inflation later years lhreshoJd • no e\ideru:edisinflati lhat on caused output losses{with one exception) • expon marketssignificant

Be!E • urbanization • World Bank · fi�aJ 1990-96 • fuedeffec ts.. • initial conditions significant Borens:ztein. • repressed inflation indexupdated deficit panel lags, ioleJ'l!Ctive (but h:n-e minor effect in Sah.ay. and • natural resources b) EBRD� • inflation • differential effect expLaining cross-sect.ioool Zeuelrneyer • m-erintate of pre \"3riable must bepresent transition reform • greater speedof reforms. more growth

©International Monetary Fund. Not for Redistribution Camj)(ho • initial income • government • investment share 1990-97 • Sarroand Levine- • lowR2 valuese\en in panel (1999) • CIS dummy share • schooling cross Renelt • im•estmenL schooling not • population gro'A'1b section, specifications significant. sometimes ;v m panel • initial income, OSmost 0 �ignifican�;bill income DO{ 2 incl. �gnificant ifClS is m ;v -< HaHy1ysh�n • MOOT cluste�: • Wor ld Bank • inflation ncipal 1990-98 •lags • >tabili7..ation. liberalization • pri )> and van • overincl�ve index updated componen�:>: panel • tim�: dependence significant z Rooden dJst.ortions by EBRD Oegal. political of v.uiables • institutions significant but 0 (2000) O\-enll Jusing • inslimtional not pre..eminem (j) ;v sun-ey by \'3riab1e • importance of iiDtial 0 EBRD.Wo rld available forlii()£e conditions declinesO\'elr time � Bank. Heritage.. thanone year • imponance of institutional I ovec Freedom House. increases time z Euromoney -f ;v Kaufman, • 300 indicators 1990-95 • insuumeming • accountabilit)'. instability. )> Kraay. and aggregated into cross- govecnance go,·emmem effectiveness.. z en 6 using Zoido- clusters section (178 regulatory burden. ::::::j Lob:uon ( 19991 unobsen'Cd countiries: rule law.of highly f>ignilicant 0 componems OECD. non- (graft not significant) z model OECD) )> 0 m BerkowiiZ • income • new enterpris.:� .t7 regions of • new enterprise • strong correlation De\\ 0 and De Joog • shadowprofits Russia. 93- insuumented by enterprises and growlh. )> • of defense therefore price 0 (2000) share 97 degree of liberalization, m deregulation pri,'atization. priHtization, 0 priceliberalization. positive effect "TI m Ia.� rate!>. • higher defense share bas < decentralization positire effect 6 of power) m z 0 m

©International Monetary Fund. Not for Redistribution o- o- Table 2. (continued)

Variables Used. By Type Method andResuhs

Authors Initial Liberalizaliooor S�abili:z.ationf Factors lnstitotional Years Metbodologi<:al Key findings on growth (Year) conditions reform inflation (in\·estment. ...mable> peculiarities (I C) e.-.: pons, labor)

Campos (2000J • ci\·il MJCiety 1990-97 panel • interaction • all botacconmability • quality of among \'llriables signillcant bureaucracy • rime dummy • in Central and Eastern Europe. • nile of law • sample split into ROL, CIVIL dominate • accountability CEE andOS • in CIS, Ro2. qual bur. and lr.UlSp.arency dominate 0 • for percap ita.almoSt no dummies • bankcomp. variable, including privatization noeffect

li) hard budget degreeand • prl'-'31.zationi plus constraints; agency climate in stitution.allcompetitive (ii) bankruptcy eO\ironment \'ef)' significant courts: • need minimum threshold of (iii) EBRD legaL c,ompetition,agency effectiveness to get pri\-atization benefiiS

Moers (00) • initial income •World Bank • inHation • secondary • Wortd Bank. 1990-95 • serial testingof • stabilization andli beralization • industrial �bare in de)( schooling CEER.EBRD. cross- institutional significant alone; •CMEAuade • EBRD index • tertiary Euromoney. etc. settioo variables • institutions significant alone •War • pm•ate share schooling survey sources withother bul DotYoith aboYe CODirOI -rule of law CODirOI • variables correlation of -im"CStmenl Jaw variables liberalization and institutions -propeny righiS make impossible separating -

©International Monetary Fund. Not for Redistribution Oleh Havrylyshyn

exchange rate anchor-and the results are almost always significant: that stabi lization contributes to growth, or at least that growth does not occur until stabi lization is achieved. A majority of the studies use only an inflation variable to reflect stabilization, positing (and finding) a negative sign in growth regressions. The few that include both inflation and a fiscal deficit (Wolf, 1997; Loungani and Sheets, 1997; Berg and others, 1999) show varying results. Berg and others tind that "fiscal balance survived [exclusion tests] in most models," (p. 50) but the contemporaneous effect is puzzling as it implies that "tight fiscal policies sustain production in the state sectors but negatively impact the private sector." Loungani and Sheets concluded that "fiscal deficits have tended to be stimulating" (p. I 0), while Wo lf reaches the contrary conclusion, as do Fischer, Sahay, and Vegh ( J 996 and 1998b)-tbough the latter find the fiscal variable loses statistical significance in presence of other explanatory variables, in particular liberalization. From the above, one concludes that the empirical literature is nearly unanimous on the nega tive impact of inflation on growth bur has nor been able ro disentangle the sepa f rate efects of fiscal deficits and inflation on growth. This may be explained by either (or both) of two difficulties. First, the regressions are not in most cases derived from a structural model of growth in transition and they exclude a simul taneous determination of inflation and growth, though some do estimate separate equations for inflation, in which fiscal deficits generally show up significantly and, as expected, positively conelated (de Melo, Denizer, and Gelb, 1997; Fischer, and Gelb, Sabay, and Vegh (1998b); Aslund, Boone, and Johnson, 1996). Second, fiscal deficits may be too nanowly measured, and underestimate the "public sector deficit," which includes off-budget transfers, tax easing, central bank credits. and/or accommodation of directed credits by banks plus-most important for CIS countries-arrears problems. Only a handful of studies consider the role of exchange rate anchors explic itly. Fischer, Sahay, and Vegh (1998b), in particular, test for fixed exchange rate regimes with a dummy variable and conclude that "a fixed exchange rate and smaller fiscal deficits seem especially important in reducing inflation and raising growth rates." It is important to note that, in their model, the fixed-rate variable is negative and significant in the inflation regression, and positive and significant in the growth regressions that do not include inflation as a separate variable but in which the fiscal variable loses significance in presence of other explanatory vari ables. Thus, the impact of the exchange rate anchors on growth could be indirect only, working through the inflation stabilization effect. Recall that most other growth regressions have inflation as an argument in the growth equation. Thus the specification, not being in structural form, may not fully capture the mechanisms of effects on growth. Another missing variable here may be arrears, as noted; if a peg is intended to enforce fiscal discipline but this is achieved by running arrears, the coefficient estimate for a peg without an arrears correction will be at best weakened. Nevertheless, other studies, while somewhat more skeptical of the clear conclusion about fixed rates, do not fundamentalJy contradict it. CottareUi and Doyle (1999) explore the possibility of disinflation causing output losses and find that such difficu1ties only occur in presence of exchange rate pegs, but even then, "the losses seem likely to be due to undervalued pegs, rather than pegs per se."

©International Monetary Fund. Not for Redistribution RECOVERY AND GROWTH IN TRANSITION: A DECADE OF EVIDENCE

Hernandez-Cam ( 1997) also fi nds ·•results provide some-but not much-support for the view that fixed nominal exchange rates have contributed (to output growth but perhaps indirectly) by helping to bring down inflation."

Liberalization and Structural Reforms

The second set of fmdings concern liberalizing reforms; virtuaUy all the studies that include a reform variable have affirmed the de Melo, Denizer, and Gelb result on the strong and positive role of liberalization and structural reforms. Whether the model was a simple one relating only to growth and some index of structural reforms (Sachs, 1996; Selowsky and Martin, 1997), or a more complex one reflecting also the effects of stabilization, initial conditions, and conflicts (as in Aslund, Boone, and Johnson, 1996; Fischer, Sahay, and Vegh, 1996; Hernandez-Cata, 1997; de Melo, Denizer, and Gelb, 1997; Havrylyshyn, lzvorski, and van Rooden, 1998; and Berg and others, 1999), the conclusion was frrm: more reforms are associated with better growth performance. The results are not without clear country exceptions, Belarus and Uzbekjstan being the key ones today, Bulgaria and Romanja earlier. Belan1s remruns an understudkd case, but three points suggest the numbers overstate performance or its sustainability. Much of its growth was based on exports to Russia bartered for energy and other products; its largely unrefonned system meant a high degree of directed credits to the economy and a much higher inflation than its neighbors; and after the Russia crisis, its exports and growth feU sharply. 12 Zettel meyer (1998) shows for Uzbekjstan that structural and macroeconomic policies alone cannot explain the better-than average performance and finds initial conditions, in particular the low degree of industrialization and cotton export potential, helped cushion the decline and perhaps promote an early recovery. Fischer and Sahay (2000) point to the rurected credits and vulnerabiUty of Russia-oriented exports. Bulgaria and Romania, despite Umited structural reforms and stabilizatjon, saw an early growth recovery, probably fueled by directed credits, but the unsustrunability of such growth was made clear by the collapse in 1996-97. Related to these exceptions, Aslund, Boone, and Johnson (1996) observe an interesting dichotomy in the literature concerning the pace of reforms: while theoretical work on transition has often shown a gradual pace mjght lead to less early decline of output (e.g., Aghion and Blanchard, 1993, Dewatripont and Roland, 1995), empirical studies generally conclude that fast and early reforms result in early and sustained recovery. The strong findjngs on stabilization and structural reforms would appear to vindicate, at least broadly, the major elements of the Washington Consensus (which as footnote 4 noted does incorporate institutional development notwith standing statements of critics). An important policy questjon is whether some of the components are more important than others (e.g., privatization) or whether, as Aziz and Wescott (1997) conclude for all developing countries, it is the package that counts. The results for transition countries tend to confirm the view that all

12Furthermore. as IMF (2002) reports. a methodological problem with Lhe industrial production index results in an overestimate of 1-2 percentage points.

©International Monetary Fund. Not for Redistribution Oleh Havrylyshyn policies matter to some extent, as does institutional development, though some components are found to be separately important. Furthem1ore, when the reform variable is a single aggregate index, its coefficient is invariably significant; when components of liberalization (price, internal, external liberalization, privatization or private sector share) are considered separately, the results are more mixed. Thus Fischer, Sahay, and Vegh (1998b) find that aggregate liberalization is highly significant, but separately external liberalization or private sector environment (including banking) are less so, while internal liberalization (prices, and competi tion environment) is not significant, though still positive. Havrylyshyn, Izvorski, and van Rooden (1998) use these three components plus a legal reforms index of the European Bank for Reconstruction and Development (EBRD), but argue further that the underlying structural relations differ in the period of decline (1990-93) and recovery (1994-97).13 They in fact find different effects of liberal izing reforms; internal/price liberalization has at first a negative effect-which is consistent with the Kornai and Blanchard views of the necessary early decline in the old sectors. But the common view that the early pain is quickJy rewarded with substantial gain is also confmned, the coefficient being clearly positive for the second period. The "no pain, no gain" view finds strong support in many studies, in partic ular, Aslund, Boone, and Johnson (1996); Wolf (1997); Hernandez-Cata (1997); and Heybey and Murrell (1999). In contrast, Berg and others (1999) make an important contribution on this issue by using more sophisticated lag structures and separating out public and private sector. They conclude that "effects of structural reforms are mostly positive from the beginning." Their separation of public/private may help clarify that, while the larger old sector continues to decline in response to reforms, this decline is not large and in some cases is ambiguous. For example, components of reform such as external liberalization may have an immediate posi tive impact on state sector growth. On the other hand, the smaller new sector reacts to hard-budget signals of reform immediately and unambiguously. Thus, in early years, there may be negative growth, but Berg and others interpret this not as the "pain of reforms," but as costs of incomplete reforms. This is a useful refi nement, which is not fundamentally at odds with the rest of the literature. Finally, on the separate effects of components, it is notable in Havrylyshyn, Izvorski, and van Rood en ( 1998) that for the second, or recovery period, every one of four elements is separately significant, but when the aggregate index plus one of the elements is included serially to test for the possibility of combined effects, only internal liber alization remains significant-both in the decline period when it is negative and in the recovery period when it is positive. For all these separate elements, the coeffi cient is larger in the recovery period, suggesting the cumulating payoffs to growth over time. Tbis may help explain the strong CIS recoveries of 1999-200 1 despite the plateauing of reform efforts there since 1997. Consider next privatization, which of course has been very controversial in itself, and bighly criticized by many, perhaps most vocally in Stiglitz (1999). The

13Havrylyshyn and van Rooden (1998) update data to 1998 and confirm the Havrylyshyn, (zvorski, and van Rooden ( 1998) results for the recovery period.

©International Monetary Fund. Not for Redistribution RECOVERY AND GROWTH IN TRANSITION: A DECADE OF EVIDENCE

vast empirical literature on effects of privatization in transition has been recently reviewed in Havrylyshyn and McGettigan (2000), as well as Dj ankov and Murrell (2000). While, not surprisingly, there remain substantial arguments about whether privatization could have been done better, avoiding the worst of insider privatization inequity and subsequent stagnation, the evidence of these two surveys suggests that in many-if not all-cases, improved performance and increased efficiency ensued. The effect was stronger when outside control could be more effective, as with foreign investors, or in de novo enterprises whose resources frequently came from idle capital and labor of old enterprises. This suggests privatization has indeed led to reallocation and X-efficiency gains. But the evidence is far from overwhelming and leaves plenty of room for the view that all is not well with privatization.I4 In particular, the evidence is limited on two important dimensions: Havrylyshyn and McGettigan (2000) conclude in their survey that research has paid too much attention to the method of privatization and not enough to the "institutions of an approp1iate enabling market environment," and bas also largely ignored the contribution of de novo enterprises. A glance at the third column in Table 2 (Liberalization or Refonn) does indeed show that use of privatization as an explanatory variable in growth does not gener ally consider the market environment, or method, or de novo enterprises. True, one of the reform indices used by several studies covers broadly the element "private sector environment," but it is a general category in tl1e EBRD indices comprising several subcomponents including progress in restructuring of state firms. For what it is worth, the results are usually statistically significant. The separation of public/private may have clarified that, while the larger old sector continues to decline, the smaller new sector reacts to hard budget signals of reform immedi ately. Thus, in early years, there may be negative growtll in total but, as noted, Berg and others interpret this not as pain of reforms, but as costs of incomplete reforms. By separating public and private sector. they show reforms have some initial negative effect on public sector, but are immediately offset by gains in private sector. Thus, tlle bigger the private sector share, tlle larger the positive impact on growth. One of the most recent studies on privatization (Zinnes, Eilat, and Sachs, 2001, in this Special Issue of IMF Staff Papers) goes much farther in trying to disentangle the effects of formal change of ownership and tlle market envil·onment facing firms. The market environment is measured by a composite (and subjective) index reflecting a true change of tl1e firms' objective function to profit maximiza tion, a true hard budget constraint, and an effective agency climate. The results in panel regressions for output levels (instead of growth rates) are compelling. Change of ownership alone has virtually no effect; combined with a strong envi ronment index (a multiplicative variable), the effect is positive and highly signifL cant, but there is a minimum threshold for the market environment index before one sees a positive impact of the change of ownership. Zinnes, Eilat, and Sachs may be overly rhetorical in claiming that the insignificance of change of owner-

14Nellis (2001)considers all these factors and the evidence, and presents some counterarguments to the revisionist critics like Stiglitz.

©International Monetary Fund. Not for Redistribution Oleh Havrylyshyn

sllip variable implies a "failing grade" for the Washington Consensus. First, the Washington Consensus clearly talks of the role of the state in providing legal, market institutions, and property rights.l5 Second, Zinnes, Eilat, and Sachs are not able to show a strong negative effect of privatization in a poor markel environ ment. Their study is, nevertheless, a powerful statement about the crucial impor tance of the institutional changes needed to achieve a competitive hard-budget environment and good corporate governance. This jargon boils down to saying, you need good institutions to ensure that the new Komai market incentives truly lead to the Blanchard optimal reallocation of resources. While it is often stated that new enterprise formation is critical, perhaps because of data difficulties onJy one substantia] econometric study relating growth and new enterprise formation has been done. Berkowitz and De Jong (2000), using data from 47 regions of Russia, obtain results strongly supportive of the conven tional view and affi rm the more qualitative assessment of Aslund, Boone, and Johnson ( 1 996) of the early growth surge in Poland.

Initial Conditions

A rhird set of conclusions relates to initial conditions (e.g., overindustrializa tion and price distortions in Soviet period) and other factors specific to countries such as wars; it is generally agreed these do have a country-specific effect, though different studies attribute a different magnitude of importance. De Melo, Denizer, and Gelb (1997) group many different initial conditions into two clusters with principal components analysis and find a substantial impact; while Aslund, Boone, and Johnson (1996) argue that the more inward-looking and generally overindus trialized economies in the former Soviet Union faced a bigger hurdle than did Central Europe. Fischer and Sahay (2000) conclude that the length of communism and distance from Western Europe plus other initial conditions "go far in accounting for performance." But other studies suggest that the effect declines over time, and with refi nement of variable measurement. Berg and others (1999), applying lags and exclusion tests, conclude that achievement of stabilization and progress in structural reforms (i.e., policies) explain most of the differences between the better growth performance of Central and Eastern Europe and the poorer performance of the CIS countries. Initial conditions do have a limited impact, but this is strongly adverse only in early years, diminishing quickly over time (with a half-life of five years). Havrylyshyn, lzvorski, and van Rooden ( 1998) also find that initial condition variables are significant-even when crudely measured by share of industry-but they go on to show that these effects not only decline over time, but they also are not immutable effects and could have been offset with a greater effo rt at refonns. Thus, for example, in Ukraine, the higher industry share compared to the CIS average did contribute to a lower growth rate,

I5Jt often takes fresh blood and someone uninvolved in heated academic debates to cut to the heart of lhe mauer: Moers (1999), in a preliminary Ph.D. paper, concludes: "Thus it seems not so much the case that policies of the "Washington Consensus" are wrong. but rather they are incomplete. or at least not balanced enough·' (p. 37).

©International Monetary Fund. Not for Redistribution RECOVERY AND GROWTH IN TRANSITION: A DECADE OF EVIDENCE

but that difference in growth could have been made up by achieving a level of the reform index only slightly higher (on a scale of0-1.0, a rise from 0.57 to 0.65) about equal to Kazakhstan's. This result is disputable, of course, being based on a linear impact of refonns; if there are positive nonlinear effects, or threshold levels, it would mean that a larger leap forward on reforms may be required to offset initial conditions, but, at the same time, the eventual impact on growth would then be greater. Rather than estimate the impact of initial conditions on growth, another way of looking at this issue is to hypothesize that initial conditions play an indirect role through the politics of determining how committed a government is to stabilize and undertake structural reform. Fischer and Sahay (2000) recognize this but do not model it explicitly. Wolf (1997) takes this approach in a clear, simultaneous two-equation model, and de Melo and others ( 1997) also address this, with a sepa rate regression explaining the degree of liberalization achieved. Both studies conclude that reforms are affected by initial conditions, but more analysis needs to be done to detenni11e if any additional separate effect on growth exists.

Institutions

The fo urth set of conclusions concerns institutions: put simply, they matter too and increasingly so over time. Despite the fact that both proponents of the Washington Consensus and of the institutionalist school discussed the need for institutional development in the early 1990s (Fischer and Gelb, 1991; Murrell, 1 992), the fir st group of growth studies did not attempt to include institutional variables-perhaps because of the lack of data. Since 1997, the reverse has been true, with most studies incorporating some such measure, usualJy based on surveys of policy environment (Freedom House, Heritage Foundation, Transparency International, Wa ll Street Journal, World Bank), and most finding strong evidence that these variables are significant and generally of the expected sign. One interesting peculiarity is that some studies that purport to "correct" the omission of institutions in studies explaining growth either omit the other explana tory factors altogether (Kaufmann, Kraay, and Zoido-Lobaton, 1999; Campos, 2000), or find poor results for the oilier policy variables (Brunetti, Kisunko, and Weder, 1997; Moers, 1999). Moers clearly admits tlUs is odd and explains that it is attributable to an econometric problem, given the small cross-country samples and high correlation of liberalization indices and institutional development. Despite tllis, he still concludes that macroeconomic stabilization is important. But Havrylyshyn and van Rooden (2000) using time series for institutional variables, undertake panel regressions for much larger observations sets and, with a principal components analysis, form three clusters of institutions: legal, political, and overall. They are able to obtain regression results that retain the significance of stabilization, liberalization, and initial conditions variables, and in addition show that institutional variables are also separately significant. Using a time multiplica tive transformation, they further demonstrate that, while initial conditions become less important over time, institutions become increasingly important. That is, early growth can be stimulated by stabilization and liberalization even if institutional

©International Monetary Fund. Not for Redistribution Oleh Havrylyshyn

development lags, but, as time goes by, sustaining growth requires continuous institutional improvement; this is consistent with the result on privatization of Zinnes, Eilat, and Sachs (2001).

Factor Inputs

The fift h and final set of conclusions demonstrates the unimportance of tradi tional factor input explanations for growth, differentiating transition economies from others covered in the large recent literature on growth (a good overview is in Barro and Sala-i-Martin, 1995). Wolf (1997); Havrylyshyn, Izvorski, and van Rooden (1998); Campos (2000); and De Broeck and Koen (2000a) all conclude that investment has not been a significant determinant of growth-the last study being based on factor decomposition and total factor productivity calculations. This should not be so surprising given the short period of time since recovery began in transition countries and, as Havrylyshyn and others ( 1999) note; "for over-industri alized, distorted, and inefficient transition econonries, recovery only comes after some elimination of the wasteful old production ... and usually cannot be based on a large investment effort to build the new before the proper incentives for efficient resource use are in place" (p. 17). The conceptual framework noted in Section I and in Figure Al in the Appendix strongly suggests a sequence of various efficiency improvements, with traditional factor input expansion coming only later. Indee-d, the best guess one can make about why most of the studies did not even include variables representing factor inputs (and guess one must, as none explain this) is that they considered-quite sensibly it turns out-that these were not important explanatory variables. The bottom line here is that growth in transition-in a so far rather short period of less than a decade-is more a special short-run phenomenon than the conventional long run considered in the genera1 literature. Thus, it is not surprising that expansion of factor inputs have not yet been important. Further suggestive evidence comes from trends in the investment to GOP ratio discussed in Havrylysbyn and others (1999). Of the 17 countries with sustained growth to date and adequate data on investment, the most common pattern for the investment of GOP ratio is a decline from the central plan period levels of 30 percent and more to near 20 percent or even lower. Further, for the 17 countries, an upturn in the ratio of investment to output preceded the recovery in only 3 cases, while it coincided with the beginning of recovery in 5 countries and actu ally lagged the upturn in output in 9. It is therefore not surprising that the econo metric specification of recent empirical studies on growth during transition, while broadly analogous to those of the new growth theory, ignores the long-tenn factors such as investment, and focuses on efficiency-improving factors such as macro policies, structural refonns, and property rights climate.

Some Common Methodological Difficulties

As the studies listed in Table 2 address a common question (what explains growth in transition countries), it is not surprising that they have to contend with common analytical difficulties concerning measurement and choice of modeling

©International Monetary Fund. Not for Redistribution RECOVERY AND GROWTH IN TRANSITION: A DECADE OF EVIDENCE

structure. Here we explore some of the main methodological difficulties and differences and bow these may affect some of the results described above. The first measurement problem concerns valuation of output. With the excep tion of Loungani and Sheets (1997), all the studies content themselves with offi cial GDP values, notwithstanding the consensus view that unofficial activity may account for as much as 40-45 percent of total output in some countries. The reasons given include, in particular, the unavailability of consistent estimates of unofficial economy for the full sample of countries over an extended period. While this is understandable, it may still be troublesome if there is reason to believe that growth in the offtcial sector differs substantially from growth of the unofficial sector. Loungani and Sheets do a great service on this score by demonstrating for a smaller sample of countries that using an index of electricity output (perhaps the most common estimator of total output) does not change the resuJts of growth equations. Herm'indez-Cata ( 1997) includes a correction variable (ratio of measured output to power consumption); the results on the major policy variables are not materially affected, but the variable is significant and replaces the role of the FSU and war dummies-suggesting that the biggest underreporting is in those cases. Key writings on the unofficial economy (Johnson, Kaufmann, and Shleifer, 1997) fully confirm the latter point and emphasize that strong sustainable growth only comes in a policy environment that induces shadow activities to become open and official. At a minimum, this impues that in the recovery and positive GDP growth period, shadow activities become relatively less important. It does also imply, however, that in the decline period official output falls more rapidly than total output, which further implies that estimated coefficients of explanatory vari ables such as initial conditions or liberilizing reforms overstate the negative impact on output in the early period. As already noted, Aslund (200 1) makes some very rough corrections for this as well as several Soviet-accounting problems, and claims that the correction strengthens the positive effect of reforms. This issue deserves more analysis. Looking more closely at unfavorable initial conditions that are found to have a negative effect and to decline over time, correction of the unofficial GDP bias is unlikely to do more than imply an earlier fading away of these negative effects. A more important modification of conclusions could come about on the hypothesis of "no pain, no gain"-that is, reforms inevitably cause early pain. Most studies, with the exception of Berg and others (1999), conclude that, while reforms have unquestionably a positive influence on growth in the long run, the short-run effect is negative, as closures or cutbacks by inefficient producers are faster than expan sion of new efficient production is elsewhere. It is possible that correcting official GDP data gives a lower output fall and/or higher positive growth; thus the nega tive effects may be less strong and a ceteris paribus estimate of the reforms coef ficient would yield a positive value. The puzzle remains, since GDP correction has not been done for a systematic large sample to allow its use in growth empirics. Berg and others came to their "no pain at all'' conclusion despite the fact that they use official GDP data. In the end, however, all analysts agree that reforms are good for growth, whether or not they cause some pain early on, so the logical puzzle may have limited impact on policy implications.

©International Monetary Fund. Not for Redistribution Oleh Havrylyshyn

Measurement difficulties also occur with fiscal deficits, as already noted, and this may be part of the reason why the literature cannot give clean results on how state-sector finances, broadly defined, affect growth. But the conclusion about inflation control is universal and one could not ask for stronger empirical support on the crucial role of stabilization. Indeed, it may not be necessary to have solid econometric estimates of how fu1ancial imprudence by governments-be it budget deficits, off-budget operations, tax relief, or soft credits-contributes to inflation. Once it is accepted that inflation harms growth, the policy argument for fiscal prudence, transparency, and credit tightness is sufficiently well grounded. Perhaps the measurement difficulty that 1eads to the most serious problem for policy conclusions is that relating to institutions. Not only the transition literature, but also the general literature on institutions, has great difficulty coming to grips with what exactly it is one means by institutions, and how they can be objectively measured. 16 Studies that do consider institutions are broadly in consensus that the ex post measurement problem is not overwhelming, and the subjective indices are sensible. But if institutions are but a state of being (subjectively) measurable by surveys, it is d.ifficult to be as concrete in policy recommendations as with mone tary and fiscal measures. This serious policy dilemma is beyond the scope of this paper. The modeling difficulties for transition countries are in principle substantial, but in practice may not comprise a serious impediment to drawing valid and useful conclusions. The general literature on growth itself includes ongoing debates about the right structural form to capture the spirit of the new endogenous growth theory (Barro and Sala-i-Martin, I 995) or the right methodology (cross-country, time-series, panel-see, for example, Quah, 1999). When one turns to transition economies and their recovery in the short mn, as opposed to the medium- and long-term equiUbrium path, the difficulties mount. Consider Figure A l; the point CPI represents the starting point before transition begins with Soviet level ineffi ciency in the allocation between goods and factors, plus X-inefficiencies for pure wastage. Any reasonable reading of what has to happen-and in retrospect generally did happen-suggests several events of output decline or expansion occur before a new market-detennined equilibrium is approached and before an equilibrium path of growth (arrow 5 in Figure A l) is embarked upon. First, output will fall to a lower level of capacity (CPU) compared to CPI; then it could of course rebound toward CPI (arrow I). The important gains of Russia and Ukraine in import substi tution since their August 1998 devaluation are of this kind, though clearly preceded by some minor restructuring. But any such rebound is probably hidden in a simultaneous move involving the many resource shifts that Blanchard (1997) theorizes about and which a vast literature on restructuring (reviewed in Djankov and Murrell, 2000) describes. These resource shifts include technical efficiency gains, or moves toward the efficiency frontier (arrow 2); perhaps illustrated by

16See Fukuyama (2000) for a general view of the institutional measurement problem. Kaufmann, Kraay. and Zoido-Lobaton ( 1999) is a good comprehensive source for the myriad efforts to use survey data that measure institutional development.

©International Monetary Fund. Not for Redistribution RECOVERY AND GROWTH IN TRANSITION: A DECADE OF EVIDENCE

improvements in efficiency of Polish state enterprises as documented by Belka and others (1995); gains f1·om achieving efficient factor proportions (arrow 3) the labor-shedding and resultant increases in labor productivity as shown in De Broeck and Koen (2000a and 2000b); and finally the gains from less production of unwanted goods and the shift of resou1·ces to production of other goods (anow 4)-the sectoral shifts noted in Berkowitz and De Jong (2000) and the product-line shifts as desctibed in EBRD (1997). All of this adds up to a complex, but eventu ally clear, picture reminiscent of Harberger's ( 1998) vision of the growth process, as a "thousand and one Little changes" at the firm level that happen daily and add up to measured growth in GDP. Modeling this for econometric analysis seems a daunting task and it is there fore no surprise that virtually all the growth regressions reviewed here take an ad hoc (but we wouJd say sensible) approach to specification, with only limited efforts to derive a structural form from first principles. There are two exceptions. Hemandez-Cata (1997) starts with a standard production fu nction with a state and private sector, the latter with higher productivity. Liberalization increases share of the private sector, but time lags in resource allocation result in underutilization of capacity. As inflation and certain state variables (initial conditions) can affect both the liberalization and the pick-up of capacity, the model is readily modified to a form identical to other studies with no factor inputs, but instead policy variables Jjke reforms, inflation/stabilization, exchange regime, and initial conditions vari ables or dummies. Berg and others (1999) begin with a three-equation model in which the growth equation already omits factor inputs on the right-hand side while the two other equations reflect policy determination and structural changes since transition, to capture the possible changes in impact of policies or initial condi tions on growth as transition progresses. In the econometrics-which is unques tionably the most detailed of all the studies--only the first equation is estimated, the other two providing a rationale for using certain instrumental variables and for a methodology of "general to specific" estimation, which allows the large panel data set to choose the proper specification. Table 2 includes a summary of what variables they use in the end to derive conclusions. All other studies, if less formal, tend to specify quite similar regression equa tions, and have more or fewer variables depending on whether they are strict cross sections or panels. As Table 2 and the preceding discussion conclude, these different studies reach more or less similar broad conclusions on the importance of various factors ru1d how the effects of initial conditions fade over time. However, more complex questions-such as the importance and timing of partie war elements of reform, ideal policy combinations. and differential impacts of structural reforms and stabilization over time (the "no pain, no gain" debate)-are onJy addressed by studies with panel data, which have more possibility for econo metric fine-tuning. But the greater methodological sophistication does not resolve all debates. For example. consider the two studies that are based on more formal derivation from structural equations. Hermmdez-Cata strongly concludes that reforms initially have a negative effect on output as state sector losses are not immediately offset by private sector expansion. Berg and others find evidence of a negative effect on the state sector, but the associated losses are more than offset

©International Monetary Fund. Not for Redistribution Oleh Havrylyshyn

by contemporaneous private sector expansion. Similarly, Hernandez-Cata finds only weak support for fixed exchange rates, while Berg and others show strong evidence that fixed exchange rates are important. It is not the intent here to resolve these differences one way or another; since the results use different models, periods, and econometric techniques, comparing them is difficult. The use of a more formally derived model and more thorough econometric analysis helps illus trate subtler points, but does not change the broad conclusions of what matters for growth, and resolves some but not all the debated issues.

IV. Conclusions: Considerable Consensus, Some Debates, and Some Puzzles

What is most striking about this survey of empirical studies analyzing deter minants of growth in transition is the wide scope of consensus compared with the limited range of continued debate or puzzles. It is agreed that no royal road to growth exists but that a wide range of good policies is needed, including financial stabilization, market liberalization. and market-friendly government rules and institutions. One should not misunderstand this, however, since the wide agree ment on what policies are needed to achieve strong sustainable growth still leaves large differences in views on how to implement these policies, what sequence to follow, and what relative importance to give to each. Many of these large differ ences are explored elsewhere in this Special Issue of IMF Staff Papers. Here, the emphasis remains on the determinants, and this section summarizes the key points of consensus, some of the outstanding debates, and a handful of interesting puzzles.

Five Points of Consensus

The overwhelming area of consensus is that traditional fa ctor inputs have no role in explaining growth over time and across countries. The empirical evidence confinns the short-run nature of both the decline and the recovery, as well as the clear perception that Soviet-type economies were welJ inside their production frontiers and off their comparative advantage equilibrium. Investment ratios have no explanatory power or significance to speak of; their trend over time suggests almost a reverse causation-that is, they begin to rise only after growth has begun to recover. Efficiency gains appear to be the main, if not sole, source of growth. It is, of course, implausible that efficiency improvements come without new invest ment, but this paradox is easily resolved. The econometric analysis finds no signif icance of aggregate investment in the output recovery, which in no way denies the contribution of individual investment projects in spurring growth in sectors and firms. Nor does the result deny that aggregate investment will become increasingly critical in sustaining growth as the recovery proceeds and the new equilibrium growth path of a market economy is reached. There is Little or no argument that a decline in output was inevitable, and that it had to be about as large as the 20-25 percent fall seen in most of Central Europe;

©International Monetary Fund. Not for Redistribution RECOVERY AND GROWTH IN TRANSITION: A DECADE OF EVIDENCE

whether the stronger decline farther east was also inevitable or could have been mitigated by better policy is discussed below under "Debates." There is equally strong consensus that financial stabilization, in partkular inflation control, is a necessaryfirst step before sustainable growth can occur. But it is also agreed that stabilization is not sufficient, and thus market liberalization or structural refonn (correction of various price and qualitative distortions of central planning) are identified in many econometric studies as the most important fa ctor-though the statistical measure of "relative impmtance" varies. It is diffi cult to conceptualize the precise meaning of attributing "most" importance to one determinant, given the solid evidence that other factors-like initial conditions, stabilization, and some minimum of institutions-are also necessary. A fair state ment might be that, given some minimum of these other factors, the differences in growth performance are attributable in particular to differences in the progress on market reforms. Related to this is the conclusion on the role of privatization and expansion of the private sector. Empirical evidence suggests it was on balance beneficial to growth, though some debate continues as noted below. There is slightly Less consensus on initial conditions, though no one disagrees that they have some, at least initial, at least minor impact. The evidence is clear that the negative burden of Soviet inheritance (excess industrialization, 17 energy inefficiency, inertia of price distortions, geographical and historical distance from the market, civil wars) does have a hindering effect on growth, albeit not clearly separate from its indirect effect as a brake on speedy refom1 implementation. The econometrics also reveals that this effect is Jess important than stabilization and structural reform, declines over time, and can be offset by stronger progress in transition. On institutional development, there is a somewhat surreal consensus, with all agreeing that good institutions are one of the necessQ/y components in the recipe. For what it is worth, studies by proponents of the Washington Consensus and its critics all obtain the same strong, econometric results ex post: institutional indices have significant and positive coefficients. But the debate is not on whether insti tutional development-like patriotism is a good thing, but on what it comprises and whether it should be added to the transformation recipe befo re, during, or after the stabilization and liberalization.

Five Continuing Debates

The least well resolved-and arguably most important-continuing debate concerns the timing and sequencing of institutional reforms. From a historical perspective, it is difficult to assess empirically a relevant counterfactual: had there been greater and/or earlier efforts on developing effective judicial, rule-of-law, conditions, easier entry, stronger exit rules, and competition, would there have

171t is fair to ask why the very high human capital of the military-industrial complex should not have been ajavorab/e initial condition. One interpretation would be that it was, but poor policy implementa tion precluded quick and effective utilization of this comparative advantage endowment. We are not aware of any study that looks at this.

©International Monetary Fund. Not for Redistribution Oleh Hovrylyshyn been even better performance? The only, very weak, suggestive econometric evidence is lhe result in Havrylyshyn and van Rooden (2000) that the impact of institutions increases over time. In other words, good economic policies (macro and micro) may be enough to get growth started even with poor institutions, but as time goes on, institutional development must proceed apace or growth wiU falter. TIUs is, however, only suggestive evidence, and much more empirical analysis would be needed to move to consensus even on the rustory. For a futw·e policy perspective, the lesson that more weight should be given to institutions is clear, but two dilemmas remain. First, what exactly is to be done by policymakers and particularly by outside parties such as lhe IMF and the World Bank? Second, there is insufficient hard empirical evidence to refute the alternative view that institu tions must come first and only then should liberalization and privatization proceed. An equally difficult debate continues on the speed of reforms, or gradualism vs. Big Bang-and it is reflected in this Special Issue of JMF Staff Papers. Little explicit analysis has been done, with the exception of Heybey and Murrell ( 1999) on the one hand arguing speed does not help. and Berg and others ( 1999), on the other hand showing it does. A large part of the problem is definingspeed-and the two studies do it sUghtly differently. Another difficulty is whether an earlier start or faster speed is more important. Given the wide consensus that antireform vested interests (the most important being new capitalist rent-seekers) have captured the process in many countries, it is hard to deny a hypothesis that, say, Ukraine and Russia, might have needed an even bigger Big Bang in the rnid- l990s to get same effect as Poland did in 1989-90. In support of those who say speed is less impor tant than the accumulated, absolute level of reforms reached. there is suggestive evidence from the high rates of growth finally achieved in CIS countries in 1999-2000-and apparently continuing. It will be useful in a few years to ask how much this is attl'ibutable to cumulative benefits of reforms, exogenous effect of energy prices, a sharp rebound from the very low levels reached (recall Armenian and Georgian growth in the mid-1990s). or finally the long-term convergence effects described in the general growth literatw·e. Does "distance from the marker"matter or, more usefully, how does it matter? The empiLical evidence on such initial conditions as years under communism and geograprucal distance from Western Europe is strong, and seems to say Poland or the Czech Republic were inevitably better placed to perform better than Ukraine or the Kyrgyz Republic. The fact that these immutable conditions-or "bard conditions" as Zinnes, Eilat, and Sachs (2001) usefully categorize them-makes the issue all the more important. But such a view is troubling in several respects. First, it is not yet clear from the literature whether these conditions affect growth performance directly, or only indhectly through the political economy effect of slower if not actually frozen reforms. Second, even on the latter interpretation, there are many other considerations of perhaps greater importance, such as the anchor effect of EU accession prospects. Finally, there is a potentially malevolent misinterpretation of this distance hypothesis. In many of the slower reformers, the old and new interests that oppose reforms often find it very convenient to use such an argument to delay reforms. Observers of transition should be very cautious of

©International Monetary Fund. Not for Redistribution RECOVERY AND GROWTH IN TRANSITION: A DECADE OF EVIDENCE

statements such as "the populace has a Soviet mentality, they are not ready for the market": often these statements reveal that the speaker is not ready for a real, competitive market economy in which capitalism is an opportunity for all the people and not just a few. While, so far, privatization has been fo und to be generally beneficial, and it seems at least no worse than continued public ownership, there remains both a strong perception and some empirical evidence that "badly done " privatization can actually be harmfu l. The perception is particularly strong in many of the OS countries, where the inequity of privatization favoring a privileged few-a percep tion with considerable truth-may lead to even more damaging indirect effects as the populace opposes further market reforms. The empirical evidence of actually negative effects is Limited-as in the Zinnes. Eilat, and Sachs article-but too compelling to ignore altogether. Finally, let me turn to the debate on how much pain may be caused in the early part of the transition. As discussed earlier, most analysts are of the view that some early pain is inevitable, as stabilization and reform starts, but that the eventual benefits compensate for this and vindicate the aphorism of '"no pain, no gain." The analysis by Berg and others, by separating public and private sector, contradicts tl1e general views as it finds ilie benefits of reform are seen immediately in the private sector and, of course, ilie aggregate value of these benefits increase equally as the share of the private sector grows. The contradiction may be more semantic than substantive. Transition is not instantaneous; hence the good policies themselves and subsequent results take some time. In the meantime, the old (i.e., public) sector undergoes the pain of closure and cutbacks of employment probably faster than new expansion occurs in private. Thus, transition is inevitably painful. But it, nevertheless, remains the case. as Berg and otllers argue, iliat it is not reforms that give tl1e negatives (the pain), but the lack or slowness of reforms. As soon as reforms (including privatization) are made, their positive impact is immediate.

A Handful of Puzzles

Perhaps the first puzzle that future analysis might eventually answer is why the transfonnational recession was quite as deep as it was, even in Central Europe. As noted, there may still be a debate about why it was greater in the CIS and Baltic countries, but a lot of insight has been gained on this relative performance; the puzzle is about the absolute decline; was the 20 percent or so seen minimally in the Czech Republic, about what one should have expected in the Kornai transfor mational recession, about equal to the amount of "useless" or negative value added Soviet production? ln a sense, Berg and otl1ers (1999) do propose an answer, saying initial conditions explain not only the variation across the group. but tl1e actual decline as well. and that policy reforms do not have any further negative impact on output levels. Aslund (2001) also addresses this, and is strongly skeptical tl1at ilie decline was anytl1ing like official GDP shows. But the vast majority of analyses concur the decline was large and often say or imply it has been surplisingly large. This may not ever be adequately answered as it is difficult to envision the correct counter factual, never mind measure it.

©International Monetary Fund. Not for Redistribution Oleh Havrylyshyn

Apart from transitional factor inputs, two variables surprisingly did not show econometric significance: exports and fo reign direct investment. 18 Qualitative assessments of the more successful performers in Central Europe and the Baltic countries, in contrast, attribute a considerable role to both these variables. Foreign investments have been substantial, especially in the more advanced transition economies, as the article by Garibaldi, Mora, Sahay, and Zettelmeyer (2001) in this Special Issue of lMF Staff Papers shows. Perhaps the problem is an econo metric one, as these two variables may be strongly correlated to each other and to the "good policy" variables, thus hindering efficient estimation. A continuing puzzle remains on the low decline and high growth of Belarus and Uzbekistan. Certainly, the authorities of these countries and many critics of the standard reform school continue to point to these cases as successful gradual reforms. But even official figures show that growth there bas slowed a lot, while the rest of the CIS show an accelerating trend. To resolve this issue, it will require either a clear reversal as was seen earlier in Bulgaria and Romania, which also bad positive growth with limited reforms, or much more compelling explanation by one school of thought or the other. One quite unexplored hypothesis is that stronger central discipline has pennitted a recouping of some of the Joss from the 1960s Soviet-level efficiency that occurred in the Brezhnev and subsequent periods. But even if a convincing case is eventually made that retention of central plan discipline mitigates any negative effect, these two countries are likely to remain, in practice, the exception rather than the rule.

APPENDIX

Efficiency Improvements, Reallocation, and Investment as Proximate Sources of Growth in the Transition: A Conceptual Framework

The mechanism or proximate sources of growth in the recovery phase of tran sition is illustrated in Figure Al.

• CP* represents the point of potential production on the production possibil ities functions (PPF) under central planning if full capacity is utilized, factors are not wasted (as at Xl 1, at which more K and L are being used to produce 1 unit of good B than at the efficiency frontier), and shadow prices of K and L are reflected in the central plan's factor allocation. In this case, the only error under CP* is to ignore world prices. • CP! is the point where factor and technical inefficiency of central planning also exist, but capacity is fully utilized; historically this is roughly equivalent to pre-1990 production.

18Cotarelli and Doyle ( 1999) do show significant and positive coefficients for anexport proxy, growth of the markets for a given country's exports, but this is a rather indirect and, hence, insufficiently, compelling result. A few other studies that did include export growth found no significance.

©International Monetary Fund. Not for Redistribution RECOVERY AND GROWTH IN TRANSITION: A DECADE OF EVIDENCE

Figure A 1. Growth in the Recovery Phase of Transition

Good A K

Good B L

• CPU reflects a decline in production levels (as in post-1990), and lower utilization of existing capacity. • XJ1 is inside the efficiency frontier of the unit isoquant for good B; K and L are used wastefully relative to the theoretical best practice point for central planning, XI1 *. • M1 is a point on the PPF "inherited" from resource accumulation during the period of central planning. No aggregate net investment is necessary in moving to Mt. but there could be gross new investment in the expanding sector, so that allocation among goods can change, reflecting adjustment of production (allocation) to world prices. • M2 is an efficient goods allocation point with new net investment. Using Figme Al, five types of changes (structural shifts) can be defined that provide growth in the sense of more output for a given level of factor availability, and more factor inputs. 1. Recovery from capacity underutilization (from CPU to CPJ). 2. Technical efficiency or X-efficiency gains = movement to efficiency fron tier by eliminating wasteful usage of production factors (from CPI to CP*, and from XI1 to XI1*). 3. Efficiency gains from achieving optimal factor proportions (from XJ 1* to BCP*, and also from CPI to CP*). 4. Resource reallocation gains (from CP* to M1, that is, production of more B and Less A, hence, also BCP* to BM1• 5. Net factor-expansion gains (from M1 to M2, and from BM1 to BM2, plus an analogous shift to a higher level isoquant for good A).

©International Monetary Fund. Not for Redistribution Oleh Havrylyshyn

References

Aghion. PhWppe. and Olivier Jean Blanchard. 1993. "On the Speed of Transition in Central Europe:· EBRD Working Paper 6 (London: European Bank for Reconstruction and Development). Aslund, Anders, 200 I, "The Myth of Output Collapse after Communism·· (unpublished: Washington: Carnegie Endowment for International Peace).

__. Peter Boone, and Simon Johnson, 1996. "How to Stabilize: Lessons from Post Communist Countries," Brookings Papers 011 Eco11omic Activity: 1, Brookings Institution, pp. 217-313.

Aziz. Jahangir. and Robert F. Wescott. 1997, '·Policy Complementarities and the Washington Consensus;· rMF Working Paper 97/1 18 (Washington: Tntemational Monetary Fund). Balcerowicz. Leszek, 1994, "Economic Transition in Central and Eastern Europe: Comparisons and Lessons," Australia// Economic Review, Vol. I 05 (January-March), pp. 47-59. Barro, Robert J., and Xavier Sala-i-Martin. 1995, Economic Gr011'1h (New York: McGraw Hill). Belka, Marek, Saul Estrin, Mark Schaffer, and I. Singh, 1995. "Enterprise Adjustment in Poland: Evidence from a Survey of200 Private, Privatized and State-Owned Firms," CEPR Discussion Paper No. 233 (London: Centre for Economic Policy Research). Berg, Andrew. Eduardo Borensztein, Ratna Sahay. and Jeromin Zettelmeyer. 1999, "The Evolution of Output in the Transition Economies: Explaining the Differences," JMF Working Paper 99n3 (Washington: lnternational Monetary Fund).

Berkowitz, Daniel, and David De Jong, 2000, ·'Growth and New Enterprise Formation in Russia" (unpublished: Pittsburgh, Pennsylvania: University of Pittsburgh). Blanchard. Olivier Jean. 1997. The Economics of Post-Communist Transition (Oxford: Clarendon Press).

__, and Michael Kremer, 1997, "Disorganization;· Qutmerfy Journal of Economics, Vol. 112 (November), pp. 1091-1 125. Bloem. Adriaan M .. Paul Cotterell. and Terry Gigantes. 1996, ·'National Accounts in Transition Countries: Distortions and Biases," IMF Working Paper 96/130 Washington: International Monetary Fund). Bnmetti, Aymo, Gregory Kisunko, and Beatrice Weder, 1997, "lnstitutions in Transition: Reliability of Rules and Economic Performance in Former Socialist Countries;· World Bank Policy Research Working Paper 1809 (Washington: World Bank). Campos, Nauro, 1999, "Back to the Future: The Growth Prospects of Transition Economies Reconsidered," Cerge-El Working Paper Series (Czech Republic); No. 146 (June), pp. 1-37.

__, 2000. "Context Is Everything: Measuring Institutional Change in Transition Economies." World Bank Policy Research Working Paper 2269 (Washington: World Bank). Christofferson, Peter. and Peter Doyle. 1998, '·From Inflation to Growth: Eight Years of Transition." IMF Working Paper 98/99 (Washington: International Monetary Fund). Cottarelli. Carlo. and Peter Doyle. 1999. Disinflation i11 Transition 1993-1997. IMF Occasional Paper No. 179 (Washington: International Monetary Fund). De Broeck, Mark, and Vincent Koen, 2000a, ·Tbe Great Contractions in Russia. the Baltics. and the Other Countries of the Fonner Soviet Union: A View from the Supply Side;· lMF Working Paper 00/32 (Washington: lntemational Monetary Fund).

©International Monetary Fund. Not for Redistribution RECOVERY AND GROWTH IN TRANSITION: A DECADE OF EVIDENCE

___. 2000b. ·'The Soaring Eagle: Anatomy of the Polish Take-Off in the 1990s.'' IMF Working Paper 00/6 (Washington: international Monetary Fund). De Melo, Martha, Cevdet Denizer, and Alan Gelb, 1997, '"From Plan to Market: Patterns of Transition." Chapter I in Macroeconomic Stabilization in Transition Economies, ed. by Mario Blejer and Marko Skreb (Cambridge, Massachusetts: Cambridge University Press).

___, and Stoyan Tenev, 1997. "Circumstance and Choice: The Role of Initial Conditions and Policies in Transition Economies." World Bank Policy Research Working Paper 1866 (Washington: World Bank). Dewatripont, Mathias, and Gerard Roland, 1995, "The Design of Reform Packages Under Uncertainty.'' American Economic Review. Vo l. 85 (December). pp. 1207-23. Djankov, Simeon. and Peter Murrell. 2000, '"Enterprise Restructuring in Transition: A Quantitative Survey" (unpublished: College Park, Maryland: University of Maryland). Easterly, William, and Stanley Fischer, 1995. ''The Soviet Economic Decline: Historical and .. Republican Data, Wo rld Bank Economic Review, Vol. 9 (September). pp. 341-71.

European Bank for Reconstmction and Development. 1997, Transi1ion Repon 1997: Enterprise Peiformanc:e and Growth: Economic Tr ansition in Eastern Europe and the Fo rmer Soviet Union (London). Fischer, Stanley. and Alan Gelb, 1991. "The Process of Socialist Economic Transformation,'' Journal of Economic Perspectives, Vol. 5 (Fall). pp. 91-105. Fischer. Stanley, and Ratna Sahay, 2000, ·'The Transition Economies After Ten Years,'' IMF Working Paper 00/30 (Washington: International Monetary Fund). Fischer, Stanley, Ratna Sahay. and Carlos A. Vegh, 1996. ·'Stabilization and Growth in .. Transition Economies: The Early Experience. Journal of Economic Perspectives, Vol. lO (Spring), pp. 45-66.

__, 1998a, "How Far ls Eastern Europe from Brussels?'' IMF Working Paper 98/53 (Washington: International Monetary Fund).

__, 1998b. "From Transition to Market: Evidence and Growth Prospects," lMF Working Paper 98/52 (Washington: international Monetary Fund). Fukuyarna, Francis, 2000, "Social Capital and Civil Society,'' IMF Working Paper 00/74 (Washington: International Monetary Fund). Garibaldi. Pietro, Nada Mora, Ratna Sahay, and Jeromin Zettel meyer, 200 I. "What Moves Capital to Transition Economies?" IMF Staff Papers, Vol. 48 (Special Issue), pp. 109-145. Harberger, ArnoldC., 1998, "A Vision of the Growth Process;· American Economic Review, Vo l. 88 (March), pp. 1-32. Havrylyshyn, Oleh, lvailo Lzvorski, and Ron van Rooden, 1998, ·'Recovery and Growth in Transition Economies 1990-97: Stylized RegressionAnalysis," IMF Working Paper 98/141 (Washington: international Monetary Fund). Havry1yshyn, Oleh, and Ron van Rooden, 1999, ''Institutions Matter in Transition. but So Do Policies," lMF Working Paper 00/70 (Washington: International Monetary Fund). Havrylyshyn, Oleh, and Donal McGettigen, 2000, "Privatization in Transition Countries." Post Soviel Affairs. Vo l. 16, pp. 257-86.

Havrylyshyn. Oleh, and others, 1999. Growth Experience in Tr ansition Cowuries, 1990-98. IMF Occasional Paper No. 184 (Washington: International Monetary Fund). Hernandez-Cat<\, Ernesto, 1997, "Liberalization and the Behavior of Output During the Transition from Plan to Market." IMF Stajj'Papers. Vol. 44 (December). pp. 405-29. Heybey, Berta. and Peter MurreU, 1999. "The Relationship Between Economic Growth and the Speed of Liberalization During the Transition." Policy Refo rm, Vo l. 3 (June), pp. 121-37.

©International Monetary Fund. Not for Redistribution Oleh Havrylyshyn

International Monetary Fund, 2002, "Republic of Belarus: 200 I Article IV Consultation-Staff Report; Public Information Notice on the Executive Board Discussion; and Statement by the Executive Director for the Republic of Belarus," Country Report No. 02/23 (Washington: IMF, February). Johnson, Simon, Daniel Kaufmann, and Andrei Shleifer, 1997, ''The Unofficial Economy in Transition," Brookings Papers on Economic Activity: 2, pp. 159-239. Kaufmann, Daniel, Aart Kraay, and Pablo Zoido-Lobaton, 1999, "Governance Matters," World Bank Policy Research Working Paper 2196 (Washington: World Bank). Kornai, Janos, 1994, "Transformational Recession: The Main Causes," Journal of Comparative Economics, Vol. 19 (August), pp. 39-63.

___, 1998, "The System Paradigm," Working Paper No. 278, The Davidson Institute (Ann Arbor, Michigan: University of Michigan). .. Loungan.i, Prakash, and Nathan Sheets, 1997, Central Bank Independence, Inflation, and Growth in Transition Economies," Journal of Money, Credit and Banking, Vo l. 29 (August), pp. 381-99. Mercer-Blackman, Valerie, and Anna Unigovskaya, 2000, "Compliance with IMF Program Indicators and Growth in Transition Economies;· lMF Working Paper 00/47 (Washington: International Monetary Fund). Moers, Luc, 1999, "How Important Are Institutions for Growth in Transition Countries?" Tinbergen Institute Discussion Paper 99-00412 (Amsterdam). Murrell, Peter, 1992, "Evolution in Economics and in the Economic Reform of the Centrally Planned Economies," in The Emergence of Market Economies in Eastern Europe, ed. by Christopher Clague and Gordon C. Rausser (Cambridge. Massachusetts: Blackwell), pp. 35-53. Nellis, John, 200 I, "Time to Rethink Privatization in Transition Economies?" Chapter 7 in A Decade of Transition: Achievements and Challenges, ed. by Oleh Havrylyshyn and Salah Nsouli (Washington: International Monetary Fund), pp. 160-93. Pinto, Brian, Marek Belka, and Stefan Krajewski, 1993, "Transforming State Enterprises in Poland: Evidence on Adjustment by Manufacturing Firms," Brookings Papers on Economic Activity: I, Brookings Institution, pp. 213-70. Quah, Danny, 1999, "Cross-Country Growth Comparison: Theory to Empirics," CEPR Discussion Paper 2294 (London: Centre for Economic Policy Research). Roland, Gerard, 200 1, ''Ten Years After ...Transition and Economics," IMF Staff Papers, Vol. 48 (Special Issue), pp. 29-52.

___, and Thierry Verdier, 1999, "Transition and the Output Fall," Economics of Transition (EBRD), Vo l. 7 (January), pp. 1-28. Sachs, Jeffrey D., l 996, "The Transition at Mid-Decade,'' American Economic Review, Papers and Proceedings, Vol. 86 (May), pp. 128-33.

_, and Andrew M. Warner, 1996, "Achieving Rapid Growth in The Transition Economies of Central Europe," HIID Development Discussion Paper No. 544 (Cambridge, Massachusetts: Harvard Institute for lntemational Development), pp. 1-58. Sala-i-Martin, Xavier, 1997, "I Just Ran Four Million Regressions," NBER Working Paper 6252 (Cambridge, Massachusetts: National Bureau of Economic Research). Sarel, Michael, 1996, "Nonlinear Effects of Inflation on Economk Growth," IMF Staff Papers, Vol. 43 (March), pp. 199-215. Selowsky, Marcelo, and Ricardo Martin, 1997, "Policy Performance and Output Growth in the Transition Economies," American Economic Review, Papers and Proceedings, Vol. 87 (May), pp. 349-53.

©International Monetary Fund. Not for Redistribution RECOVERY AND GROWTH IN TRANSITION: A DECADE OF EVIDENCE

Stem. Nicholas, 1997, "The Transition in Eastern Europe and The Former Soviet Union." Chapter 2 in Lessons From the Economic Tran sition: Central and Eastem Europe in the 1990s, ed. by Salvatore Zechinni (Boston: Kluwerfor the OECD), pp. 35-58. Stiglitz, Joseph E., 1999, "Whither Reform: Ten Years of Transition," The Wo rld Bank Economic Review (September). Wolf, Holger C., 1997, "Transition Strategies: Choices and Outcomes'' (unpublished; New York: Stern Business School). Zettelmeyer, Jeromin, 1998, "Uzbek Growth Puzzle," IMF Working Paper 98/133 (Washington: International Monetary Fund). Zinnes, Clifford, Yair Eilat, and Jeffrey Sachs, 200 l, "The Gains from Privatization in Transition Economies: ls Change of Ownership Enough?" IMF Staff Papers, Vol. 48 (Special Issue), pp. 146-70.

Escaping the Under-Reform Trap

ANDERS ASLUND, PETER BOONE, and SIMON JOHNSON*

Most fo rmer Soviet republics have fa llen into an economic and political under refo rm trap. An intrusive srate imposes high tax rates and drives entrepreneurs into the unofficial economy, which further aggravates the pressure on of ficial busi nessmen. Tax revenues and public goods dwindle, fu rther reducing incentives to register business activity. This economic under-reform trap has a political coun telpart. Remarkably, Communist parties remain popular and opposed to estab Lishing the rule of law precisely in those places where they were able to delay and derail reform. No electoral backlash prompts the refo rms necessary to Leave the under-refO im trap. The best way out of the trap in countries such as Russia and Ukraine is increased economic and political competition among the elite. [JEL £65, H41, K42, P52]

ruts of East-Central Europe and most of the former Soviet Union have fallen ir to an economic and political under-reform trap. Managers hide their firms' activitiesP underground in order to escape regulation and reduce the bribes they have to pay. High levels of underground activity keep tax revenues low, which meam the government cannot afford to provide public goods, such as Jaw and order, thus further lowering the incentives for entrepreneurs to enter the official economy. These economies are caught in a trap: as few pay taxes, the tax burden upon those who do becomes unbearable, inducing entreprenems to stay under groun.j even though this keeps their economic efficiency low and prevents growth.

*Anders Aslund is Senior Associate at the Carnegie Endowment for International Peace in Washington. D.C. Peter Boone is Director of Research at Brunswick UBS Warburg in Moscow. Russia. Simon Johnson is Associate Professor in the Sloan School of Management at MIT. Simon Johnson thanks the MlT Entrepreneurship Center for support. Vi ctoria Levin has kindly provided research assistance for this article.

©International Monetary Fund. Not for Redistribution ESCAPING THE UNDER-REFORM TRAP

At the same time, senior officials and powerful private individuals Live well through corrupt deals and sharing rents. This economic under-reform trap has also an important political dimension. Remarkably, there is little evidence of an electoral backlash against the people who led countries into the under-reform trap. Countries that have not had much reform tend to have little governmental change, quite strong Communist parties, and little impetus for more reform. Conversely, however, countries that manage to reform find that the political support for reform subsequently strengthens. Until recently, these problems were perceived as temporary matters of transi tion, but it is now clear that many former Soviet bloc countries have become trapped in a rent-seeking equi]jbrium. Slow and ineffectual reform created the opportunity for corrupt bureaucrats and politicians to become entrenched and extract bribes from firms. High inflation offered huge temporary rents, and the longer it lasted the richer the rent seekers became. Slow privatization facilitated extortion by government officials. By the end of 2000 it was possible for an economy to be stabiUzed, widely privatized, and liberalized, yet remain trapped by corruption and a large underground economy. Most former Soviet republics appear trapped in this way. Can a country break out once it is deeply in the trap? In this paper, the possi bilities for two countries-Ukraine and Russia-are compared, and a two-part solution is offered. Economically, rents need to dwindle through competition and new entry, while political power needs to be dissipated as a consequence of competition among the elite. The policy goal should be to foster such competition. lt seems more difficult for Ukraine than for Russia to escape from the trap, but even Ukraine has some reason to hope. This paper updates findings first presented in Aslund, Boone, and Johnson (1996). The economics of the under-reform trap were studied in Johnson, Kaufmann, and Shleifer (1997). Here we explain the politics of the trap, both in terms of electoral outcomes and the pattern of competition among the elite.

I. The Economic Under-Reform Trap

Post-communist market economies fall into two main groups: those with sustained recovery fo llowing a relatively radical transforn1ation, and those stuck without sustained growth because of gradual refonn. In 1999, measured real GDP in East-Central Europe was 5 percent lower than in 1989, but in the Commonwealth of Independent States (the former Soviet Union without the Baltics), real GDP was just 56 percent of its 1989 level (ECE, 2000, p. 225). While performance has diverged dramatically, measured GOP exaggerates the discrepancy, because the unofficial economy has expanded much more in the Fonner Soviet Union than in east-central Europe. Table 1 shows the results of esti mating the share of the unofficial economy in total GOP through 1995 using the electricity consumption-based methodology presented in Johnson, Kaufmann, and Shleifer (1997). Two divergent development paths are evident. The rather liberal East-Central European countries started with a relatively large unofficial economy, which peaked in J 992, and then declined moderately. In contrast, the unofficial

©International Monetary Fund. Not for Redistribution ...() 0 Ta ble 1. Share of Unofficial Economy

Official Total Official TO(a) Share of lhe Unofficial Economy GOP GDP GDP GOP (percem of GOP) t()(a) in 1994 inl994 in 1995 in 1995

Countries 1989 1990 1991 1992 1993 1994 1995 {1989 = 100) (1989 = 100) (1989 = 100)(1989 = 100)

)> Eastern Europe :J Bulgaria 22.8 25.1 23.9 25.0 29.9 29.1 36.2 72.3 78.7 73.7 89.2 0. 9:! Czech Republic 6.0 6.7 12.9 16.9 16.9 17.6 11.3 81.0 92.4 84.3 89.3 en 27.0 28.0 32.9 30.6 28.5 29.0 83.4 84.3 84.7 87.1 Hungary 1:1.1 en::J>. Poland 15.7 19.6 23.5 19.7 18.5 15.2 12..6 92.0 91.4 98.3 94.9 c :J Romania 22.3 13.7 15.7 18.0 16.4 17.4 19.1 72.7 68.4 77.7 74.7 P- Slovakia 6.0 7.7 15.1 17.6 16.2 14.6 5.8 77.9 85.8 83.1 82.9 -u Lilhuania 12.0 113 21.8 39.2 31.7 28.7 21.6 43.9 54.1 45.1 50.6 3 Moldova 12.0 18.1 27.1 37.3 34.0 39.7 35.7 41.7 60.9 43.0 58.8 0 :J 12.0 14.7 23.5 32.8 36.7 40.3 41.6 51.3 75.5 49. 1 74.0 Russia c.... Ukraine 12.0 16.3 25.6 33.6 38.0 45.7 48.9 44.2 71.6 39.0 67.0 0 ::::J Uzbekistan 12.0 11.4 7.8 11.7 10.1 9.5 6.5 85.0 82.6 84.0 79.0 ::J en 0 Source: Johnson . Kaufmann. and Shleifer, 1997. :J Note:We report hereall Eastern Europeanand former Soviet Union countries for which these data are available.

©International Monetary Fund. Not for Redistribution ESCAPING THE UNDER-REFORM TRAP

share in former Soviet countries was initially small and rose sharply until 1994 before flattening out. Strikingly, while the Polish unofficial economy share declined from 1989 to 1995 by 3 percent of GOP, its share in Russia and Ukraine surged by 30 percent and 37 percent of GOP, respectively. More recent work (through 1997) shows the same pattern has continued-the unofficial economy is typically 2-3 times larger, as a percent of GOP. in the former Soviet Union than in east-central Europe (Eilat and Zinnes, 2000). Adjusting for the unofficial economy, the contraction in real GOP has been much lower than commonly supposed. In Russia, for example, output probably fe ll no more than 20 percent during transition, but qualitatively. the large unoffi cial economy is an essential part of the under-refonn trap. A large unofficial economy means that few pay taxes, inciting governments to raise tax rates for those few while prov iding minimal public goods, thus inducing entrepreneurs to opt for the underground economy in a vicious circle. Conversely, a marked difference has developed especially between Central Europe and CIS countries in tem1s of general government spending as a percent of both official and total GOP. In Central Europe (Poland, the Czech Republic, Slovakia, and Hungary), general government spending declined insignificantly from 51 percent of GOP in 1989 to 46 percent of GOP in 1999 (Tanzi, 1999). By contrast, general government spending in the CIS fe ll on average from 46 percent of official GOP in 1992 to 27 percent in 1999 (Tanzi, 1999). The low level of tax collection is no reflection of liberalism in most CIS coun tries. While their level of actual tax collection varies greatly, the tax burden on those who actual ly pay taxes is much higher than in Central Europe (Johnson, McMillan, and Woodruff, 2000b). In fact, either high state revenues or high de facto tax rates have characterized post-communist countries. The functioning of the tax system is a reflection of a broader qualitative differ ence between these two subregions. A group at the World Bank originally calcu lated a composite structural reform index, which mainly reflects external liberalization and price liberalization but also privatization (De Melo, Denizer, and Gelb, 1997). For an ideal market economy, this index would be 1.00. As early as 1992, central Europe had an average structural reform index of 0.83, while it stood at 0.29 in the CIS. Over time, the CIS countries have caught up, but when central Europe recorded 0.90 in 1999, the CIS countries reached an average of only 0.63 (see Table 2). Because of slow initial reforms, the CIS countries have become stuck at a lower level of Liberalization and privatization. Whj)e East-Central Europe and the former Soviet Union have converged greatly in stabilization, Liberalization, and privatization, this is not true of fairer taxation, fewer regulations, and the rule of law. Countries with less distortionary tax and regulatory systems have larger official economies, collect more tax revenues, and provide more public goods. The fom1er Soviet Union is particularly bad at protecting entrepreneurs from ext01tion by government officials (Frye and Shleifer. 1997; Shleifer, 1997). The survey reported in Johnson, McMillan, and Woodruff (2000a) indicates that the effective or perceived tax rates are almost twice as high in Russia and Ukraine compared with three East-Central European countries (Poland, Slovakia, and

©International Monetary Fund. Not for Redistribution Anders Aslund. Peter Boone. and Simon Johnson

Table 2. Structural Reform Index. 1990-99

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999

Central Europe Poland 0.68 0.72 0.82 0.82 0.83 0.79 0.79 0.81 0.86 0.86 Czech Republic 0.16 0.79 0.86 0.90 0.88 0.82 0.82 0.82 0.90 0.90 Slovakia 0.16 0.79 0.86 0.83 0.83 0.79 0.79 0.77 0.90 0.90 Hungary 0.57 0.74 0.78 0.82 0.83 0.82 0.82 0.87 0.93 0.93

Southeast Europe Romania 0.22 0.36 0.45 0.58 0.67 0.65 0.64 0.66 0.76 0.82 Bulgaria 0.19 0.62 0.86 0.66 0.63 0.61 0.57 0.67 0.79 0.79

Sal tics Estonia 0.20 0.32 0.64 0.81 0.83 0.77 0.78 0.82 0.90 0.93 Latvia 0.13 0.29 0.51 0.67 0.71 0.67 0.74 0.74 0.86 0.86 Lithuania 0.13 0.33 0.55 0.78 0.79 0.71 0.74 0.74 0.82 0.82

CIS Russia 0.04 0.10 0.49 0.59 0.67 0.64 0.71 0.72 0.64 0.64 Belarus 0.04 0.10 0.20 0.33 0.42 0.50 0.44 0.37 0.37 0.37 Ukraine 0.04 0.10 0.23 0.13 0.33 0.54 0.57 0.59 0.65 0.65 Moldova 0.04 0.10 0.38 0.51 0.54 0.64 0.64 0.64 0.76 0.76 Armenia 0.04 0.13 0.39 0.42 0.46 0.54 0.61 0.61 0.76 0.76 Azerbaijan 0.04 0.04 0.25 0.31 0.33 0.40 0.44 0.51 0.61 0.61 Georgia 0.04 0.22 0.32 0.35 0.33 0.50 0.61 0.66 0.79 0.79 Kazakhstan 0.04 0.14 0.35 0.35 0.42 0.50 0.64 0.66 0.79 0.72 Kyrgyzstan 0.04 0.04 0.33 0.60 0.71 0.71 0.67 0.70 0.82 0.79 Taj ikistan 0.04 0.11 0.20 0.26 0.42 0.40 0.40 0.39 0.55 0.58 Thrkmenistan 0.04 0.04 0.13 0.16 0.29 0.27 0.27 0.36 0.36 0.36 Uzbekistan 0.04 0.04 0.26 0.30 0.50 0.57 0.57 0.54 0.57 0.50

Sources: De Melo, Denizer, and Gelb (I 997): Havrylyshyn and Wolf ( 1999), p. 34; authors' calculations from EBRD (1998). p. 26. and EBRD (1999). p. 24. Notes: This index was originally established by De Melo, Denizer, and Gelb (1997), with World Bank assessments for 1990-94. They also indicated how their assessments were related to EBRD indices. Havrylyshyn and Wolf ( 1999) updated their series for 1995-97, while we have updated correspondingly for 1998 and 1999. The fonnula is straightforward. The first element is 0.3 times EBRD's indices for price liberalization and competition policy. The second element is 0.3 times EBRD's index for trade and foreign exchange liberalization. The third element is 0.4 times EBRD's indices for large- scale privatization, small-scale privatization and banking refonn. Each index is normalized to reach a maximum of I. Thus. this index represents liberalization to 73 percent, while the rest is privatization. The weights have been arbitrarily selected, but actually it does not matter much what weights are chosen for the countries relative standing to one another, as the covariance is great.

Romania). The evidence suggests that Central Europe, in particular, has made substantial progress towards refonning its institutions, while most of the former Soviet Union lags far behind, being caught in an under-reform trap, with high effective taxation and intrusive regulation. The most powerful entrepreneurs respond with various forms of "state capture" (Hellman, Jones, and Kaufmann, 2000), effectively privatizing public goods such as taxation and regulation to their own benefit, driving out competitors

©International Monetary Fund. Not for Redistribution ESCAPING THE UNDER-REFORM TRAP

who are trying to obey the law. Small entrepreneurs exit the offi cial economy, and about half of the economy is now underground. Russia, Ukraine, and most of the CIS have become true rent-seeking societies with a suboptimal equilibrium of low economic efficiency (for modeling of such an equilibrium, see Johnson. Kaufmann, and Shleifer, 1997).

II. The Electoral Under-Reform Tra p

The economic under-reform trap has a political counterpart. ln a democracy, it would be reasonable to presume that informed citizens would vote for political parties advocating policies that have brought about economic growth in other countries. In general, among postcommunist countries, market reform and democ racy are strongly positively related, as Figure I shows. Full democracies have usually unde1taken advanced market reform, while semi-democratic states have usually pursued limited reform, and real dictatorships have done Uttle reform, as is specified in Figure 2.

Figure 1. Democracy and Market Reform. 2000

7 • Turk menistan • "' Uzbekistan 6 • • Tajikbtan .0 Belarus :JQ; •Kazakismn �� 0 0t: 5 • Kyrgyz Republic "0 t: II "' ,.... •Aib,wia "' u Russia ·= s Armeni'' v 4 • •Cmali:t �� 0.. II � - • • Georgia Ukminc 8 E :c e 3 • • Macedonia 1= :::. Mtlldova "0c • Bulgari:t <.> � Romania lltlll2al') !.L. 2 • E"onia Poland Latvia Ctech Rep Lithuania • • • • Slovenia � Slovak•:• I () 0.1 0 . 2 0.3 0.4 0.5 0.6 0.7 0.8 ll.'.l

World Bank/EBRD Structural Reform Index (from 0 =low w I = high)

The countries of particular interest to us are those that have gotten stuck in an under-reform trap-those in the middle of Figure l: Ukraine, Moldova, Bulgaria. Romania, and Russia. These countries have underperformed in terms of economic reforms and economic growth (see Tables 1 and 2), and even so, the responsible governments have repeatedly survived relatively free elections. How could this have happened?

©International Monetary Fund. Not for Redistribution Anders Aslund, Peter Boone, and Simon Johnson

Figure 2. Democracy and Market Reform

Advanced Reform Limited Reform Limited Market Reform

Free Poland Czech Republk Estonia Hungary Latvia Lithuania Slovenia Slovakia Romania Bulgaria

Partialy free Croatia Albania Macedonia Anuenia Georgia Moldova Ukraine Kyrgyztan Russia Azerbaijan

Not free Kazakhstan Belarus Tajikistan Turkmenistan Uzbekistan

Note: Countries were classified by Freedom House (2000) as "free'' "panly free," or "not free.'' Theborder line between "advanced reform" and "limited reform'' has been put at 0.80 on the EBRD structural reform index for 2000. and as borderline between "limited reform" and "little market refom1," the index value 0.62 has been chosen (EBRD. 2000).

Assesslng Election Results

In order to understand the politics of under-reform, we scrutinize election results. We review only parliamentary election outcomes because presidential elec tions tend to be dominated by personal factors and offer a narrow choice, while local elections are rarely relevant for national policy. Only election results in coun tries with a reasonable degree of political freedom can be considered. Nineteen countries in the postcommunist region fulfill this criterion, being classified as either free or partly free by Freedom House (Karatuycky, Motyl, and Graybow, 1999, p. 15). Seven countries do not qualify as even partly free, and they are there fore excluded (Azerbaijan, Belarus, Kazakhstan, Tajikistan, Turkmenistan, Uzbekjstan, and Yu goslavia). To categorize the electoral outcomes, we need some killd of party structure. Hence, Kyrgyzstan is omitted, because it bad its flrst multi party elections in 2000, and even then parties played only a minor role. Party structures vary much more in postcommunist countries than in We stern Europe. For our purpose, electoral outcomes are most easily classified by the Commurust party and its direct successors. Although the Commurust parties have been renamed in most countries, one evident successor party usually exists, and splinter groups tend to alter their appearance.

©International Monetary Fund. Not for Redistribution ESCAPING THE UNDER-REFORM TRAP

Figure 3. Communist Party Electoral Performance and Market Economic Transformation, End-1999

Communist Parties Communist Parties Doing Well Doing Badly

Radical transfom1ation Hungary Croatia Poland Czech Republic Estonia Latvia Lithuania Slovakia Slovenia

Incomplete Bulgaria Armenia transformation Macedonia Georgia Moldova Romania Russia Ukraine

Note: The borderline between "radical transformation" and "incomplete transformation'' has been put m 0.80 on the stnuctural reform index for 1998 (sec Table 2). and as borderHne between ·'communist parties doing well'' and ·'communist parties doing badly'' 20 percent or the votes in the las t elections up to summer 2000 has been chosen (see Table 3)

Even Communist parties evolve. They can be divided roughly into two groups: those turning Social Democratic and those remaining Communist parties (some times becoming nationalist). Among the democratic countries, the old Communist parties have become Social Democratic in Poland, HLmgary, Slovakia, Lithuania, Slovenia, and Macedonia, while they remain bard line in the Czech Republic and throughout the fonuer Soviet Union. This dichotomy remained largely true at the end of 2000.

Parliamentary Election Results

All the democratic or semidemocratic countries in the region have undertaken some market economic reforms. If such reforms were truly unpopular, we would expect Communist parties to gain popularity and even win electoral majorities. However, Table 3 shows that Communist parties have fared poorly regardless of country and policy, with the exception of the marginal cases of Albania (where the "democratic" forces abandoned democracy in 1997) and Mongolia in 2000 (which is disregarded here as a very special case). By 1997, no other Communist party in the region bad attained one-third of the votes cast in the most recent democratic election, and the Communist party was not the senior partner in any government. The Communist parties in Italy, France, and Finland were actually larger during the Cold War. Apparently, the popular nostalgia for communism bas been very limited. Results from all the parliamentary elections are summarized in Table 3, and these offer some intriguing conclusions. At the end of 1999, these countries could

©International Monetary Fund. Not for Redistribution Table 3. Ex-communisfVo te Share in Parliamentary Elections in Post-communist Countries (Percent)

Countr)' FlfSl Election Yore Share Second Election Vote Share ThirdElection Vote Share Fourth Election Vote Share

Albania March-April 1991 56.2 March 1992 25.6 May 1996 Rigged (20.4) June-July 1997 52.8 Armenia 1990 minority July 1995 12.1 May 1999 12.0

Bulgaria June 1990 47.2 Oct. l991 33.1 Dec. .199-t 43.5 April 1997 22.1 Croatia April-May 1990 26.5 Aug. 1992 5.4 OcL 1995 Tm } Jan. 2000 n.a. 1 CzechRepublic June 1990• 13.6 June 1992 14.2 May 1996 10.3 June 1998 11.0 Estonia March 199(}! Minority SepL 1992 13.61 March 1995 5.9 March 1999 6.1 � Georgia Oct.-Nov. 1990 Tiny minority Oct. 1992 2.7-4 Nov. 1995 3.8 Oct. 1999 Tiny V>)>o Hungary March-April 1990 10.9 May 1994 33.0 May 1998 32.9 c ::J Latvia Man:h-April 1990 2 minority June 1993 12.0� Sept.-Oct. 1995 12.95 Oct. 1998 14.7� p. Lithuania Feb.-Man:h 1990 2 minoriry Oct-Nov. 1992 42.6 Oci.-.�OV. 1996 9.5 Oct. 2000 3 1..1 � Macedonia Nov.-Dec. 1990 25.8 Oct. 1994 48.3 � Oct. 1998 25.2 Moldova Feb. 19902 majoriry Feb. 1994 22.0 March l998 30.1 � ro Mongolia July-Aug. 1990 61.7 June 1992 57.0 June 1996 40.5 July 2000 Majority 0 0 Poland June 19896 minoriry Oct. 1991 12.0 Sepl 1993 20.4 SepL 1997 27.1 ::J Romania May 1990 66.3 Sept. 1992 37.9 Nov. 1996 21.5 NO\'. 2000 37.0 � Russia March 19902 majority Dec. 1993 12.4 Dec. 1995 22.7 Dec. 1999 24.3 0 ::J Slovakia June 19901 13.6 June 1992 15.2 Sept.-Oct. 1994 13.1 SepL 1998 l.f.7 0. Slovenia April l990 22.7 Dec. 1992 13.6 Nov. 1996 16.1 Oct2000 15.8 U> Ukraine Man:h 19902 majority March 1994 minoriry March 1998 24.7 3 0 .J Sources: Inter-Parliamentary Union (www.ipu.org). Chronicle of Parliamentary Elections and De\elopments.. 23-29 (1988-95}� Commission on Security and L Cooperation inEurope. Elections in Central and Eastern Europe. December 1990; Facts on File World News Digest 0 J Note: Countries that were classified by Freedom House (Karamyck.y. Motyl, and Grabow. 1 999) as not freehave beenexcluded. They were Azerbaijan. Belarus. .J

©International Monetary Fund. Not for Redistribution ESCAPING THE UNDER-REFORM TRAP

be divided into countries where the Communist parties received more than 20 percent of the votes cast in the latest parliamentary election (nine countries), and countries where they obtained Jess than 20 percent (eight countries). The threshold is a natural division with no country in the interval of 16-21 percent. The summary classification is displayed in Figure 3, matching this political division by their degree of market reform in 1998 (with the threshold set at 0.8), splitting the demo cratic transition countries into four quadrangles. Three typical electoral paths can be distinguished. First, the top left box in Figure 3 contains two countries that have undertaken radical marketizatioo, and even so, their Communist parties are doing well, namely Poland and Hungary. Their communists were almost routed in 1991 and 1990, respectively. They have staged strong comebacks to about 30 percent of the vote, however, as they were thoroughly reformed early on, becoming right-wing Social Democratic parties. These countries' successful market economic transformation convinced even communist leaders to reform, securing their political survival, and their conversion led to the formation of a market economic consensus. Second, in nine countries the Communist parties are now weak, receiving only I 0-16 percent in the latest elections. These countries have all gone through substantial market economic transformation and achieved substantial and lasting growth. They should all be in the right upper box in Figure 3 (the latecomers, Armenia and Georgia, have almost qualified). AJI these Communist parties have undergone far-reaching reform, with the exception of the Latvian National Harmony party, but not fast enough to jump onto the market economic band wagon. A broad consensus in favor of market reform was achieved without them. This does not mean, however, that they are discredited for good. In Lithuania, the former communists made a strong comeback in the elections of 2000 as social democrats. The bottom left box in Figure 3 represents the political under-reform trap. It contains six countries where Communist parties have been relatively successful, while the market economic transformation has been insufficient. These six coun tries are Bulgaria, Romania. Macedonia. Moldova, Ukraine, and Russia. Two different patterns are in evidence, depending on whether the Communist parties stayed in power or not. In Bulgaria and Romania the former Communist parties held on to power in democratic elections until 1996-97. The cause of their demise was rampant economic crises, for which the communist governments were clearly responsible. In these two countries, the conupt policies of partial reform were discredited, and people voted against their bitter experiences of communist governments in the post-communist period. In Bulgaria, a break with the system of rent seeking occurred, while the communists returned to power in Romania in 2000, illustrating the tenacity of the under-reform trap. In Russia, Moldova, and Ukraine, the old Communist parties have remained strong, usually gathering 20-30 percent of the votes cast. While not fo rmally in power in Ukraine or Russia, t the communists have been highly influential in

IMoldovan governments have been both unstable and hard to categorize precisely.

©International Monetary Fund. Not for Redistribution Anders Aslund, Peter Boone, and Simon Johnson

parliament, as the electoral system made them over-represented, allowing them to block reform legislation in coalition with other parties. Whether in government or not, strong Communist parties have ensured that the state has remained all-intrusive, with high taxation and extraordinary rent seeking. The natural political outcome has been a state dominated by powerful oligarchic business groups. The concentrated economic power around the state also circumscribed democracy, while politics became polarized between oligarchs and communists. Ironically, both groups favor similar economic policies and agree on having a rent-seeking state. Accordingly, the EBRD (1999) finds that state capture is the greatest in these semi-reformed countries (see also HeUman, Jones, and Kaufmann, 2000). The communists remain strong electorally because of dissatisfaction with the new semi-refonned system, which produces little growth but great inequality, while noncommunists vote for the strongest organized contenders against the communist threat. With few independent entrepreneurs, the Liberal right is too weak to counterbalance both communists and oligarchs. Thus communists remain reasonably strong and unreformed when market-oriented economic transforma tion is slow. This is the political under-reform trap.

Ill. Stickiness of the Reform Trap

The countries of greatest interest for our analysis are Russia and Uk:raine.2 Both countries fa iled to attain significant economic growth until 2000 because of late and partial refonns. To understand their problem, we fir st examine the initial rents in these societies. Next, we check the ensuing effects on the dominant economic-political groups. Finally, we consider the stability of the situation and how it may change.

Early Rents in Russia and Ukraine

A mixture of state enterprise managers, new entrepreneurs, government offi cials, commodity traders, bankers, and outright criminals have grown rich on government subsidies and regulations, that is, rent seeking. Today 's rent seeking can be traced back to four dominant early forms of rent seeking in both Russia and Ukraine. The first form of rent seeking was to buy commodities on the domestic market, which were cheap because of price regulation, and sell them abroad at the world market price. This required access to the commodities and export permits. About 40 percent of Ukraine's exports in 1992 were commodities (mainly metals; IMF, 1993b, p. 113), and their average domestic price was about 10 percent of the world market price. Hence, the total export rents amounted to some $4. 1 billion or 20 percent of GDP in 1992. The beneficiaries were managers of state metallurgical companies, commodity traders, foreign trade officials, and some politicians. In

2This section draws on Aslund ( 1999 and 2000).

©International Monetary Fund. Not for Redistribution ESCAPING THE UNDER-REFORM TRAP

Russia, which had more commodities to export (mainly oiJ and metals), the export rents in 1992 amounted to $24 billion or 30 percent of GOP (IMF, 1993a). The second method was to import certain commodities, notably food in Russia and energy in Ukraine, at a low subsidized exchange rate and resell them at a higher domestic price. The main beneficiaries were a small number of gas importers in Ukraine and food importers in Russia. In Russia, the import subsidies amounted to 17.5 percent of GOP in J 992 (IMF, 1993a), and they were probably as sizable in Ukraine, largely pertaining to energy imports from Russia. The third form of rent seeking was access to subsidized credits. In 1993, Ukraine experienced l0,155 percent inflation, but huge state credits were issued at an interest of 20 percent a year-that is, state credits were sheer gifts, and they were given to a privileged few. In 1992, net credit expansion to enterprises was no less than 65 percent of GOP and 47 percent of GOP in 1993 (calculated from IMF, l993b, p. 109; IMF, 1995, pp. 73, 105). In Russia, the net credit issue of the central bank was 31.6 percent of GOP, and it was less in 1993 (IMF, 1993a). Thus, these rents were greater and more lasting in Ukraine than in Russia. The fourth form of rents was direct budget subsidies, which amounted to 8.1 percent of GOP in 1992 and 10.8 percent of GOP in 1993 in Ukraine (IMF, 1995, p. 94). ln Russia, they were 10.8 percent of GOP in 1992 and 9.4 percent of GOP in 1993. In both countries, these subsidies were concentrated on agriculture and energy-that is, gas in Ukraine and the coal industry in both countries. The subsi dized industries became totally crirninalized by a struggle over these subsidies. In total, rents as a share of GOP were huge in both countries but somewhat higher in Ukraine than in Russia in both 1992 and 1993 because of greater credit emission in Ukraine (Aslund, 1999, 2000). Much of these rents have been accumulated abroad in tax havens. In this way, a small select group of privileged insiders usurped a huge share of GOP in the early years of transition and grew strong. They have no reason to abandon their enormous power and wealth, which are based not on property, but on diverted financial flows. For society, the result bas been sharply rising income differentials. Ukraine has reached a Gini coefficient of 47 and Russia 48, which is about the same as the Latin American average (Milanovic, 1998, p. 41). By 1996, macroeconomic stabilization was happily taking hold at long last in both coun tries, as the original rents dwindled, but the rent seekers stayed rich and powerful, inventing new rents. Both Russia and Ukraine have well-developed oligarchic power structures of so-called fl flancial-industrial groups, whose essence is to use political influence to extract rents. A good relationship with the presidency has been essential to success. In both cases, control over the media has been helpful to developing a power base. In Ukraine, the oligarchs sit in parliament themselves and lead large parties, while Russian oligarchs tend to hire parliamentarians and purchase specific votes. The Russian oligarchs started as small traders. Then they became bankers and commodity traders, moving on to the production of raw materials for export in the mid-1990s, and now they are going into manufacturing. The Ukrainian oligarchs remain commodity traders, but a few have taken to metallurgy and the production

©International Monetary Fund. Not for Redistribution Anders Aslund, Peter Boone, and Simon Johnson

of energy (Brady, 1999; Aslund, 2000). The government and the leading busi nessmen live in a curious symbiosis and mutual dependence. In order to be re elected, politicians need the support of the top businessmen (particularly those who control the media), but the tycoons can be ousted by politicians if they turn against the political leaders. Especially in Ukraine, an iron triangle of government, businessmen. and parliament ruled the country until 1998. AU three groups favored a maximum of regulation and state interference to maximize rent seeking and corruption, while the effects on the population as a whole were of little consequence. The govern ment malfunctioned in the interests of the rent-seeking elite. This model of self reinforcing rent seeking looked like a frightfully stable suboptimal equilibrium, reminiscent of the distribution of power in a stagnant African country (Collier and Gunning, 1999). Yet, the nature of the oligarchs' interaction with one another and the govern ment changes swiftly. The main institutional change occurred with privatization. Previously, the budding oligarchs merely colluded with state enterprise managers and government officials. Now, they became owners of large enterprises, which were by necessity visible. Their dependence on the state diminished. They started defending private property rights, and they represented different industrial inter ests. Their greater security and more clearly defined interests intensified their competition. As a result, their competition was transformed from obscure court intrigue to public politics. 1n Russia, this occuned in 1994, after the initial voucher privatization, while in Ukraine it happened only in 1998-99, after the large metal lmgical companies had been privatized in obscure inside privatizations. In parallel. the legal situation changed. At the end of communism, large-scale business was pervasively criminal and the murder rate rose steeply. Government was unable to provide d1e legal institutions that make contracts enforceable. Crime grew increasingly organized, and major businessmen established d1eir own secu rity organizations. Greater order reduced the murder rate (Aslund, 1997). In Russia, the internecine murdering among major businessmen ceased in 1994, while it continued until 1996 in Ukraine. In recent years, the police have reestab lished their monopoly of violence, but their services are often bought by busi nessmen (Hellman, Jones, and Kaufman, 2000). The demand for court services is rising steeply, prompting higher bribes for judges, while civil servants are deprived of bribes, since they no longer have much to sell. After most property has been distributed by state officials, the focus of bribery moves to judges responsible for the security or redistribution of property.

Policies for Breaking the Trap

Looking back on the past decade, the fast qualitative changes are striking. and the stability of dle suboptimal equilibrium must not be exaggerated. For instance, overindustrialization is soon to become a memory, as value detraction has faded. From 1989 to 1998, the share of industry in GDP in Russia and Ukraine contracted by 15 and 14 percent, respectively (World Bank, 2000). Future economic reform in the former Soviet Union should be seen in this new light. The best option of

100

©International Monetary Fund. Not for Redistribution ESCAPING THE UNDER-REFORM TRAP

swift, comprehensive, and radical reform is no longer available, leaving two contrary alternatives, either collusion or competition. The advantages of collusion or deal making have been analyzed by some economists and political scientists. Andrei Shleifer and Daniel Tre isman (1998 and 2000) and Treisman ( 1998) have argued that it facilitated privatization and macroeconomic stabilization in Russia. Trus was certrunly true for privatization. Ukrrune. however, has abounded with deal making with no benefit for reform. A competitive approach appears, instead, more likely to reduce rent seeking. In their book, Mercantilism as a Rent-Seeking Society, Robert B. Ekelund and Robert D. To llison argued that mercantilism ended in Britain because of a compe tition between the royal court and the parliament over monopoly rents and juris diction. In a similar vein, Andrei Shleifer and Robert Vishny (1993) have suggested that corruption should be fought through competition rather than prohi bition, which often implies a reinforcement of the monopoly of corruption, char acteristic of the Stalinist system. Shleifer and Vishny ( 1998) and Shleifer and Tre isman (2000) have devised methods for how to organize competition among rent seekers to drive bribes and rents down toward zero. Wben rents dwindle, smart operators might find productive business more lucrative. Competition is the opposite of a political and economic monopoly, and it involves both politics and economics. A competitive approach to reform can be described as seven related goals: (1) to split the political and business elite into different groups; (2) to form independent political associations and enterprises with secure legal bases; (3) to encourage competition among different groups; (4) to acrueve political and economic discontinuity; (5) to diminish the state's stranglehold on economy and society; (6) to achieve a maximum of transparency; and (7) to break up or regulate monopolies. This competition should aim to diminish rent seeking, thus enhancing economic efficiency. There is a fundamental difference between taxation and regulation with regard to desirable reforms. In taxation, competition must be avoided, as it leads to the kind of overgrazing that has proliferated in the CIS and in many developing coun tries (Shleifer and Treisman, 2000; Johnson, Kaufmann, and Zoido-Lobaton, 1998). In regulation, however, competition can minimize bribery. A first goal is to create more independent bodies. Throughout the post-Soviet region, rent seeking has prospered around so-called natural monopolies in energy and transportation (oil, gas, coal, pipelines, railways, and telecommunications), but many of these are not natural monopolies. Telecommunications have been divided and successfully turned competitive in Russia. Similarly, Russia has broken up its oil industry into a dozen major oil companies. Although only one of these privatizations was undertaken through competitive bidding, the Russian oil industry bas been transformed into a rughly competitive industry. The natural gas monopoly Gazprom, on the contrary, retains most ministerial regulatory functions, forming the mrun bastion of asset stripping by managers through barter and transit pricing, while it is badly mismanaged. Coase's Theorem (Coase, 1988), that the initial distribution of property does not matter because property can be traded, has been much maligned. It presup poses the absence of transaction costs, which are very high in the transition

101

©International Monetary Fund. Not for Redistribution Anders Aslund, Peter Boone, and Simon Johnson

economies. Considering high transaction costs, however, it can be modified into a less stringent hypothesis: it is more important that privatization be undertaken than how it is undertaken. Historically, states have lived without private property for centuries, while robber barons tend to be rusciplined by the market within a few decades. Over time, more property legislation is promulgated, and property rights are reinforced through practice, as reflected by increased usage of courts by enter prises (Hendley and others, 1997). According to the EBRD (1997), the private sector had grown to account for 70 percent of Russia's GDP by 1997, while the Ukrainian private sector only deliv ered 50 percent of GDP in that year but slowly expanded to 60 percent by 1999 (EBRD, 2000). The combination of a growing private sector and the enhancement of property rights amount to a threat to the old habits of rent seeking. Certain state bodies also need autonomy, especially regional and local organs, from the central state. Today, they have little independence and thus responsibility, whlle the center is unable to supervise them, which inspires cheating by local organs and arbitrary rule by the center. Most taxes are shared between two or three administrative levels in a shifting and arbitrary fashion, which leads to additional tax revenues of regional organs being confiscated by the central state (Kravchuk, 1999). Ideally, each tax would be fully allocated to one adrrurustrative level, for instance, VAT, excise taxes, and foreign trade taxes to the central state, and prop erty taxes to local authorities. To make the division crystal clear, each tax base should be allocated to one adm.irustration level, and each kind of expenditure should be entirely financed by one level, for instance, defense by the central government, but schools by local organs. Finally, the state tax service should be divided between the three adminlstrative levels as well to clarify responsibility and accountability (Sbleifer and Treisman, 2000). Fiscal autonomy would encourage competition between different administrative levels. A large number of private enterprises are needed to support political competi tion. Autonomy for srna.U entrepreneurs has empirically proven rather easy to create. Essentially, only two things are needed-a simplified registration system and a simple lump sum tax. The early introduction of wide-scale lump sum taxes spearheaded the evolution of many small private enterprises in Poland, which later formed the social, political, and economic base of radical market economic tran sition (Aslund, 1985). In 1998, Ukraine successfully launched a fixed Jump-sum tax for small entrepreneurs. It has stimulated the development of millions of smaU enterprises, which will be the base of a true market economy. If the state is characterized by rent seeking, state resources are used for private enrichment and the repression of enterprise. Then, falling state revenues are desirable to limit rents. With falling state revenues, resistance against a rational tax reform is also likely to fade, as leading rent seekers will no longer perceive a cumbersome tax system as an effective deterrent against business by outsiders. Or they might be too weak to block a liberal tax reform. Thus, Georgia, Kazakhstan, and Kyrgyzstan undertook radical tax reforms after state revenues had fallen sharply. It is not enough, however, that state activities are cut for a lack of funds. Public mandates must not be left unfunded but rather eliminated. A major cause of corruption is that semiautonomous state organs try to raise their own revenues to

102

©International Monetary Fund. Not for Redistribution ESCAPING THE UNDER-REFORM TRAP finance state programs that remain on the books but are no longer funded by the central government (Rose-Ackerman, 1999). Such public fund-raising .is usually undertaken through extortionary fees and penalties, being the most ineffective taxes and amounting to a major cause of corruption. Tbe most plausible interpretation of growing barter and arrears up to 1998 was that they fo1med a mechanism of extracting government subsidies and then priva tizing these subsidies through management theft. Total enterprise subsidies amounted to no less than 16 percent of GOP in 1998 (Pinto, Drebentsov, and Morozov, 1999). Otherwise, the cost of a bruter deal should be deterring at about 25 percent of the deal (Djankov, 1999, p. 131). This extraction of implicit subsi dies through barter is fac ilitated through its nontransparency. With the financial crash in Russia, barter fell like a stone from 54 percent of all interindustry trans actions to 21 percent in August 2000 (Russian Economic Barometer, 2000). A parallel development has occurred in Ukraine. The government was forced to harden its own budget constraints, demanding tax payments in cash, which elimi nated the possibilities to extract tax rebates through barter or offsets. The effect of the dwindling barter has been not only greater monetization but also a leveling of the playing field, intensified competition, and economic growth. One of the most erroneous ideas in much of the transition literature is that political and economic continuity are vital, because a valuable administrative capital exists that must be utilized (Stiglitz, 1999). On the contrary, a maximum of discontinuity is desirable in both the political and economic spheres. The EBRD (1999) Transition Report 1999 shows that the greater the discon tinuity in the ruling elite, the more radical and successful the market economic reform. Progress in liberalization "was twice as high in countries where the polit ical executive was replaced as in those where the incumbent from the communist era remained in office" (p. 106). Transition countries with more competitive polit ical systems and more unstable governments have tended to achieve greater progress in economic reform. The five transition countries with the most frequent changes in government have been the three Baltic States, Poland, and Bulgaria, of which four have been among the most successful reformers (p. 1 12). The apparent explanation is that turnover of personnel leads to greater transparency and more checks on corruption. The greater turnover of governments can also be seen as a reflection of more effective democracy, one of the best checks on a rent-seeking elite (see Figure I). The same is true of certain economic discontinuity. Drazen and Grilli (1993) have modeled how economic crises may facilitate reform by unde1mining harmful vested interests both financially and politically. The Russian financialcrash under mined the ruling oligarchy, thus facilitating substantial reforms, notably budget and tax reforms, which contributed to unexpected economic growth in 1999 and 2000. Thanks to its hyperinflation of 1996-97, Bulgaria changed political leader ship through democratic elections and undertook far-reaching liberal economic reforms, seemingly breaking out of its under-reform trap. The Romanian crisis of 1996-97, on tbe contrary, does not appear to have been sufficiently severe. Insufficient reforms proved ineffective and allowed the comeback of rather unre formed communists through democratic elections in 2000.

103

©International Monetary Fund. Not for Redistribution Anders Aslund, Peter Boone, and Simon Johnson

Economically, market reform in Russia and Ukraine has proceeded. Both countries are now liberal and open economies with large exports in relation to their GDP (see Table 2). Some monopolies are truly natural, but they can be structured to reduce rent seeking. For example, the power companies in Russia are aligned with individual regions. and regional governors have substantial control over tariff setting and personnel appointments. Therefore, these companies have become the fiefdoms of regional leaders, who typically keep tariffs low, prohibit electricity companies to cut off local enterprises, and force them to supply regional govern ments for free. If several regional companies are merged. they fall outside of the control of each governor. A popular idea in attempts to reduce corruption through administrative harass ment has been the attempted establishment of one-stop shops for licensing and fo reign investors in various countries, but such efforts run counter to the compet itive approach and they have failed in many countries, as this implies a monopoly of corruption. Instead, several agencies could be allowed to issue the same licenses and permits, prompting them to provide better and cheaper services in competition with one another. For example, a visit to Bishkek in 1998 revealed that the city architect had swiftly registered all land for housing and issued titles for the land, because two competing agencies were doing the same, and it was enough to get title from one of them. Similarly, enterprises could be allowed to register with different agencies, as long as multiple registrations are not required. Organs at federal, regional, and local level could all be allowed to issue all licenses that exist. The basis of civil society is public learning, which is probably a reason for the brevity of each method of rent seeking. As people have understood, reforms have become necessary. Either followers have jumped into the game to make money for themselves, for instance, driving rents down through arbitrage, or others have reacted against the rent seeking and demanded structural changes. For instance, cheap credits could be publicly defended in 1992 in Russia and Ukraine, but everybody knew they caused hyperinflation by 1994. The quality, and thus freedom, of the media is crucial for this public learning. Besides, much of the old rent seeking has been built on structural inertia and fear of change, but as the necessary structural changes proceed through downsizing by default, many obsta cles to reform fade away. Our line of argument fits the model of Daron Acemoglu and James Robinson (2000) of political losers as barriers to economic development. They argue that socially beneficial economic reforms are being blocked, "when the political power of the incumbent is threatened by economic innovation" (p. 128). Their model indicates that the incumbent monopolist is more likely to block the introduction of new technologies when political rents from staying in power are greater, when monopoly profits from blocking are greater, and when the tax revenue they can collect from rivals are smaller. While this paper discusses an oligarchy rather than a monopoly, it shows how this confluence of political and economic oligarchy works in practice. A corollary of this reasoning is that an oligarchy is much better than monopoly and dictatorship. Belarus and Turkmenistan present a far more stable under-reform

104

©International Monetary Fund. Not for Redistribution ESCAPING THE UNDER-REFORM TRAP

trap, which is held together by an iron triangle of 80 percent state ownership, minimal liberalization of the economy and full dictatorship (EBRD, 2000).

IV. Conclusions Russia and most former Soviet republics have actually become rather normal market economies, because most market economies have weak rule of law, a few dominant business groups, and a close relationship between government and busi ness. Russian and Ukrainian levels of corruption and political conflict are not unique. They have fallen into a fairly common trap: high corruption drives the economy underground and offers few incentives to enter the official economy. We have discussed this as an under-reform trap, but this is really a special case of a more general development trap.3 There are three possibilities for market economies trapped with weak institu tions. First, there may never be growth, as bas been the experience in much of Africa for decades. Second, if it is in the interest of some part of the elite, some growth may occur, and this creates the incentive for managers and governments to behave better and steal less. Growth becomes self-fulfilling and self-supporting, as in poorer parts of Europe over the past 40 years. Eventually, institutional reform occurs, but it usually requires the involvement of a third party (like the European Union). Third, rapid growth may alternate with episodes of economic collapse. This has been the experience of Latin America over the past 40 years, and it may now be the case of East Asia. Institutions cannot be built without political demands for institutions, and such demands depend as much on the nature of the economic system as on the political system. For a dictatorship or bus.iness hegemony, any institution is harmful, as it implies a reduction of absolute power. An oligarchy may find an equilibrium, but it is potentially unstable, as is an oligopoly. The question whether there is a suffi ciently strong demand for sound market economic institutions in Russia and Ukraine is a question whether the economic and political pluralism is strong enough to allow these countries to break out of their under-reform trap. The finan cial crash in Russia broke many oligarchs both financially and politically, while broadening the number of big businessmen. It also increased the demand for stable financial institutions in all groups of society. Meanwhile, increasing competition is bound to reduce rents, and eventually rent seekers may opt for profits instead. The driving force of positive change has to be entrepreneurs who feel they are at a disadvantage because of the privileges of others.

3for evidence on the importance of institutions in economic development more generally, and the long-lasting effects of early institutions (such as how a country was colonized), see Acemoglu, Johnson, and Robinson (2000). For the size and determinants of the underground economy around the world, see Friedman and others (2000).

105

©International Monetary Fund. Not for Redistribution Anders Aslund, Peter Boone, and Simon Johnson

REFERENCES

Acemoglu. Daron, and James A. Robinson, 2000, "Political Losers as a Barrier to Economic Development," American Economic Review: Papers and Proceedings, Vol. 90 (May), pp. 126-44.

Acemoglu, Daron, Simon Johnson, and James A. Robinson, 2000, "The Colonial Origins of Comparative Development: An Empirical Investigation," NBER Working Paper No. 7771 (Cambridge, Massachusetts: National Bureau of Economic Research).

Aslund, Anders, 1985, Private Enterprise in Eastern Europe: The Non-Agricultural Private Sector in Poland and the CDR, 1945�3 (London: Macmillan).

---, 1997, "Economic Causes of Crime in Russia," in The Rule of Law and Economic Reform in Russia, ed. by Jeffrey D. Sachs and Katharina Pistor (Boulder. Colorado: Westview Press).

---, J 999, "Why Has Russia's Economic Transformation Been So Arduous?" paper presented at the World Bank's Annual Bank Conference on Development Economics, Washington, April.

---, 2000, "Why Has Ukraine Failed to Achieve Economic Growth?" in Economic Refo rm in Ukraine: The Unfinished Agenda, ed. by Anders Aslund and Georges de Menil (Armonk, New York: M.E. Sharpe).

---, Peter Boone, and Simon Johnson, 1996, "How to Stabilize: Lessons from Post Communist Countries." Brookings Papers on Economic Activity: 1, Brookings Institution, pp. 217-313.

Brady, Rose, I 999, Kapitalivn: Russia 's Struggle to Free Its Economy (New Haven, Connecticut: Yale University Press).

Coase, Ronald H., 1988, The Finn, the Market, and the Law (Chicago, fllinois: University of Chicago Press).

Collier, Paul, and Jan Willem Gunning, I 999, ''Explaining African Economic Perfonnance," Journal of Economic Literature, Vol. 37 (March), pp. 64-111.

De Melo, Martha, Cevdet Denizer. and Alan Gelb, 1997, ''From Plan to Market: Patterns of Transition," in Macroeconomic Stabilization in Transition Economies, ed. by Mario I. Blejer and Marko Skreb (New York: Cambridge University Press).

Dj ankov, Simeon, 1999, "Enterprise Restructuring in Russia," in Russian Ente17Jrise Refo rm: Policies to Further the Transition, ed. by Harry G. Broadman, World Bank Discussion Paper No. 400 (Wash jngton: World Bank).

Drazen, Allan, and Vittorio Grilli, 1993, "The Benefits of Crises for Economic Reforms," American Economic Review, Vol. 83 (June), pp. 598-607.

Economic Commission for Europe (United Nations), 2000, Economic Survey of Europe, No. I (April).

Eilat, Yair, and Clifford Zinnes, 2000, "The Evolution of the Shadow Economy in Transition Countries: Consequences for Economic Growth and Donor Assistance," CAER II Discussion Paper No. 65 (Cambridge, Massachusetts: Harvard Institute for lntemational Development).

Ekelund, Robert B.. and Robert D. Tollison. 1981, Mercantilism as a Rent-Seeking Society (College Station, Texas: Texas A&M University Press).

European Bank for Reconstruction and Development, 1995, Transition Report 1995 (London).

---, 1996, Transition Report 1996 (London).

---, 1997, Transition Report 1997 (London).

106

©International Monetary Fund. Not for Redistribution ESCAPING THE UNDER-REFORM TRAP

---. 1998, Tr ansition Report 1998 (London).

---, 1999, Transition Report 1999 (London).

---, 2000, Transition Report 2000: Employmem. Skills and Tr ansition (London). Friedman, Eric, and others, 2000, "Dodging the Grabbing Hand: The Determinats of Unofficial Activity in 69 Countries," Journal of Public Economics, Vol. 76 (June), pp. 459-93. Frye, Timothy, and Andrei Shleifer, 1997, "The rnvisible Hand and the Grabbing Hand,'' American Economic Review, Papers and Proceedings, Vol. 87 (May), pp. 354-58. Havrylyshyn, Oleh, and Thomas Wolf, 1999, "Growth in Transition Countries, 199 1-1998: The Main Lessons," paper presented at the International Monetary Fund conference "A Decade of Transition," Washington. February. Hellman, Joel S., Geraint Jones. and Daniel Kaufmann, 2000, "Seize the State, Seize the Day: State Capture. Corruption and lntluence in Transition," Policy Research Working Paper No. 2444 (Washington: World Bank). Hendley, Kathryn, and others, I 997, "Observations on the Use of Law by Russian Enterprises," Post-Soviet Affairs, Vol. 13 (January-March). pp. 19-41. International Monetary Fund. 1993a. Russian Fe deration Economic Review (Washington).

---, 1993b, Ukraine 1993 Economic Review (Washington). ---, 1995, Ukraine 1994 Economic Review (Washington). Johnson, Simon. Daniel Kaufmann, and Andrei Shleifer, 1997. "The Unofficial Economy in Transition," Brookings Papers on Economic Activity: 2, Brookings Institution. pp. 159-239. Johnson, Simon, Daniel Kaufmann, and Pablo Zoido-Lobaton, 1998. "Regulatory Discretion and the Unofficial Economy," American Economic Review: Pape rs and Proceedings, Vol. 88 (May), pp. 387-92. Johnson, Simon, John McMillan, and Christopher Woodruff. 2000a, "Why Do Firms Hide? Bribes and Unofficial Activity After Communism." CEPR Discussion Paper No. 2105 (London: Centre for Economic Policy Research). ---, 2000b. ·'Entrepreneurs and the Ordering of Institutional Refom1: Poland, Slovakia, Russia, and Ukraine Compared," Economics of Tr ansition, Vo l. 8, pp. 1-36. Karatnycky, Adrian, Alexander Motyl. and Charles Graybow, eds., 1999, Nations in Transit, 1998 (New York: Freedom House). Kravchuk. Robert S .. 1999, "The Quest for Balance: Regional Self-Government and Subnational Fiscal Policy in Ukraine," in State and Institution Building in Ukraine, ed. by Taras Kuzio, Robert S. Kravchuk, and PauJ D' Anieri (New York: St. Martin's Press). MiJanovic, Branko, 1998, Income, Inequality, and Poverty During the Transition from Planned to Market Economy (Washington: World Bank). Pinto, Brian. Vladimir Drebentsov, and Alexander Morozov, 2000. "Dismantling Russia's Nonpayments System: Creating Conditions for Growth,'' World Bank Technical Paper No. 471 (Washington: World Bank). Rose-Ackerman, Susan, L999, Corruption and Government: Causes, Consequences, and Reform (New York: Cambridge University Press). Russian Economic Barometer, 2000, JII Quarter (Moscow: REB). Shleifer, Andrei, 1997, ·'Joseph Schumpeter Lecture: Government in Transition," European Economic Review, Vol. 41, pp. 385-410.

---, and Robert W. Vishny. 1993, "Corruption," Quarterly Joumal of Economics. Vol. 108 (August), pp. 599-6 17.

107

©International Monetary Fund. Not for Redistribution Anders Aslund, Peter Boone. and Simon Johnson

---, 1998. The Grabbing Hand: Govemment Pa thologies and Their Cures (Cambridge. Massachusetts: Harvard University Press).

Shleifer, Andrei, and Daniel Trei sman. J 998, The Economics and Politics of Transition to an Open Market Economy: Russia (Paris: OECD Development Center).

---, 2000, Without a Map: Political Ta ctics and Economic Reform in Russia (Cambridge, Massachusetts: MJT Press). Stiglitz, Joseph E., 1999, "Whither Refonn? Ten Years of Transition," Annual World Bank Conference on Development Economics, Washington, April. Tanzi, Vito, 1999, "Transition and the Changing Role of Government," Finance and Development, Vo l. 36 (June), pp. 20-23. Treisman, Daniel S., 1998, "Fighting Inflation in a Transitional Regime: Russia's Anomalous Stabilization," Wo rld Politics. Vol. 50 (January). pp. 235-65.

United Nations: Economic Commission for Europe, 2000, Economic Survey of Europe, No. I (New York).

World Bank, 2000, Wo rld Development Indicators, CD-ROM (Washington).

108

What Moves Capital to Transition Economies?

PIETRO GARIBALDI, NADA MORA RATNA SAHAY. and JEROMIN ZETIELMEYER*

The transition economies in Europe and the fo rmer Soviet Union between 1991 and 1999 differed widely in terms of total capital flows and the share and compo sition of private flows. With some exceptions (notably Russia), the main source of private inflows was fo reign direct investment. Portfolio investment was volatile, and concentrated in a handful of countries. Regressions show that direct investment can be well explained in terms of economic fundamentals, whereas the pres- , ence of a financial market infrastructure and a property rights indicator are the only explanatory variables that seem to have a robust effe ct on portfolio investment. [JEL F2 1, F32, P27]

ross-border capital flows are an important part of the transition story. Access to external financing determines the size of the current account deficit that a countryC can run when it is reorienting its production and consumption structures and rebuilding its capital stock. So, there may be a link between the ability to attract capital inflows and the speed of transition. Moreover, the composition of capital fl ows likely affects how much they are associated with the transmission of know-how and changes in corporate governance, and the degree to which they are subject to sudden reversals. Finally, the magnitude, composition, and stability of

'Pietro Garibaldi is Associate Professor of Economjcs at Bocconj University. Nada Mora is a Ph.D. candidate at MIT. Ratna Sahay is a Division Chjef and Jeromin Zettel meyer is a Senior Economist. both in the IMF's Research Depanment. This paper was written while Garibaldi and Mora were at the Research Depanment of the IMF. An earlier draft of this paper was presented at the conference on "A Decade of Transition: Achievements and Challenges" held at the IMF in February 1999 and at the 1999 European Economic Association Annual Conference. We are grateful to John Cuddjngton, conference panicipants and seminar participants at the !MF, and particularly Oleh Havrylyshyn for comments and suggestions. Research assistance by Grace Juhn and Hruyan Shi is gratefully acknowledged.

109

©International Monetary Fund. Not for Redistribution Pietro Garibaldi. Nada Mora, Ratna Sahay, and Jeromin Zettelmeyer

capital flows during transition may be indicative of broader aspects of the transi tion process. How long did the economy rely mainly on official and exceptional financing, such as debt rescheduling? How large and volatile were pottfolio inflows? What was the role of private inflows other than FDI and pottfolio investment? The objective of thls paper is twofold. First, it seeks to document the magru tude and composition of capital flows to the transition economies of Central and Eastern Europe and the former Soviet Union from the early 1990s until 1999. We describe the main trends in time, as well as cross-sectional variation between the main regions, and within regions. We particularly highlight cross-country hetero geneity in the behavior of (foreign) direct investment and portfolio investment, wh.ich played an important role in many countries as a source of financing and in the case of pottfolio investment-as a source of macroeconomic volatility. We then explore to what extent inward direct investment and portfolio investment flows in time and across countries can be accounted for using individual country characteristics and some common controls. The narrow focus on inward direct investment and pottfolio investment (as opposed to a broader capital flows concept) enables us to interpret the empirical regularities we find in terms of the behavior of private foreign investors. Existing analyses of aggregate capital flows to transition economies concen trate primarily on the determinants and effects of foreign direct investment flows to Central and Eastern Europe (see, in particular, Lansbury, Pain, and Smidkova, 1996; Holland and Pain, 1998; Brenton, di Mauro, and Liicke, 1999; and Bevan and Estrin, 2000). This study takes a broader view, characterizing capital flows more generally and exploring a wide range of potential determinants of both direct investment and portfolio investment in a larger sample of 25 transition economies that include the countries of the former Soviet Union. To our knowledge, the only study that adopts a similar perspective is Claessens, Oks, and Polastri, (1998), who analyze capital flows for 21 transition economies until 1996. Claessens, Oks, and Polastri provide more qualitative and descriptive country information than we do, particularly on the early transition years, and nm some regressions for broader capital flow aggregates in addition to foreign direct and portfolio investment. Our paper analyzes in more detail the heterogeneity of levels and composition of capital flows to transition economies in quantitative terms, and covers three more years of data.1 In view of the relatively large data set, we are able to consider a wider range of potential explanatory variables for direct investment and portfolio investment flows than the earlier Literature and use standard specification search techniques to arrive at a parsimoruous model rather than assuming a small number of potential determinants from the outset. In addition, this paper contains some new data work. We build on the direct investment and portfolio investment data published in the IMP's 2000 Balance of Payment Statistics Ye arbook by extending it for the earlier years and filling in the missing values using infonnation from other sources, including IMF desk

IFor a survey of the recent past. see also the EBRD's Transition Report 2000, which emphasizes developments during 1999-2000.

110

©International Monetary Fund. Not for Redistribution WHAT MOVES CAPITAL TO TRANSITION ECONOMIES?

economists. To control for direct legal restrictions to both direct investment and portfolio investment flows in our regressions, we also created indices of direct investment and portfolio investment restrictions based on qualitative information recorded in the IMF's Annual Report on Exchange Arrangements and Exchange Restrictions. The paper highlights both the quantitative importance of capital flows to tran sition economies as a group-which, since about 1993, have been higher than the average flows to other emerging market economies-and the high degree of heterogeneity across the transition countries. Foreign direct investment was gener ally the most important source of private capital, but portfolio investment also played a major role in a handful of countries, as did bank lending. Russia stands out as a special case in many respects: in spite of substantial official financing, it was a net exporter of capital over the period we study, and until 1998 it attracted much more portfolio investment than foreign direct investment. The regression results show that the cross-country pattern of foreign direct investment can be explained reasonably well by standard macroeconomic fundamentals, including institutional indicators. However, we find that it is much harder to "explain" the behavior of portfolio investment.

I. Capital Flows to Transition Economies: The Main Facts

This section presents the main stylized facts regarding the overall level, trends, and composition of net capital inflows to transition economies. By "net capital inflows" we mean all recorded net inflows that lead to a net liability vis-a-vis foreign residents.2 This includes all capital flows, both official and private, except those based on current transactions, transfers, and unrecorded flows (which would be classified as "errors and omissions"). Net capital inflows are disaggregated on the basis of the analytical presenta tion of the balance of payments suggested in the Balance of Payments Statistics Ye arbook 2000 (Part 1, p. xi), into five categories: (foreign) direct investment, portfolio investment, other investment, use of IMF credH and loans, and excep tional financing. Exceptional financing comprises debt forgiveness and arrears accumulation. Portfolio investment includes investment in both debt and equity securities below a certain threshold (usually 10-20 percent of total equity; above that threshold, the transaction would be considered a direct investment). "Other" investment is a broad category including trade credits, deposits, and loans. Direct investment and portfolio investment are virtually all private. Other investment is mainly private, but also includes non-IMF official loans. We ftrst present the main facts on the level and composition of capital flows for transition economies as a whole and their main subgroups, and then look at the cross-country distribution of drecti investment and portfolio flows.

2In the standard classification of the balance of payments (see IMF, 2000,Part I), this is equal to the balance on the financial account minus changes in reserve assets.

111

©International Monetary Fund. Not for Redistribution Pietro Garibaldi, Nada Mora, Ratna Sahay, and Jeromin Zettelmeyer

Level and Composition of Flows to Transition Economies

A broad comparison of capital flows to transition economies with flows to other emerging markets and developing countJ.ies is presented in Figure l. Even at this level of aggregation, we show flows to Russia separately, given the size of its economy (30-40 percent of the transition group, in terms of both GDP and popu lation) and the fact that it is quite atypical, as we shall see. Figure la shows that total net inflows to transition economies excluding Russia have been consistently above both the overall developing country average and the average to Latin America and Southeast Asia since 1995.3 Net flows to Russia have been consis tently smaller, reflecting in part Russia's current account surpluses over the period. The lower two panels of Figure I show the two main items of private capital flows tl1at are the focus of this paper: direct investment and portfolio investment. Direct investment exhibits a similar pattern across countJ.·y groups, with the excep tion of Russia. Not counting Russia, direct investment inflows into transition economies are high, somewhat higher than in other emerging markets, and followed a sharply rising trend (even as a percentage of GOP). In Russia, net direct investment inflows were much smaller, and direct investment in dollar levels fell in 1998 and 1999; however, this was more than offset by the sharp reduction in dollar GDP (by over 30 percent) resulting from the 1998 devaluation. Before 1997, portfolio investments to transition economies excluding Russia were generally below the international average. Since then, they have been about average or slightly above average-that is, transition economies other than Russia seem to have been somewhat less affected by the post-1997 contraction in emerging market financing. In the case of Russia, there were no inflows until 1995 followed by a sharp bubble that peaked in the first half of 1998 and that was followed by outflows after the August crisis. Since annual inflows for 1998 are an average of precrisis inflows and postcrisis outflows, tl1e peak shows in 1997 in the annual data. Moving to a more disaggregated picture of capital flows to transition economies, Figure 2 looks at total flows and the composition of flows for five groups of countries: Central and Eastern Europe ("CEE," comprising the Czech Republic, Hungary, Poland, Slovak Republic, and Slovenia); Southeastern Europe ("SEE," comprising Albania, Bulgaria, Romania, and Croatia), the three Baltic countries, Russia, and the remaining countJ.·ies of the former Soviet Union (other FSU). We show net inflows over the longest transition period consistent with stable group membership. Because of missing data, particularly in the early t:ran sition years, this leads to slightly different starting periods for each group.4

3By "total net capital inflows•· we mean all recorded net inflows that lead to a net liability vis-a-vis foreign residents. tn the standard classification of the balance of payments (see IMF. 2000) this is equal to the baJance on the fi nancial account minus net changes in reserve assets. This definition excludes unrecorded flows (which are part of "net errors and omissions") and capital transfers, but is otherwise very broad; in particular, it includes some forms of exceptional financing (such as arrearsa ccumuJation) as well as lending by the £MF and other official creditors. 4Because of lack of data, the ··other Fsu·· group excludes four countries: Georgia, Tajikistan, Turkmenistan. and Uzbekistan.

112

©International Monetary Fund. Not for Redistribution WHAT MOVES CAPITAL TO TRANSITION ECONOMIES?

Figure 1. World Region: Capital Flows, 1991-99 (in percent of GO!?annual data)

a. Total Capital Flows (net) 1 0

2 ··· ._ Developing 0

-2

-4

-6

-8 1991 93 95 97 99

b. Foreign Direct Investment (net) 5 r------�

2 ,, ,, ,,

,, ,, -z;,�;;, ica and ASEAN-4 _ , :: __..,.. , f?ussia -----·------· ------··----·-·-- ...... - - .... -...... -- -- ·---...... - --- - · 0 ��----�----��----�----�----�------�----�----� 1991 93 95 97 99

c. Portfolio Investment (net)

5 .------.

Russia Latin Americc1 wul ASEAN-4

-1 �----�------�----�----��----�----�------�----�----� 1991 93 95

Sources: IMF. World Economic Outlook Database 2000: IMF. lwemarional Financial Statistic.�·: and author�· calculations.

11J

©International Monetary Fund. Not for Redistribution Pietro Garibaldi. Nada Mora, Ratna Sahay, and Jeromin Zettelmeyer

However, the findings of the figure are quite robust to changes in the defmition of the time window; in particular, they would not be affected by either choosing the same period in calendar time (1995-99) or in transition time (say, 1993-97 for Central-Eastern and Southeastern Europe, and 1995-99 for the others).S Figure 2 shows stark differences in both the levels and the composition of capital flows to the region. The Baltic countries received by far the largest inflows as a percentage of GDP (and also in per capita terms) over the period. These inflows consisted mostly of direct investment and "other investment," in about equal proportion; the latter refers mainly to private lending to banks and corpora tions. Central-Eastern Europe, Southeastern Europe, and the other FSU (other than Russia) come second. In GDP terms, the other FSU received slightly higher total inflows, while Central-Eastern Europe received higher private inflows (the sum of direct investment, portfolio investment and other investment) than both Southeastern Europe and the other FSU. In both dollar and per capita terms, however (not shown), net inflows were much smaller in the other FSU, reflecting, in part, their lower levels of development. Note also the differences in the composition of flows across the three country groups. In Central-Eastern Europe, the main source of net inflows was direct investment, followed by sizable portfolio investment and other investment (mostly private lending). IMF financing was actually negative over the period, reflecting the repayment of loans prior to 1993, and there was almost no "exceptional financing." In Southeastern Europe, there were almost no net portfolio inflows, and both direct investment and other investment were important sources of private fi nancing. In the other countries of the former Soviet Union, finally, almost all private inflows took the form of direct investment, with very little private lending and almost no net portfolio investment (as we shall see in the next figure, there were some portfolio investment inflows prior to the 1998 crisis, but they are offset by outflows after the crisis). Instead, the bulk of nondirect investment financing was official lending, primarily by the IMF and the World Bank, and "exceptional financing," which comprises increases in external arrears (mainly to Russia and other energy exporters such as Turkmenistan) and debt reschedulings. Debt reschedulings also played an important role in Southeastern Europe. Figure 2 shows again that Russia is highly unusual, in several respects. First, it is unique in experiencing cumulative net outflows over 1994-99. Given its status as a large commodities exporter, this is not in itself surprising. The "errors and omissions" in Russia's balance of payments were consistently negative and large (since 1995, about 2.5 percent of GDP on average). If "errors and omissions," which often reflect some form of capital flight that have not been officially

5The source of the data in Figures 2 and 3 is the IMF's 2000 Balance of Pa yments Ye arbook (analyt ical presentation), or equivalently, the lmemational Financial Sratisrics. These data have the advantage of being consistently compiled, but the disadvantage of containing major gaps in coverage. For this reason, in the first (February 1999) draft of this paper. which considered annual data until 1997. data were obtained directly from lMF desk economists for the purposes of this overview. By the time we revised the paper, we felt that the coverage of the IFS had sufficiently improved to warrant switching to that source, which has the advantage of being public. This said, large gaps remain, particularly for the earlier transi tion years. These gaps are not an issue in Figures I b and I c and also in our regressions. which use longer series for direct and portfolio investments compiled from several sources (see below).

114

©International Monetary Fund. Not for Redistribution WHAT MOVES CAPITAL TO TRANSITION ECONOMIES?

Figure 2. Total Capital Flows (net), by Region (In percent of GOP In the same period)

a. Total Flows

10 -

8 -

6 -

4 -

2 -

-2

-4

-6

-8 Central-eastern South-eastern Baltics. 1993-99 Russia. 1994-99 Other FSU. 1995-99 Europe. 1993-99 Europe. 1993-99

b. Composition Flows 12

-2 = exceptional financing - usc of lMF credits and loans -4 - olher invesm1en1 = portfolio invesm1cnt -6 - direct investment -8 Central-Eastern South-Eastern Ballics, 1993-99 Russia. 1994-99 Other FSU, 1995-99 Europe. 1993-99 Europe. 1993-99

SourceS: IMF, 111/emational Fi nancial Statistics; and aulhors' calculations. Regional definitions: Ceocral-eastcrn Europe includes Czech Republic, Hungary, Poland, Slovak Republic, Slovenia); Southeastern Europe excludes Macedonia, Bosnia-Herzegovina and Yugoslavia due to data incompletness (i.e.. Albania, Bulgaria, Croatia, Romania): FSU stands for "former Soviet Union." In Figure 2 and 3 the "olher FSU" group comprises Arme11ia, Azerbaijan. Belarus, Kazakhstan. Kyrgyz Republic and Ukraine. Georgia. Tajikistan, Tu rkmenistan and Uzbekistan are excluded due to data incomplemess.

recorded, were to be included in the definition of capital fl ows, net capital outflows from Russia would be even larger. Second, and perhaps more surprising, foreign direct investment was comparatively small. On a cumulative basis, it was

115

©International Monetary Fund. Not for Redistribution Pietro Garibaldi. Nada Mora, Ratna Sahay, and Jeromin Zettelmeyer

overshadowed by substantial net portfolio inflows, large "exceptional financing" (debt reschedulings and increases in an·ears), and even IMF loans. However, all these financing items together were more than offs et by very large outflows of "other" investment, mostly of the private sector. This is usually interpreted as capital flight.6 Figure 3 presents the composition of flows at each point in time. Data for 199 J and J 992 for the Central-Eastern Europe and Southeastern Europe groups have been added since they were available for these groups. Direct investment as a share of GDP fo llowed an upward trend in all country groups except for Russia. Given the experience of positive growth in most transition economies following the initial "transitional recession," this implies a much sharper trend in dollar levels of direct investment, or in direct investment per capita. The figure also shows the "bubble" in portfolio investment in Russia, the "other FSU" countries and, to a lesser extent, Southeastern Europe prior to the 1998 crash. In 1998 itself, net port folio investment was still large and positive in Russia, as pre-crisis inflows were not completely offset by the post-August collapse, but already negative in the "other FSU" group. In 1999, however, it was negative in both cases. Figure 3 also illustrates the changing role of exceptional financing and IMP lending. With one exception-Bulgaria, which received large TMF financing in 1997 in connection wid1 a successful exchange-rate based stabilization program the role of IMF lending as a significant net source of financing in Central and Eastern Europe, Southeastern Europe, and Baltic countries was limited broadly to the flist years of transition (up to 1992 for Central and Eastern Europe, reflecting Polish debt restructuring, 1994 for Southeastern Europe, and 1995 for the Baltics). In Russia and the other countries of the former Soviet Union, however, it remained significant until 1998. "Exceptional financing" was nonexistent in the Baltics, and plays a role in Central and Eastern Europe only in the early years (up to, and including, 1993). Southeastern Europe is an intermediate case, with exceptional financing on a declining path, but still exceeding J percent of GDP on average until 1996, and remaining at close to 0.4 percent of GDP since then. In Russia and the other countries of the former Soviet Union, exceptional financing was impor tant throughout the period. While it seems to have been on a declining path until 1997, it jumped back in the aftermath of the Russian crisis of 1998.

Cross-Country Distribution of Direct Investment and Portfolio Flows

In this subsection we disagg.regate the two types of capital flows that are the subject of our econometric analysis in the next section-direct investment and portfolio investment. Unlike the previous subsection, which takes a net (inward plus outward) flows perspective, here we look at inward foreign direct investment and portfo1jo investment. We do this to simplify the interpretation of the results from the regression analyses, as reflecting the decisions of foreign investors. To practice, however, the inward flows and the net flows series are quite correlated, because the investments by transition economy residents abroad that were

6See Loungani and Mauro (2000).

116

©International Monetary Fund. Not for Redistribution WHAT MOVES CAPITAL TO TRANSITION ECONOMIES?

Figure 3. Composition of Capital Flows (net), by Region and Over Time (in percent of Annual GOP)

• exceptionaJ financing o IMFcred it o other investment • portfolio investment • direct investment

a. Central Eastern and Southeastern Europe: 1991-99 b. BaJtics: 1993-99 14 14 12 12 10 lO 8 8 6 6 4 4 2 0 -2 -4 -4 -6 -6 -8 -8 � -10 - t o L-w, -��-, 9-3 � �9-s -L--�n� --�w� 1992 94 96 98

c. Russia: 1994-99 d. Other FSU: 1995-99 to �------� 10 8 8 6 6 4 4 2 2 0 �----��--�� 0 -2 -2 -4 -4 -6 -6 -8 -8 -10 -10 -12 -12 4 L--�-- � ---L- �--L -� L___t__ _L __ .J_ __Jc..._-l.,.__ .J._ L.._ __ __, -1 � - -14 I 992 �� 1�2 � % •

e. Central-Eastern Europe: 1991-99 f. Southeastern Europe: 1993-99 14 12 10 8 6 4 2 0 -2 -4 -6 -8 -1 -10 IWI 93 95 97 w �

Notes: FSU denotes former Soviet Union. Due to data incompleteness for the earlier transition years, Macedonia is excluded from the Central-eastern Europe group in this figure, and data prior to 1993 excludes Croatia. For the same reason, Georgia, Taj ikistan, Turkmenistan and Uzbekistan are excluded from the "Other FSU" group.

117

©International Monetary Fund. Not for Redistribution Pietro Garibaldi, Nada Mora, Ratna Sahay. and Jeromln Zettelmeyer

Figure 4. Inward Foreign Direct Investment 1992-99, by Country

As a percentage of 1992-99 GOP Cumulative, in US$ Per Capita

Azerbaijan Hungary Estonia Czech Rcpu()lic Kazakhstan E.

Uzi>ekistan Tajikistan L___ _L ____ L___ _L __ __J 0 2 4 6 g 10 12 14 16 18 0 500 I 000 1500 2000

Sources: IMF, lntemcaional Financial Statistics; IMF. World Economic Outlook Database; and autbors' calculations. recorded as outward direct investment and portfolio investment flows are rela tively small. Note that even inward direct investment and portfolio investment are "net concepts" in the sense that they reflect the sum of net flows to transition economies with outflows due to the liquidation and repatriation of assets owned by foreign residents. In the period we are studying, this plays little role for inward direct investment, which is always positive or zero, but it does play an important role for inward portfolio investment, which is negative in some countries, in particular, due to outflows in 1998 and 1999 (see Figures 4 and 5, and the Appendix Tables in Garibaldi and others, 2002, for the complete series). In addition to inward portfolio investment from the ba lance of payments, we also show a related concept, namely, bond, equity, and Joan issues in international financial markets (Figure 6). The sum of bonds and equity overlaps with the defi nition of inward portfolio flows, but differs from this definition in two important respects. First, it only captures primary market issues, that is, it ignores any secondary market activity (and thus all outflows). Second, it ignores domestic

118

©International Monetary Fund. Not for Redistribution WHAT MOVES CAPITAL TO TRANSITION ECONOMIES?

Figure 5. Portfolio Investment Liabilities 1995-99, by Country

Cl 1995-97 portfolio investment liabilities • 1998-99 ponfolio investment liabilities

As a Percentage of 1995-99 GDP Cumulative, in US$ Per Capita

Estonia Solvcnia

Hungary Estonia C7.ech Republic Czech Republic Slovak Republic Hungary

Croatia Croatia

Lithuania Slovak Republic

Russia Lithuania

Solvenia Rus.•ia

Latvia Poland

Poland Latvia Kazakhstan Kazakhstan

Macedonia Romania

Georgia Macedonia Armenia Ukra me

Azerbaijan Am1eniu Albania Georgia

Taj ikistan Belarus

Turkmenistan Kyrgyz Republic Uzbekistan Azerbaijan Belarus Albania

Kyrgy.r. Republic Tajikistan Romania Turkmenistan Ukraine Uzbekistan

Bulgaria Moldova Moldova --'==t:::::..._.__ _._ _ _._-..J _ ___.__ _ _,______. ___, _ L Bulgaria .___ __,_ -4 -2 0 2 4 6 8 -200 0 200 -100 600 800

Sources: IMF. lntemational Financit1i Statistics: fMF. World Economic Outlook Database: and author�· calculations.

issues, which may, in part, be purchased by foreigners and thus give ri se to capital inflows (as was the case, for example, with the Russian GKO issues). The main reason for reporting bond, equity, and loan issues even though the concept misses important components of portfolio investment is that we know exactly what is reported, and that it is much less likely to contain errors than the direct investment and portfolio investment data. The latter are compiled from individual country authorities, which may have weak reporting systems, particularly in the early tran sition years. In contrast, international issues data are collected from the major financial centers and involve no serious reporting problems. Moreover, the gross international issue series receives considerable attention from the major financial markets. The data underlying Figures 4 and 5 (see Table Al in Garibaldi and others, 2002), is an extension of data published in the International Financial Statistics. Missing values, and in particular data for the earlier years, were filled in using information from IMP desk economists, the IMP's World Economic Outlook

119

©International Monetary Fund. Not for Redistribution Pietro Garibaldi, Nada Mora. Ratna Sahay, and Jeromln Zettelmeyer

Figure 6. Bond, Equity, and Loan Issues in International Markets 1992-99. by Country

Cll 992-99 bonds • 1992-99 equities 01992-99 Loans

As a Percentage of 1992-99 GDP Cumulative, in US$ Per Capita

Hungary Hungary Estonia Sol venia Slovak Republic Slovak Republic Thrkmenistan Estonia Lithuania Croatia Croatia Czech Republic Sol venia Lithuania Czech Republic Poland Kyrgyz Republic Russia Russia Latvia Latvia Thrkmcnbtan Katakhstan Romania Pot:wd Kazakhstan Moldova Kyrgyz Republic Romania Ukraine Tajikistan Moldova Ukraine Bulgaria Uzbekistan Macedonia Azerbaijan Uzbekistan Bulgaria Tajikhran Macedonia A7.erbaijao Georgia Georgia Albania Albania Amenia Amenia Belarus Betaros -2 0 2 4 6 8 0 500 I 000 1500 2000 2500 3000

Source: IMF BEL (Bonds. Equities, and Loans) database.

database and-for the early transition years of some Central European economies-data provided by Claessens, Oks, and Polastri (1998). The data underlying Figure 6 come from the IMP's Emerging Market Bonds, Equities and Loans database, which underlies the IMF's quarterly Emerging Market Financing. The main lesson of Figures 4-6 is that there is heterogeneity of inward direct investment and portfolio investment (we refer from now on only to inward invest ment), even within some of the broad country groups classified in the previous subsection. A look at the cross-country distribution of fo reign direct investment (Figure 4) yields some surprises. Few would have guessed that Azerbaijan is the direct investment leader in the transition group, by a large margin, when direct investment is measured as a percentage of GDP, or that Armenia edges out the Czech Republic and Poland as a recipient of direct investment as a shMe of GDP. In part, this truly reflects unusually high direct investment, for reasons which we will be exploring in the next subsection (for example, oil-related investments in

Azerbaijan), but it may also have to do with problems with measuring (dollar)

120

©International Monetary Fund. Not for Redistribution WHAT MOVES CAPITAL TO TRANSITION ECONOMIES?

GDP. It is well known that the initial output decline may have been exaggerated in several transition countries, particularly in the former Soviet Union.7 Perhaps more important, dollar GDP is computed using market exchange rates, which were probably strongly undervalued, particularly in the early transition years and in the countries of the former Soviet Union. 8 For these reasons, it may be preferable to use either PPP-adjusted GDP or population to scale direct investment flows. Figure 4 shows that the choice of normalization has a significant impact on the cross-country rankings. After scaling by population, the Central-Eastern European countries and the Baltic countries appear as clear djrect investment leaders. However, Azerbaijan and Kazakhstan are still leaders in the "other FSU" group, suggesting that natural resource endowments may be an important factor driving direct investment flows. The portfo)jo investment and bond, equity, and loan issue rankings (Figures 5 and 6) are relatively less sensitive to the choice of scale variable. In both cases, the Central-Eastern European countries, the Baltic countries, and Russia rank ahead of others. The main surprise of Figure 5 is that the position of Russia as a top recipient of net inward portfolio investment flows over 1995-99 is not as pronounced as Figures l-3 would perhaps have led one to believe. Trus is due to the fact that portfolio investment inflows to Russia were extraordinarily large during only one year, 1997, when they exceeded 4 percent of GDP. In 1995 and 1999, however, Russia suffered negative inward portfolio investment. In contrast, as Figure 3 showed, Central-Eastern Europe received much more sustained inflows. Similar observations apply to the bond, equity, and loan issue data.

II. What Moved Capital to Transition Economies? Econometric Results

Using panel data for 26 transition economies ranging from the early transition years (1990-92, depending on the country) to 1999, we explore how well poten tial explanatory variables account for the behavior of inward djrect investment and inward portfolio investment flows. We chose these two series for two reasons: first, because they are important from a policy perspective (for example, because of their potential impact on growth and structural transformation) and second, because we have a good idea of what asset classes they contain. That is less clear in the case of wider flow concepts. The class "other investment," in particular, mixes a wide range of different types of flows, from lending by foreign banks to domestic capital flight, wruch makes it harder to interpret regression results. For several reasons, we decided not to include in the regression analysis the bond, equities, and loans issue series, which was discussed at the end of the previous subsection. While we know with great precision (asset by asset) what is included in this series, it would not be easy to interpret the results because primary

7See, among others. Berg ( 1993), Gavrilenkov and Koen ( 1994), and Kaufmann and Kaliberda (1996). ssee Halpern and Wyplosz (1997) and Krajnyak and Zettelmeyer (1998).

121

©International Monetary Fund. Not for Redistribution Pietro Garibaldi, Nada Mora, Ratna Sahay, and Jeromln Zettelmeyer

issue flows mix decisions about the demand for capital-i.e., the decision to issue bonds, equities, or loans on the part of domestic authorities and residents-and decisions by foreign investors to supply capital. A similar identification problem exists for portfolio investment flows, but to the extent that these reflect inflows and outflows into an existing asset stock that is traded on secondary markets-as opposed to primary issues-it is not as severe a problem. At the limit, with large domestic asset markets, the supply of domestic portfolio assets would be elastic, enabling us to perfectly identify foreign demand for assets-i.e., to interpret the coefficients as reflecting foreign investor behavior.9 In the next subsection, we briefly review some methodological issues, including model selection and the selection of an appropriate scale variable (GOP or population). We then discuss a set of candidate explanatory variables for the regression analyses, and then we present our findings.

Methodology

In view of the large set of (possibly correlated) potential determinants of capital flows to transition econonues, a challenge for any econometric exercise is to maintain a reasonable degree of parsimony while avoiding misspecification of the model. To deal with this, we used a general-to-specific model selection approach in the context of a dynamic panel of the form:

Yu = ao + aiYi,l-1 + a(L)Xu + �Z; + E;." where y;,1 is the dependent variable (either direct investment or portfolio invest ment: see below), Xu denotes a vector of time-varying variables, a(L) denotes a lag polynomial of coefficients, Z; is a vector of initial conditions and time-invariant controls (e.g. regional dummies), and E;,, is the enor term. The time index denotes years since the beginning of transition, while the index i refers to countries. The advantage of this approach is that it enables us to "test down" among a reasonably large set of potential right-hand-side variables in the context of a flexible dynamic specification. This minimizes the chances of bias due to omitted variables and inappropriate dynamic restrictions. In practice, the initial lag length was limited to one or two lags, depending on the variable, given the shortness of our time series. The presence of a lagged dependent vruiable precludes the use of the standard fixed or random effects estimators.10 We thus used OLS estimates, which are consistent and efficient only if the model does not contain unobserved country specific effects. This means that the specification we start out with must be suffi ciently general to contain all country-specific effects that ru·e correlated with other

9This said, secondary market volume was small relalive to new issues in limes when markelS were rapidly expanding, such as in Russia prior to the 1998 crisis. For example. the decision to issue GKOs in 1997 was surely driven by variables that also influenced asset demand (e.g., the fiscal balance as a solvency indicator). We wilJreturn to this point when we interpret the results. 1°Fixed and random effects rely on a strong exogeneity assumption that is automatically violated in dynamic panels; see Keane and Runkle ( 1992).

122

©International Monetary Fund. Not for Redistribution WHAT MOVES CAPITAL TO TRANSITION ECONOMIES?

right-hand-side variables, a further reason to start out with a very broad initial specification. The main drawback of our approach, as in any general-to-specific exercise, is the "path dependency" problem-the fact that the parsimonious specification obtained may be sensitive to the order of elimination, which in turn reflects the researcher's priors in giving some variables a relatively better chance to remain in the model (by ordering them last). It turns out that in this san1ple there is very little path dependency, given our initial model. To confirm this, a longer version of this paper (Garibaldi and others, 2002) supplements the approach described above with a version of Leamer's (1983) "Extreme Bounds Analysis," which tests the robustness of the main explanatory variables surviving in our parsimonious models. The models shown below pass these robu stness tests except for caveats pointed out in the text. A final methodological issue concerns the choice of the variable used to normalize capital flows-that is, dollar GOP, PPP-adjusted dollar GOP, or popu lation. As we saw in the previous subsection, this bas a significant impact on the relative levels of direct investment and (to a lesser extent) portfol io investment across countries. The standard approach is to normalize using GDP evaluated at market exchange rates, but this could introduce distortions due to the large fluctu ations of both exchange rates and real output in our sample. Our preferred approach is thus not to impose any particular normalization on the data, but instead to use unsealed capital flow data (i.e. simply measured in millions of dollars), while controlling for several alternative scale variables on the right-hand side. The scale variable that appears to be most relevant from the perspective of predicting capital flows can then be selected as part of the broader general-to-specific procedure. II

Explanatory Va riables

We divide the potential determinants of capital inflows, beyond the scalevari ables discussed in the previous subsection, into three groups: those that are likely to affect both direct investment and portfolio investment, those that are specific to direct investment, and those that are specific to portfolio investment.

111f both the capital now data and the scale variable enter the regression in natural logs. this encom passes the standard approach (which is equivalem 10 imposing a coefticienl of unity on the relationship between log flows and log GOP) as a special case. ln any case, log-transforming the data is appropriate if one believes that the true structural model is multiplicative-which is just saying that a unit increase in a right-band-side variable would always lead to some percemage increase in capital flows per capita, rather than some absolute increase. The problem is that the log transfonnation cannot be directly applied when there are zeros in the data. To deal with this problem, we transfom1ed the DI data by taking the natUral log of I + Dl. However, with portfolio investment, which contains large negative values. we worked with untransformed data. and ran all regressions using both a popularion normalization and a GOP normalizati on.

120

©International Monetary Fund. Not for Redistribution Pietro Garibaldi, Nada Mora, Ratna Sahay, and Jeromin Zettelmeyer

General determinants

The basic variables of macroeconomic performance are inflation, the fiscal balance, and (lagged) growth. Macroeconomic stability is widely viewed by po]j cymakers, academics, and lhe press as creating a conducive environment for foreign investors, so one expects inflation to have a negative effect on capital inflows. The expected signs on the fiscal balance and growth are more ambiguous. One view of the fiscal balance is as an alternative stabilization proxy (Fischer, Sabay, and Vegb, 1997); this would imply a positive sign (surpluses are better than deficits fTom the point of view of encouraging investment). However, a large deficit might also proxy a budgetary need for foreign financing-particularly if inflation bas been adequately controlled for but not interest rates-which would suggest a negative relationship between the fiscal balance and portfolio inflows. Similarly, the prima fa cie expectation on the relationship between output growth and capital flows is probably that investors prefer entering an economy that has already turned the comer. This could be justified if profitability is related to domestic demand, or if output declines are associated with disruptions (including political and social) that are not controlled for through other variables. On the other hand, the output decline could be viewed as increasing the marginal product of new capital (by freeing resources held in the traditional sectors) and perhaps as lowering asset prices. indicators of the exchange rate regime include whether or not a prean11ounced exchange rate peg was in place, and whether the country had unified the exchange rate or applied restrictions to convertibility that resulted in multiple exchange rate practices. To the extent that a peg credibly reduces exchange rate risk, it might attract higher inflows. Multiple exchange rates and restrictions to convertibilityare typically believed to deter inflows, as they may create obstacles both for the repa triation of profits and for the import of intermediate goods that often go hand-in hand with direct investment projects. We used an indicator of liberalization and privatization, compiled by De Melo, Denizer, and Gelb (1996, 1997), and updated using similar indices compiled by the EBRD. The De Melo, Denizer, and Gelb index is an average of three subindices, which capture progress in internal (price) liberalization, external (trade) liberalization, and privatization. respectively. This reflects the view that countries that are further along in implementing structural reforms and privatizing will find it easier to attract foreign capital. Indicators of institutional quality and the legal framework are included to capture the argument that institutional and legal shortcomings-including unpre dictable and burdensome regulation, red tape, confiscatory taxation, and difficul ties in enforcing contracts-are important impediments to private business in general, and particularly to foreign investments in transition countries. We rely on a set of 5 institutional indicators compiled by the World Bank in their 1997 survey of perceptions of the qua]jty of governance (see Appendix for details). Initial conditions and other controls are included to capture potential deter minants that are unrelated to policies during the sample period, including natural resource endowment, location (e.g. distance from Western Europe), pretransition

124

©International Monetary Fund. Not for Redistribution WHAT MOVES CAPITAL TO TRANSITION ECONOMIES?

liberalization efforts, and a dummy for wars. Abundant natural resources should encourage inflows: location will affect transportation costs to export markets, and initial liberalization may have an effect (not just the most recent level of reforms might matter, but also how long they have been in place). The expected sign on the war dummy is not entirely obvious: wars should discourage inflows, but may also attract inflows to the extent that there is a Large nonresident community that helps to finance the war effort (Armenia and Croatia might be examples). We also included the three-month U.S. treasury bill rate as a basic opportunity cost variable. A final question is whether or not to control separately for market (investor) perceptions, as captured by "country risk" ratings, which are regularly published in investor magazines such as Euromoney or Institutional Investor. As we explain in more detail in the Appendix, these are overall indicators of investor perceptions that give about equal weight to a country's economic and political/institutional "performance." To the extent that we are already controlling for these factors, they should not matter. Nevertheless, we decided to include market perceptions among our regressors, for two reasons. First, whether or not perceptions matter even in the presence of variables that directly control for "fundamentals" is in itself an interesting question. More important, some institutional indicators are only avail able for a subset of our sample. Using market perception indicators to control for unmeasured institutional factors enables us to use a larger sample. Thus, current and lagged values of market perceptions are generally included in the regressions reported below. To make the coefficientsof the regression easier to interpret, we do not include the investor magazine scores directly into the regression, but rather we use the residuals of a first-stage regression of these scores on the remaining right-hand-side explanatory variables that were included in the regression. Thus, the coefficient on ow· perceptions variable represents the effect on direct investment of market perceptions that are not attributable to fa ctors that we are directly controllingfor in the regression model.

Specific determinants of direct investment

Beyond the variables discussed so far, there are four sets of variables that could be relevant specifically for direct investment. Competitiveness indicators would ideally comprise a measure such as (dollar) unit labor cost in manufacturing. However, while average wage data in manufac turing are generally available, indirect labor costs and productivity data for the manufacturing sector are much harder to come by on a consistent cross-country basis. Rather than attempting to compute unit labor costs, we separately include average monthly dollar wages. Since we control for aggregate productivity differ ences across countries by including both population and GDP measures as "scale variables" in the regression (see last subsection above), our specification encom passes aggregate unit labor costs (except for indirect labor costs). Trade liberalization. Direct investment often takes place in businesses that either are export-oriented or require imported inputs, or both. fn such cases, one would expect that the more a country's trade regime has been liberalized the

125

©International Monetary Fund. Not for Redistribution Pietro Garibaldi. Nada Mora, Ratna Sahay, and Jeromin Zettelmeyer

greater the direct investment. One indicator capturing trade liberalization is the EBRD's index of external liberalization. Restrictions to direct investment. The extent to which legal barriers that make foreign direct investment more difficult or more costly varies considerably across countries. We attempted to construct an index that captures these barriers (see Appendix for details). Method of privatization. While the extent of privatization is controlled for, at least rudimentarily, by the De Melo, Denizer, and Gelb liberalization index (see above), this disregards the potentially important effect on direct investment of how a country privatized. Direct sales with equal access by fo reigners may offer an automatic opportunity for direct investment, while insider privatizations (privati zation as a sale or gifts to the management and/or workers of a previously state owned enterprise) may create barriers.12 To capture this notion, three indicators of the privatization method based on EBRD information are included in the regres sion (see Appendix for details about the construction of these variables).

Specific determinants of portfolio investment

Securities market development. Countries that have more advanced securities markets and are equipped with a regulatory and legal framework specific to dealing with such transactions would be expected to receive higher flows of port fo lio investment. One indicator capturing the extent of institutional development in this area is the EBRD transition indicator on securities markets and nonbank financial institutions (see next subsection for more details). Restrictions on portfolio investment. This is analogous to the index on direct investment restrictions. The construction of an index of portfolio investment restrictions is described in the Appendix. Indicators of default risk. The total external debt stock and international reserves are included as default risk indicators. Both are lagged one year, to reflect a perceptions Jag and avoid endogeneity. In addition, we separately control for the ratio of short-term debt in total debt. One would expect that countries with lower short-term debt and higher reserves are less prone to default and hence attract higher portfolio inflows. Treasury bill rares. To the extent that we fully control for the factors which determine default and exchange rate risk, portfolio flows should go to countries that offer relatively higher real interest rates on their government paper. However, this effect will be attenuated or reversed to the extent that risk factors have been omitted, since in this case high interest rates may themselves proxy for these omitted factors. We used the 3-month treasury bill rate from the IMF's International Financial Statistics (IPS), and filled the gaps as much as possible with similar short-term rates from country publications and IMF desk economists. However, there still remained a substantial set of missing observations, which are typically due to countries and/or years in which treasury bill markets were nonex-

12See Holland and Pain (1998).

126

©International Monetary Fund. Not for Redistribution WHAT MOVES CAPITAL TO TRANSITION ECONOMIES?

istent or in mdimentary stages. To avoid restricting the sample size, we replaced these missing values by zero entries and at the same time defined a dummy that takes the value of one when data are available and zero otherwise.13 All the data used as explanatory variables, including sources and the notation used in the remainder of the paper, are summari zed in Table l. In addition, the Appendix gives details on the definition and compilation of those explanatory variables that are new or unusual in the context of a study of determinants of capital flows to transition economies. These include (1) the indices of restrictions on direct investment and portfolio flows that we construct for the purposes of this study; (2) indices of country-risk perceptions; (3) institutional, political, and legal proxies, particularly those compiled by the World Bank in their 1997 survey; and (4) indicators of the privatization method, which were also constructed for this study. Our sample periods are 1990-99 for Hungary and Poland; L 991-99 for Bulgaria, Romania; the Czech Republic, and the Slovak Republic; and 1992-99 for the 19 remaining transition economies (Albania, Croatia, FYR Macedonia, and Slovenia, the 3 Baltic countries. Russia, and 11 other countries of the former Soviet Union). Thls gives a potential data set of 208 observations. However, the use of lagged variables further restricts the sample period and not all data series were available for all these countries and periods. In practice, our actual sample varied between 143 and 179 observations.

Regression Results

Results for direct investment regressions

Regression results for foreign direct investment are shown in Table 2. Model l is derived by applying a general-to-specific selection process to a general model containing aU potential explanatory vruiables described above under the rubrics "general determinants" and "specific detennlnants of direct investment," except for the institutional variables from the World Bank, which are not available for 5 of the 25 economies we study. First lags (and in the case of the liberalization index, second lags as well) were included for the time-varying variables. This brought the total number of variables in the general model to 32. In columns (2), (3), and (4), we experiment with adding regional dummies and (at the cost of using a smaller sample) the World Bank institutional indicators to Model 1. Model 5, fm ally, shows a parsimonious model derived from a general model that also includes the World Bank institutional indicators, and a dummy variable for Russia, using the smaller sample.14

J3The priors on the sign of this dummy are unclear. For example. if coumries with missing 1reasury bill rates behave "as il" Lhey had positive rates, then the dummy should have a negative sign: if Lhey behave as if they had negative rates, then the dummy should have a positive sign. 14 This smaller sample excludes Croatia, Romania. Slovenia. Tajikistan, and 1\Jrkmenistan.

127

©International Monetary Fund. Not for Redistribution Pietro Garibaldi. Nada Mora, Ratna Sahay, and Jeromin Zettelmeyer

Table 1. Names and Definitions of Explanatory Variables

Variable Name Definition Source

Scale variables /pop Nmural log of population World Economic Outlook database. IMP. lgdp Natural log of GDP in U.S. dollars World Economic Outlook database.

lMF. /ppp Natural log of PPP-adjusted GDP World Economic Owlook database, IMF.

Macroeconomic growth Pirl>l lag of real GDP growth. World Economic Otulook darabase, lMF. linfav Natural log of average annual current inflation World Economic Outlook database,

lMF. err Prcannounced exchange rate regime dummy. Annual Report on E.xcha11ge

Arrangemems and Resrrictions. IMF. mcp Multiple exchange cates. Annunl Report on Exchange Arrangemems and Resrrictions. IMF. jbal General government balance as percent of GDP. Wo rld Economic Outlook database, lMF.

Struclllral Rejom1 Va riables li Liberalization index De Melo, Denizer and Gelb (1996) and Range from 0 to I, where I de notes full extended for 1996 and 1997 based on liberalization. the EBRD indices. Refer to Berg. Borensztein, Sahay, and Zenelmeyer (1999) for more details.

lnstitlllional quality and the legal framework wbpred Predictability of laws and policies World Bank, Wo rld Development Report 1997 Survey. wbprop Political stability and security of property wbgovbus Overall government-business interface wbredtape Bureaucratic red tape wbgoveff Efficiency of government in providing services (all range from .I to 6, 6 is worst)

In itial Conditions and Other Conrrols war War dummy. Authors' calculations. ldist Naturdl log of distance fr om Duesseldorf (in km).Intemet map locator. natrr Natural resource abundance dummy variable Authors' calculations. li89 Initial liberalization index. De Melo. Denizer and Gelb ( 1996). cro. rus, rkm Dummy variables for Croatia. Russia. and Thrkmenistan respectevely. bal Dummy variable for Baltic countries.

US3m U.S. 3-monlhtreasury bill rate. lmemational Firumcial Statistics, lMF.

Market Pe rception Va riable em_res Residual of Euromoney country risk rating in a Euromoney. March and September issues regression on fundantentals. &.ee text. various years.

Specific to FDI

!wage Natural log of average monthly manufacturing OECD Short-term Economic Indicators. wages in US$. ILO, and JMF. ebtrde Trade and Foreign Exchange Index; range European Bank for Reconstruction and from I to 4. where 4 denotes compiU'3.blc Development (EBRD). extended prior to standards to advanced economies. 1994 based on Ll described above.

128

©International Monetary Fund. Not for Redistribution WHAT MOVES CAPITAL TO TRANSITION ECONOMIES?

Table 1. (concluded)

Vari able Name Definition Source

diir Foreign direct invesLment restrictions inde)(. Aut.hors' ca lculations based on Atmual Range from -0.2 to 6 where 6 reflects most Repon on Exchange Arrangements and restri ctions. Resrricrions. JMF. various issues. prsal, pn·o11, The Lhreeprivatization indicators: privatization Aut.bors· calculations based on E.BRD prins by direct sale, by voucher. and in sider Transition Repons. Refer to Appendix respectively for detailed description of methodology. Specific 10 Portfolio lnvestmem ebsecse Securities market index European Bank for Recomtruction and Range rrom I to 4, where 4 denotes securities Developmem (EBRD) and exrended laws and regulations approaching IOSCO prior to 1994 based on li described standards; substantial market liquidiyt and above. capitalization: well-functioning nonbank financial institutions and effective regulation. pir Portfolio investmem restrictions index. Authors· calcularions based on Annual Potential range from 0 to 2. where 2 indicates Repon on Exchange Arrangemems and outright prohibition of portfolio llows. Restrictions. IMF, various issues. rTbiliO and rTbil/0 is the real treasury bill rate, where lntemariona/ Financial Statistics series Tbilld missing observations are replaced with a 0. 60c, �upplemented by IMF staff and Tbilld variable isa dummy that rakes on the country reportS value of I in the cases that rate� are available. debtpc External debt per capita Wo rld Economic Outlook database, LMF. respc international reserves per capita, respectively World Economic Outlook database. IMP. srdrd ratio of short-term debt to total debt World Economic Outlook database. lMF.

The main results are as fo llows. Consider first Model l. Note the rather good fit Uust over 0.9). The coefficients mostly confirm conventional priors. Direct investment flows increase with good macroeconomic performance, as measured by growth and a high fiscal balance, the state of economic liberalization (li), and reforms in the trade area (Debtrde, which measures the change in the EBRD trade reform index). Insider p1ivatization (prim;) discourages direct investment, as do direct restrictions on direct investment inflows (diir-1). As expected, countries rich in natural resources (natrr) received more direct investment�n average, almost 60 percent compared to otherwise similar resource-poor countries. Having an ex ante commitment to a fixed exchange rate (err2) is also positively associated with direct investment. Finally, note that the dummy dum.O for years in which there was zero direct investment is also highly significant and negative; this suggests some sort of start-up cost to inflows. Three aspects of Model l 's predictions are unexpected. First, current inflation (linfav) comes in with the "wrong" sign. We have no explanation for this, except potential endogeneity, which may be an issue in spite of our efforts to avoid omitted

129

©International Monetary Fund. Not for Redistribution (..> Ta ble 2. Regression Results for Foreign Direct Investment 0 (dependent variable: natural log of (1 + inward foreign direct investment in millions of US$), OLSestimates) -u

©International Monetary Fund. Not for Redistribution WHAT MOVES CAPITAL TO TRANSITION ECONOMIES?

variable problems.15 The result goes away (in the sense that contemporaneous inflation is not significant and can be eliminated from the model) when the general to-specific exercise is repeated on a smaller sample that includes the World Bank institutional variables (see Model 5). Second, after controlling for aggregate prod uctivity-which we do indirectly, by having both log GOP and log population in the model-wages are insignificant and can be eliminated. Thus, in this broad sample, 16 wage costs seem to be overshadowed by more fundamental macroeco nomic stability and governance issues. Third, the coefficient on log GDP is only about 0.3 and the sum of the coefficients on log population and log GOP is only 0.5-that is, far below unity. This implies that, even controlling for all other char acteristics that we are able to quantify, small countries were much better at attracting direct investment as a share of GDP. A doubling of country size, keeping everything else constant, is associated with an increase in direct investment by only about 50 percent. One question is the extent to which this result is driven by Russia-a very large country attracting relatively little direct investment. This question is answered in Model 2, which includes a country dummy for Russia and two other countries that tum out to be outliers: Croatia, which received significantly more direct investment than Model 1 would predict, and Turkmenistan, which received significantly less.n The coefficient on the Russia dummy is particularly large in absolute terms, and highly significant. Note that when the dummies are included in the model, the coefficient on the scale variables rises (from about 0.5 to 0.65), but remains significantly smaller than unity. Which omitted variables might be driving the three significant country outliers? For Croatia, war-related inflows from the Croatian community abroad may play a role. For Russia and Turkmenistan, plausible candidates are variables related to the business clin1ate and governance. This seems to be supported by the fact that the coefficient on the market perceptions residual (emres_di), which captures investor perceptions not attributable to differences in the remaining right band-side variables, drops in the presence of the three country dummies. The next step is to directly add governance measures to the model. To do that, we first rerun Model 2 on a smaller sample-which excludes Croatia and Turkmenistan-for which the 1997 World Bank governance measures are avail able (Model 3). The results are very close to those on the whole sample. We then add the five governance indicators. As it turns out, four of them are jointly and individually insignificant and can be eliminated, leaving just one, wbredtape, an index for bureaucratic red tape and con-uption (Model 4). As expected, the coeffi cient on emres_di falls further in the presence of wbredtape. and loses statistical significance. Surprisingly, however, tl1e coefficient on the Russia dummy is not

ISAssuming the model is otherwise well-specified. reverse causality is not a major concern for vari ables such as innation, since Dl constitutes a relatively narrow capital now item that is unlikely to have large simultaneous macroeconomic effects. l61n a sample that comprises only the more advanced transition countries in Central and Eastern Europe, Bevan and Estrin (2000) and Holland and Pain ( 1998) do find the expected negative relationship between labor costs and direct investment. l1These are the only robust outliers, in the sense that they remain significant when any other country dummy is added, and that any other counrry dummy is insignificant when added to the world.

1.31

©International Monetary Fund. Not for Redistribution Pietro Garibaldi. Nada Mora. Ratna Sahay. and Jeromin Zettelmeyer

reduced-it even rises after the inclusion of wbredtape into the model. Thus, the puzzle of why Russia received so little direct investment is nor resolved by adding this set of governance indicators to the model. Finally, we check whether the results so far are the same if the governance variables are included at the beginning in the model selection process, rather than added at a later stage. The answer is a qualified yes. Model 5 confirms the basic story about direct investment being attracted mainly by a stable macroeconomic environment, economic reforms, the privatization method, and the presence of natural resources. although there are some differences in the details.18 As far as the role of governance is concerned, the results mirror those of Model 4: wbredrape is again the only significant governance variable. and the Russia dummy stays highly significant in spite of its presence. The main surprise in Model 5 is the continued significance of emres_di as a predictor of direct investment flows, which represents any information conrained in investor ratings that is not captured by the remaining right-hand-side funda mentals including the World Bank governance variables. There are several poten tial explanations for this result. The most obvious one is that to the extent that the right-hand-side variables still do not give a full picture of the country's funda mentals, these might be captured by emres_di. However, after controlling f'or the governance variables, it is not clear in what respect the information set upon which investor magazine ratings are based could be richer than the one we are already controlling for (see the Appendix on how these ratings are constructed). Two other related explanations that we find more plausible are as fo llows. The first possibility has to do with reporting lags and errors in rel1ec:ting economic fundamentals. Not all the information that is contained on our right hand side (in particular, on governance) was available when investors made their decisions. Suppose that investors use investor magazine scores (which were avail able twice a year since the early I 990s for most countries) a� a proxy for the missing information. Ex post, these scores may turn out to have been wrong. in the sense that they either overstated or understated the quality of fundamentals in a particular country. But in the meantime, investment decisions may have been taken, which could be costly to reverse. Thus. if investors regard investor maga zine scores as a "coincident indicator'' of the quality of fundamentals. then the error component of this indicator-i.e., precisely the residual used in Model 5- should have an impact on investor decisions. The second story relates to the problem of endogeneity. in the sense of simple reverse causality. In other words, high investor ratings could be driven by high observed inflows, rather than the reverse. This, however. is implausible here, since it is lagged investor ratings that matter in Model 5, and since the regression abo controls for past inward direct investment directly. However. there may be a second, more plausible source of endogeneity-namely, that both investor ratings and capital intlows are to some extent driven by fads or biases, that is. by market

ISMainly. it is the changes in the general liberalization index, Dli. that seem to matter: change� in the specific trade liberalization index are insignilicant and cun be eliminated. and the impact of priv;uitati<>n enters not only through an adverse effect of insider privatization, b�n also via a hencliciJI effect nf diri!Ct sales (prsal). The puzzling result is that the contemporaneous ef'fect nf inll:uion i� i1hignilic:111t.

1.32

©International Monetary Fund. Not for Redistribution WHAT MOVES CAPITAL TO TRANSITION ECONOMIES?

perceptions that are not really justified by fundamentals. Since these are not directly included on the right-hand side of om regression models, they show up in the error term, which ends up being correlated with the investor score residual. emres_di. In principle, it should be possible to test this possibility by instru menting emres_di: in practice, however, instruments that are both valid and suffi ciently powerful are hard to find in this context.l9 Regardless of which of these two stories is ri ght, Model 5 suggests that investor ratings contain predictive power for direct investment flows beyond the information about fundamentals which they embody. This may be either because they are used as imperfect proxies for economic fundamentals when direct infor mation on these fundamentals is not readily available. or because they reflect market perceptions that are not attributable to fundamentals but nevertheless have an impact on investment decisions.

Results for portfolio investment regressions

We now tum to a set of regressions wbkh explore the determinants of inward portfolio investment (Table 3). As explained before, negative inward investment flows-reflecting the liquidation of positions held by foreign residents-is a frequent occwTence in our data. This precludes log-transforming the data in the way we clid in the clirect investment regression, and forces us to make a decision in advance on whether to run the regression in terms of investment per capita terms or investment per GOP. We did both, while additionally controlling for the scale variable that was not chosen for normalizing the flows. As it turns out, the popu lation normalization fits the data better, which is not surprising given the problems with measuring dollar GDP discussed above. This is the normalization used in the regressions presented below. However, the conclusions of this section would not be affected if we had instead based the cliscussion on the GOP normalization. Model I was derived by simplifying a general model containing a total of 35 variables, which include all general and portfolio investment-specific potential explanatory variables discussed in the preceding subsection, with lags where appro priate, except for the institutional indicators, which are not available for several countries. Path dependency was a somewhat greater problem than in the direct investment regressions. ln particular, we were not able to discriminate between the effects of lagged inflation and liberalization on pmtfolio investment, in the sense that the order of elimination determines which variable ends up in the parsimonious model. Thus, the variable "llnf-1" in Model 1 should be viewed as representative of the quality of macroeconomic policies, rather than as literally representing infla tion. The remaining steps, however, go through regardless of whether the lagged inflation or liberalization is included in the model. For this reason, we do not sepa rately show regressions results that include LI rather than l/nf-1.

19We tri ed using the EBRD"s index for the development of domestic financial market!; as an instru ment for emres_di-which arguably was not a detenninant of direct investment, since the latter occurred mostly through direct purchases in the comext of privatization or greenfield investmenl. and as such did not require organized stock markets. Using this instrument, a Hausman test could not reject the hypoth esis of no misspecification.

133

©International Monetary Fund. Not for Redistribution --"- (.) Table 3. Regression Results for Portfolio Investment � (dependent variable: portfolio investment liabilitiesper capita. OLS estimates) -o

(j)"-+ I (2) 3 5 !) {) {4) {) 0 Va riable Coefficient H'1llue Coefficiem l-\'1!lue Coefficient H'alue Coefficient r- ,'alue Coefficient r-,"'31ue G) Q 430 229.88 Consrant 58.1 1 2.18 64.99 2.25 200.84 3.49 217.59 4.55 fj lagged dependent 0.25 3.33 035 4.24 0.28 3.35 0.20 2.37 0. 19 1.38 �0 IJ1flation fin/cn'--1 -7.45 -1.89 --6.89 -1.60 -6.25 -1.48 z Exchange rate regime err2 18.12 19.43 2.16 25.60 2.65 z.oo 0 Market perceptions emres_pi " 2.04 3.75 L71 2.81 1.34 2.20 0. 0 - U.S. interest rate US 3m -1 1.79 -2. 18 -13.346 -228 -10.85 -1.87 -8.48 1 .57 �.82 -1.65 $ Specific 26.12 3.91 24.33 3.64 0 ebsecse p portfo Jio im·esunent Debsecse 4:!.33 2.66 43.199 2.73 33.49 2.ll ;v determinants respc-1 0.07 4.43 O.o38 1.91 0.01 0.6/ 0-+ WB -31.82 -48. 17 -49.94 -5.32 :::> gO \'emance k'bprop -2.71 -5. 10 0 dummy variable rus97 91.90 /.92 en 0 Regression R1 0.41 0.43 0.46 0.45 0.47 J 0 statistics N 177 143 143 143 143 � 6 0 k 8 8 9 5 :::> 0. afor Models (1) through (3): first lag of residual from a regression of the Euromoney country rating (see appendix) on the other potential right-hand-side deter- L

©International Monetary Fund. Not for Redistribution WHAT MOVES CAPITAL TO TRANSITION ECONOMIES?

Given our findings for direct investment, the results of Model (I) are striking in two respects. First, the fit is much worse than the fitof the corresponding direct investment regression, with an R2 of approximately 0.4 as opposed to approxi mately 0.9. Part of this large difference is explained by the absence of a log trans fmmation in Table 3, which dampened outliers and contributed to the overall good fit of the direct investment regressions. To get a sense of the importance of this effect, we re-ran Model 1 of Table 2 without the log transformation, using popu lation for the purposes of normalization. The fit declines to 0.67. which is still much larger than that of the portfolio investment regressions. Thus, the basic insight is that differences in portfolio investment over time and across countries are much harder to model, even using a very rich set of determinants, than direct investment tlows. The second striking finding in Table 3, which is related to the first, is the rela tively small number of explanatory variables that seem to play any role at all. Model 1 says that-ignoring governance indicators as a potential determinant, portfolio investment flows seem to have a systematic relationship only with past inflation (perhaps as a proxy for the quality of macroeconomic "housekeeping"), investor ratings, the exchange rate regime (where pegs are associated with bigger net inflows), the level of reserves, world interest rates, and improvements in the securities market infrastructure (Debsecse). This list is remarkable mainly for what it does not contain: solvency indicators such as the level of debt, economic growth, economic reforms as distinct from macro stability, domestic real interest rates, and direct legal restrictions to portfolio inflows. The paucity of significant explanatory variables in the portfo)jo investment regression can be given several interpretations. Poor measurement may be a problem, both in the sense that some variables could be rnismeasured (for example, debt and domestic real interest rates), and in the sense that we have not fully measured default risk directly, so that the domestic interest rate plays a role both as a proxy for country risk and as a "pull" factor attracting capital flows, with ambiguous net effects. The identification problem discussed at the beginning of Section II almost sw·ely plays a role as well, to the extent that the same variables enter the demand and supply equations for portfolio assets with opposite signs. For example, all things being equal, one would expect the supply of government bonds to increase with an increase in the fiscal deficit but the demand for bonds to decrease; these effects may offset each other in the estimation. Finally, our results may indicate that portfolio inflows in transition economies were not very sensitive to solvency indicators apart from the level of international reserves. The fact that direct restrictions to portfolio movements do not have any explanatory power either indicates measurement errors, or suggests that portfolio flows can find a way around these restrictions. Model 2 reruns Model 1 on a smaller sample that includes only the countries for which the World Bank institutional indicators are available. The results are very close, so the reduction of the sample itself does not have a major impact. Based on the reduced sample, Model 3 shows the consequences of adding the World Bank governance indicators to the model. Of the five indices, four have almost no explanatory power, and can be elimjnated from the model at extremely

1:35

©International Monetary Fund. Not for Redistribution Pietro Garibaldi. Nada Mora. Ratna Sahay. and Jeromin Zettelmeyer

high significance levels (p = 0.97). Only one variable-the indicator of respect for property rights, wbprop-tums out to be highly significant. The presence of wbprop also has a substantial impact on the remaining variables in the regression. The coefficient on reserves, in particular, drops to one-third and becomes insignif icant. The coefficient on emres_pi also falls substantially, as one would expect, although it remains significant. This leads to the question of how the results would have been affected if the World Bank indicators had been included in the general-to-specific model selec tion process from the outset. The answer is quite striking (Model 4). In the pres ence of the governance variables, and with a slightly different parametrization of the variable proxying the development of securities markets (ebsecse, which now captures the level of securities markets development rather than the change), all other country-specific variables that previously seemed to matter can be deleted from the model without loss in explanatory power (compare the R2 of Models 4 and 2). Put differently: the only "fundamentals" that seem to have a robust effect on portfolio investment are the presence of a securities market infrastructure, and some confidence that the acquired assets would not be expropriated. In the pres ence of these two variables, no other country-specific variables (or for that matter, the residual information contained in investor ratings) seem to contribute signifi cantly to explaining investment flows. As before, we subjected the parsimonious Model 4 to a series of tests to see to what extent it could explain outliers in the raw data. Consider first portfolio inflows to Russia during 1997. Given what is known about the special circum stances driving these inflows, one would not expect a highly parsimonious. fundamentals-based model to fit this episode. The dummy variable for Russia is marginally significant at 5 percent. Note that the remaining coefficients are barely affected by the inclusion of this dummy, suggesting that this particular inflow episode was indeed unrelated to fundamentals. Next, we separately added dummy variables for the top (Estonia, Czech Republic, Hungary)20 and some of the bottom (Bulgaria, Ukraine) portfolio investment recipients. None of these dummies are individuaiJy significant, nor do they affect the coefficients on ebsecse and wbprop. Only if dummies for all three top portfolio investment recipients are jointly added to the equation does the coefficient on wbprop decline substantially and become insignificant at the 5 percent level (not shown). Thus, the fact that these top recip ients also stood out in terms of their perceived protection of property rights is indeed a factor driving the large coefficient on wbprop in Models 4 and 5.

Ill. Conclusions

Capital flows to transition economies have exhibited a diverse pattern in terms of both level and composition. Since 1993, private inflows in the Central-Eastern European countries have been high and consistently around 5-6 percent of GDP. They have been even higher in the Baltic countries, where they averaged l 0

20Slovenia. which also attracted exceptionally large amounts of portfolio investment (see Figure 5). is not part of this dataset because it is not covered by the 1997 World Bank governance indicators.

1.36

©International Monetary Fund. Not for Redistribution WHAT MOVES CAPITAL TO TRANSITION ECONOMIES?

percent of GDP since 1995. Other regions, such as Southeastern Europe and most CIS countries, witnessed lower and less stable inflows, and private flows to Russia were generally negative. Official flows and "excep6onal financing" (arrears rescheduling and debt forgiveness) were important sources of fm ancing in most countries during the early transition years, and continued to play an important role in most CIS countries at least until 1998. Foreign direct investment was a large, relatively stable source of private financing in most transition economies. In contrast, portfolio investment inflows were generally smaller, much more volatile, and concentrated in half a dozen coun tries in Central-Eastern Europe, Russia (before 1998) and recently the Baltics. "Other" private investment, reflecting mainly bank lending, takes an intermediate position. It was an important source of net funds, surpassing direct investment in some countries. Also. it was generally more volatile than direct investment, although Jess volatile than portfolio investment. Russia is again an exception. It received relatively little foreign direct investment as a proportion of GOP, but instead large portfolio inflows (until 1998). Moreover, "other" investment flows to Russia were consistently large and negative, perhaps reflecting capital flight. The regression results indicate that the pattern of inward direct investment in transition economies can be well explained in terms of a standard set of economic fundamentals. These include variables reflecting macroeconomic stability, the level of economic reforms, trade liberalization, natural resource endowments, the privatization method, direct barriers to inward direct investment. and a measure of government "red tape" that reflects obstacles to investment and entrepreneurship and is cJosely related to corruption. Unlike some of the earlier papers on Central and Eastern Europe, we do not ti nd that wages have a robust effect on direct investment flows in our sample. With the exception of natural resource endow ments, initial conditions-such as time under communism, or initial liberaliza tion-also seem to have no effect (other than through their possible effect on policy choices). Interestingly, investor perceptions of the ri sk or attractiveness of a country-as measured by investor magazine ratings-seem to have some predic tive value over and above their information about fundamentals, perhaps because they reflect investor biases or fads that also affect capital flows. In contrast, portfolio investment is very poorly explained by fundamentals. In part, this may be due to measurement problems or the problem of disentangling the determinants of demand and supply of portfolio assets when secondary asset markets are still very small. The only country characteristics that seem to be robust predictors of portfolio investment are the presence of a financial market infra structure (as measured by an index of securities markets development) and an indi cator of the protection of property rights. In the presence of these two variables, none of the many other country fundamentals we control for has any robust, statis tically significant, explanatory role. Thus, our findings endorse the commonly expressed view that the quality of governance was an important factor in attracting capital flows. For direct investment, this is one important variable among several others. For portfolio investment, it seems to be just about the only country charac teristic that helps to fit a volatile series, which otherwise bears no systematic rela tionship to country fundamentals.

107

©International Monetary Fund. Not for Redistribution Pietro Garibaldi. Nada Mora. Ratna Sahay. and Jeromin Zettelmeyer

APPENDIX

Compilation of Explanatory Va riables Used in Regressions

Restrictions on Capital Flows

We quantified the existence and magnitude of capital controls in transition economies on the basis of quaUtative information reported in the IMP's Annual Report on Exchange Arrangements and Restrictions. For each country we constructed separate indices of restrictions on foreign direct investments and port fo lio investments, as well as a composite index of overall restrictions, using the qualitative information reported under the subcategory "Capital Account." The average values of the indices are reported in Table A I for all countries in our sample. Larger values indicate larger restrictions on flows. The categories covered by the index of direct investment restrictions are approval requirements, the extent to which profits can be remitted abroad, ease in liquidating assets, and preferential treatment of direct investment.21 The index ranged from -0.2 to 6, where 6 reflects the most resu·ictive environment for direct investment. The table shows that Estonia had the least restrictive environment, with an overall negative value for the direct investment index, due to the absence of direct investment restrictions and the existence of subsidies to foreign direct investment inflows. At the opposite end of the scale is Belarus, whose direct investment regulations are the most restrictive. The index of portfolio restrictions was constructed based on the 1997-2000 issues of tbe TMF's Report (which in fact cover the years 1996-99 respectively). Prior to these issues, the Report did not contain a sufficiently detailed discussion of controls on capital and money market instruments. To deal with this, we assumed that the measure constructed for 1996 could have also applied to the earlier years; this implies that there is very little time variation in this data. The index rates both inward and outward controls according to their restrictiveness (no approval requirement, registration, approval, and outright prohibition). As shown in Table A2, the Baltic countries are the most open while the other FSU countries are the most restrictive; Russia is less restrictive than the average Central-Eastern European country.

Market Perceptions Market perceptions of "country risk" are represented by ratings of i11ternational private companies such as the Institutional Investor, Euromoney, and the Economist Intelligence Unit. We used the ratings issued by Euronwney because it provides virtually complete coverage of the transition economies during our sample period.

21A detailed description on how these categories were weighted and rated is available from the authors on request.

138

©International Monetary Fund. Not for Redistribution WHAT MOVES CAPITAL TO TRANSITION ECONOMIES?

Ta ble A 1. All Transition Economies: Indices of Restrictions on Capital Flows

Direct Investment Ponfolio Investment Composite Resrricrions Restrictions Index. fndex. average 93-99 Index. average 96-99 average 96-99•

Albania 1.80 1.00 1.40 Armenia 0.43 0.00 0.00 Azerbaijan 0.80 0.63 0.71 Belarus 3.37 1.00 1.90 Bulgaria 1.27 0.38 1.01 Croatia 0.94 0.63 0.79 Czech Republic 0.29 0.09 0.05 Estonia --0.03 0.00 0.00 Georgia 0.80 0.50 0.65 Hungary 1.12 0.44 0.62 Kazakhstan 2.64 1.00 1.70 Kyrgyz Republic 1.37 1.00 0.95 Latvia 1.40 0.00 0.50 Lithuania 2.80 0.00 1.40 Macedonia 0.80 0.88 0.84 Moldova 3.11 0.63 1.56 Poland 1.64 0.46 1.03 Romania 2.80 1.00 1.90 Russia 2.57 0.63 1.81 Slovak Republic 0.82 0.61 0.75 Slovenia 1.79 0.71 1.35 Taj ikistan 1.80 1.00 2.00 Thrk.menistan 2.80 1.00 1.90 Ukraine 1.80 1.00 1.40 Uzbekistan 2.80 1.00 \.90

Central and Eastern Europe 1.33 0.62 0.97 Baltic countries 1.39 0.00 0.63 Russia 2.57 0.63 1.81 Other countries of the former Soviet Union 1.98 0.80 1.33

Source: Authors' calculations based on IMF Annual Report on Exchange Arrangements and Re�7rictions. aComposite is an equally weighted sum of direct investment and portfolio restrictions for 1997.

Since each publication issues its ratings twice a year (in March and in September), we constructed simple annual averages based on the two published ratings. For both series, ratings range from 1 to 100, where I 00 represents the best rating. The Euromoney ratings are computed on the basis of assessments of country risk experts using the fol lowing nine weighted categories: economic petformance (25 percent), political risk (25 percent), external debt indicators (10 percent), debt in default or rescheduled (1 0 percent), credit ratings ( 10 percent), access to bank finance (5 percent), access to short-tenn finance (5 percent), access to capital markets (5 percent), and discount on fotfeiting (5 percent). The average rating for the transition countries for the period 1999 was 38.6 out of a total of 100. Table

139

©International Monetary Fund. Not for Redistribution Pietro Garibaldi. Nada Mora. Ratna Sahay. and Jeromin Zettelmeyer

Table A2.A ll Transition Economies: Average Country Scores from Euromoney (1999)and World Bank Surveyo

Euromoney Survey across all Survey across firms total ratingb firmsc with foreign participationc

Albania 18.05 3.70 3.81 Amtenia 25.43 3.68 3.70 Azerbaijan 33.38 4.03 3.71 Belarus 28.40 4.33 4.07 Bulgaria 38.60 4.40 n.a. Czech Republic 61.41 3.40 3.39 Estonia 54.32 3.33 3.14 Georgia 26.25 3.84 3.79 Hungary 65.01 3.30 3.22 Kazakhstan 40.49 4.14 4.17 Kyrgyz Republic 32.83 4.05 4.06 Latvia 50.54 3.77 3.85 Lithuania 49.44 3.76 3.62 Macedonia 24.16 3.49 3.09 Moldova 30.91 4.08 4.15 Poland 62.3 1 3.66 3.71 Russia 21.94 3.95 3.90 Slovak Republic 48.69 3.78 3.80 Ukraine 30.28 4.11 3.94 Uzbekistan 29.29 4.13 4.12 Averages All 38.58 3.85 3.75 Central and E!astem Europe 45.46 3.68 3.50 Baltic countries 51.43 3.62 3.54 Russia 21.94 3.95 3.90 Other countries of the former Soviet Union 30.80 4.04 3.97

Source: World Bank World Development Report 1997 Private Sector Survey, available at hnp://www. worldbank.orglhtmVprdmg/grthweb/wdr97 .htm; authors· computations. nSurveys for Croatia, Romania. Slovenia, Tajikistan, and Turkmenistan were not conducted. hRatings range from I to I 00, where I 00 represented the best rating or the least chance of default.

A2 (first colunm) shows that the Central and Eastern European countries and the Baltics scored much better than Russia and other FSU countries.

Institutional and Legal Proxies

For institutional and legal proxies that capture governance-type issues and the existence of red tape, we use a World Bank data set compiled for the World Development Report in 1997. lt is based on a cross-country survey of firms (local or with foreign participation). The aim of the survey was to capture the institu tional uncertainty within a particular country from the viewpoint of local

140

©International Monetary Fund. Not for Redistribution WHAT MOVES CAPITAL TO TRANSITION ECONOMIES?

entrepreneurs, as opposed to the opinion of Western or international experts. The survey also recorded size, location, and whether or not the responding firm had foreign participation.22 The survey was divided into 5 broad sections. The first is "predictability of laws and policies"-whether changes in laws and policies are uncertain and whether the government takes into account concerns raised by business. The second is "political instability and security of property"-whether there is uncer tainty in the political process and whether that has an impact on business decisions and property rights. The third, "overall government-business interface," asks participants to judge whether there are strong obstacles in certain policy areas like inflation, corruption, tax regulations etc. The fourth, "bureaucratic red tape," aims to evaluate the frequency, severity, and uncertainty of "additional payments" that need to be made. Finally, the survey asks for a rating of the efficiency of govern ment in providing services such as customs, roads, and telecommunications. The responses to each of the questions was converted into a 1 to 6 numerical scale, where 6 represented the worst perception of domestic institutional constraints. To obtain a general perception of each of the five categories and a total country score, we average the five sections' responses, and in Table A3 (third and fowth columns), we report the final country scores based on all and foreign partic ipation averages. The average for all the transition countries surveyed based on averaging across all firms was 3.85 while it was 3.75 when based on averaging across finns with foreign participation. In general Central and Eastern Europe and the Baltic countries scored best, followed by Russia and the other countries of the former Soviet Union. Interestingly, firms with foreign participation rated all subgroups better than all firms. Also interesting is the fact that while all firmsrated Central and Eastern Europe better than the Baltic countries, firms with foreign participation did not. In our regressions modeling portfolio investment, we also used an EBRD index of the degree of development of security markets and nonbank financial institutions. For this index, 1 reflects little progress; 2 implies that securities exchanges and brokers have been set up along with some trading in bonds and securities, although the legal and regulatory framework dealing with such trading is still in its early stages; 3 reflects a large degree of private issues together with a more advanced legal framework, and the emergence of nonbank financial institu tions like pension funds; and 4 means a highly developed market. The regional averages for this index exhibit little time variance. The other FSU countries lag behind at 1.5 while the Central-Eastern Europe and the Baltics average 2.3. Russia jumped to a rating of 3 in 1996, putting it at the same level of development as the more advanced Central-Eastern Europe countries, such as the Czech Republic, Hungary, and Poland.

22For more details, see the World Bank's 1997 World Development Report and Brunetti, Kisunko. and Weder ( 1998) (website: http://www.unibas.ch/wwz/wifor/stafflbw/survey/index.html.)

141

©International Monetary Fund. Not for Redistribution Pietro Garibaldi, Nada Mora, Ratna Sahay, and Jeromin Zettelmeyer

Ta ble A3. Dummy Va riables for Privatization Method

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999

Albania direct• 0 0 0 0 0 0 0.5 0 0 0 voucherb 0 0 0 0 0 1 I 0 0 0.5 inside!" 0 0 0 0 0 0 0 0 0 I Armenia direc..t 0 0 0 0 0 0 0 0.25 0 I voucher/> 0 0 0 0 0 I I 1 I 0 inside!" 0 0 0 0 0 0.5 0.5 0.5 0.5 0 Azerbaijan direct!' 0 0 0 0 0 0 0 0 0.5 0 voucher/> 0 0 0 0 0 0 0 I I 0 insider<' 0 0 0 0 0 0 0 0.5 0 0 Belarus direct" 0 0 0 0 0 0 0 0 0 0 voucher/> 0 0 0 0 0 0 0 0 0 0 inside!" 0 0 0 1 1 1 1 0 0 0 Bulgaria direct" 0 0 0 0 0 0 0.5 I I voucberb 0 0 0 0 0 0 I 0.5 0.5 0.5 insider<' 0 0 0 0 0 0 0 0 0 0 Croatia di rect• 0 0 0 0 0 0 0 0 0 0 voucher/> 0 0 0 0 0 0 0 0 I 1 in sider<' 0 0 I I I 0 0 Czech direct• 0 0 0.5 0.5 0.5 I I I 0.5 I Republic voucherb 0 0 I I 0.5 0 0 I 0

insider c 0 0 0 0 0 0 0 0 0 0 Estonia direct" 0 0 0 I 1 I I l I 1 voucher/> 0 0 0 0.5 0.5 0.5 0.5 0.5 0 0 in� ide!" 0 0 0 0 0 0 0 0 0 0 Georgia direct• 0 0 0 0 0 0 0 I I I voucher/> 0 0 0 0 0 I I 0 0 0 in sider<' 0 0 0 0 0 0.5 0.5 0 0 () Hungary direct• I I I I I I I I I I voucher�' 0 0 0 0 0 0 0 0 0 0 insider' 0 0 0 0 0 0 0 0 0 0 Kazakhstan direct• 0 0 0 0.5 0.5 0.5 I I I 0 voucher�' 0 0 0 I I I 0.5 0.5 0 0 insider< 0 0 0 0 0 0 0 0 0 0 Kyrgyztan di rect• 0 0 0 0 0 0.5 0.5 0.5 I 0 voucherb 0 0 0 0 0 I I I 0 0 inside!" 0 0 l I I 0 0 0 0 0 Latvia di rect• 0 0 0 0 0.5 0.5 0.5 0.5 I 0 voucherb 0 0 0 0 I I I I 0 0 inside.-: 0 0 0 0 0 0 0 0 0 0 Lithuania direct" 0 0 0 0 0 0.5 I I I I voucher�' 0 0 l 1 l I 0 0 0 0 insider< 0 0 0.5 0.5 0.5 0 0 0 0 0 Macedonia. direct" 0 0 0 0 0 0 0 0 0 0.5 FYR voucher/> 0 0 0 0 0 0 0 0 0 0 insider< 0 0 0 I I I I I I I Moldova direct0 0 0 0 0 0 0.5 0.5 0.5 I I voucherb 0 0 0 0 I I I 0 0 0 insider< 0 0 0 0 0 0 0 0 0 0 Poland direct11 0 0.25 0.25 0.25 0.25 0.25 0.25 0.25 I I voucherJI 0 0 0 0 0 I I I 0 0 insider< I I I 0.5 0.5 0.5 0 0

142

©International Monetary Fund. Not for Redistribution WHAT MOVES CAPITAL TO TRANSITION ECONOMIES?

Table A3. (concluded)

1990 1991 1992 1993 1994 1995 19% 1997 1998 1999

Romania direct" 0 0 0 0.5 0.5 0.5 0.25 0.5 0 I voucher/> 0 0 0 0 0 0 I I 0 0 insider: 0 0 0 I I I 0.5 0 0 0

Russia directa 0 0 0 0 0.25 0.5 0.5 0.5 0 0 voucberb 0 0 I I 0.5 0 0 0 0 0 insider" 0 0 0 0 I I I I 0 0 Slovak direct0 0 0 0.5 0.5 0.5 0.5 0.5 0.5 I 0 voucher> 0 0 I I l 0.25 0.25 0.25 0 0 insider" 0 0 0 0 0.25 I I I 0 0 Slovenia direct" 0 0 0 0 0 0.25 0.25 0.25 0.25 0.25 voucher" 0 0 0 0 0 0.5 0.5 0.5 0.5 0.5 insider" 0 0 0 0 0 I I I I I Taj ikistan direct0 0 0 0 0 0 0 0 0 I I voucher/> 0 0 0 0 0 0 0.5 0.5 0 0 insider<' 0 I I 0 0 0 I I 0 0 Turkmenistan direct" 0 0 0 0 0 0 0 0 I I voucher/> 0 0 0 0 0 0 0 0 0 0 insider" 0 0 0 0 0 0 0 0 0 0 Ukraine di recta 0 0 0 0 0 0 0.25 0.25 0.5 0.5 voucher/> 0 0 0 0 0 0.5 I I I I insider" 0 0 I I 0 I 0.5 0.5 0 0 Uzbekistan direct" 0 0 0 0 0 0 0 0 0 0 vouch� 0 0 0 0 0 0 0.5 0.5 0 0 insider" 0 0 0 0 I I 1 0 0

""direct" refers to privatization through direct sales. "''voucher" refers to privatization through auctions involving vouchers (also known as "mass privitization''). <'·insider'' refers to insider privitization-either voucher privatization with significant concessions to insider. or management-employee buyouts.

Indicators of the Privatization Method

To characterize the privatization method, we constructed three indicator vari ables: prsal (privatization through direct sales), prvou (privatization through auctions involving vouchers, this is sometimes referred to as "mass privatization"), and prins (insider privatization-either voucher privatizations with significant concessions to insiders, or management-employee buyouts). Following the approach used by the EBRD in their 1997 summary of ptivatization methods, 23 we assigned a value of 1 to the series if the method it was supposed to characterize constituted d1e "primary" privatization method during this year; a 0.5 if it consti tuted the "secondary" privatization method; and a 0.25 if it constituted the "tertiary" method. If no privatization at all took place, all three indices were assigned a value of 0.

23 1997 Tra nsition Report. Table 5.7.

14.3

©International Monetary Fund. Not for Redistribution Pietro Garibaldi, Nada Mora, Ratna Sahay, and Jeromin Zettelmeyer

Our basic sow·ce on how each country privatized in each year was the EBRD's chronicle of large scale ptivatization as it appears in the country summaries of each Tra nsition Report. We chose this over a pure cross-sectional index (which could have been derived directly and much more easily from the EBRD's 1997 summary of privatization methods) because we wanted to capture important changes in the privatization method over time. For example, a switch from. say. insider privatization to sales to outsiders could have led to a spike in direct invest ment, which a purely cross-sectional variable would be unable to explain. The disadvantage is that our classification necessarily involves some judgment, over and above the judgment applied by the EBRD itself. The resulting three series are reproduced in Table A3.

REFERENCES

Berg, Andrew, 1993, '·Measurement and Mismeasurement of Economic Activity of Economic Activity During Transition to the Market,'' in Eastem Europe in Transition: From Recession to Growth? ed. by Mario Blejer and others, World Bank Discussion Paper No. 196 (Washington: World Bank), pp. 39-63.

---, Eduardo Borensztein, Ratna Sahay, and Jeromin Zettelmeyer, 1999. "The Evolution of Output in Transition Economies: Explainjng the Differences," TMF Working Paper 99/73 (Washington: International Monetary Fund).

Bevan, Alan A.. and Saul Estrin, 2000, "T he Determinants of Foreign Direct Investment in Transition Economies." CEPR Discussion Paper No. 2638 (London: Centre for Economic Policy Research).

Brenton. Paul, Francesca Di Mauro, and Matthias LUcke, 1999, ''Economic Integration and FD!: An Empirical Analysis of Foreign Investment in the EU and in Central and Eastern Europe," Empirica. Vo l. 26 (No. 2), pp. 95-121.

Brunetti. Aymo, Gregory Kisunko. and Beatrice Weder. 1998. "How Businesses See Government: Responses from Private Sector Surveys in 69 Countries," !FC Discussion Paper No. 33 (Washington: International Finance Corporation).

---, 1998, ·'Credibility of Rules and Economic Growth: Evidence from a Worldwide Survey of lhe Private Sector." World Bank Economic: Review, Vol. 12 (September), pp. 353-84.

Calvo. Guillermo A., Ratna Sahay, and Carlos A. Vegh, 1996. "Capital Flows in Central and Eastern Europe: Evidence and Policy Options," in Private Capital Flows to Emerging Markets After the Mexican Crisis. ed. by Guillem1o A. Calvo, Morris Goldstein, and Eduard Hochreiter (Washington: Institute for International Economics), pp. 57-90.

Claessens, Stijn, Daniel Oks. and Rossana Polastri, 1998. "Capital Flows to Central and Eastern Europe and the Former Soviet Union.'' World Bank Policy Research Working Paper No. 1976 (Washington: World Bank).

De Melo. Martha, Cevdet Denizer, and Alan Gelb. 1996. "From Plan to Market: Patterns of Transition," World Bank Policy Research Working Paper No. 1564 (Washington: World Bank). ---, 1997, "Circumstance and Choice: The Role of Initial Conditions and Policies in Transition Economies,.. World Bank Policy Research Working Paper No. 1866 (Washington: World Bank).

144

©International Monetary Fund. Not for Redistribution WHAT MOVES CAPITAL TO TRANSITION ECONOMIES?

European Bank for Reconstruction and Development (EBRD). Transition Report, various issues (London).

Fischer, Stanley, Ratna Sahay, and Carlos A. Vegh. 1996. "Stabilization and Growth in Transition Economies: The Early Experience," Journal of Economic Perspectives, Vo l. I 0 (Spring), pp. 45-66.

---, 1997. "From Transition to Market: Evidence and Growth Prospects;· in Lessons from the Economic Transition: Central and Eastern Europe in the 1990s, ed. by Salvatore Garibaldi Zecchini (Dordecht, Boston. and London: Kluwer), pp. 79-102.

Gavrilenkov, Evgeny, and Vincent Koen, 1994, "How Large Was the Output Collapse in Russia? Alternative Estimates and Welfare Implications," IMF Worki ng Paper 94/ J54 (Washington: International Monetary Fund).

Halpern. Uszl6, and Charles Wyplosz. I 997. ·'Equilibrium Exchange Rates in Transition Economies." Staff Papers, lmemational Monetary Fund. Vo l. 44 (December), pp. 430-61.

Holland, Dawn, and Nigel Pain, 1998, "The Diffusion of Innovations in Central and Eastern Europe: A Study of the Determinants and Impact of Foreign Direct Investment." Discussion Paper No. 137 (London) National Institute of Economic and Social Research NIESR.

International Monetary Fund. Annual Report on Exchange Arrangements and Trade Restrictions (Washington: various issues).

---, 2000, Balance of Payments Ye arbook 2000 (Washington: TMF).

---. International Financial Statistics (Washington: various issues).

Kaufmann, Daniel, and Aleksander Kaliberda, 1996, ·'Integrating the Unofficial Economy into the Dynamics of Post-Socialist Economies: A Framework of Analysis and Evidence," World Bank Policy Research Wo rking Paper No. 1691 (Washington: World Bank).

Keane, Michael P.. and David E. Runkle. 1992, "On the Estimation of Panel Data Models with Serial Correlation When lnstn1ments Are Not Strictly Exogenous," Joumal of Business and Economic Statistics, Vol. 10 (January). pp. 26-29.

Krajnyak, Kornelia, and Jeromin Zettelmeyer, 1998, ''Competitiveness in Transition Economies: What Scope for Real Appreciation?" Staff Papers, International Monetary Fund, Vol. 45 (June), pp. 309-62.

Lankes, Hans-Peter, and A.J. Venables, 1997, ''Foreign Direct Investment in Eastern Europe and the Former Soviet Union: Results from a Survey of tnvestors," in Lessons from the Economic Tr ansition: Central and Eastem Europe in the 1990s, ed. by Salvatore Zecchini (Dordecht, Boston, and London: Kluwer), pp. 555-<56.

Lansbury, Melanje, Nigel Pain. and Katerina Smidkova, 1996. "Foreign Direct Investment in Central Europe Since 1990: An Econometric Study," National Institute Economic Review, Vol. 156 (May), pp. 104-14.

Leamer, Edward E., 1983, ··Let's Take the Con Out of Econometrics," American Economic Review, Vol. 73 (March), pp. 31-43.

Loungani. Prakash. and Paolo Mauro, 2000, "Capital Flight from Russia." £MF Policy Discussion Paper 00/6 (Wash ington: International Monetary Fund).

World Bank, 1997, Wo rld Development Report 1997 (New York: Oxford University Press for the World Bank).

145

The Gains from Privatization in Transition Economies: Is "Change of Ownership" Enough?

CLIFFORD ZINNES, YAIR EILAT, and JEFFREY SAC HS'

This paper seeks to clarify what fa ctors contributed to the macroeconomic gains and losses from privatization in transition economies over the past decade. In contrast to the original "Washington Consensus," which had a tendency to equate change-of-title with privatization, we find that economic performance gains come only fr om "deep " privatization, that is, when change-of-title reforms occur once key institutional and "agency " -related refo rms have exceeded certain threshold levels. We also find that as a result of diferentf initial conditions the economic f performance responses of countries to the same policies are diferent. [JEL: G38, L33, 01 1, P31, and P37]

his paper is the thjrd in a series1 that evaluate the fL rst decade of economic reform in transition economies. Based on indicators developed in Sachs, TZinnes, and Eilat (2000a), the present paper contributes to the already large

·clifford Zinnes is Director of Research Coordination at the IRIS Center, Department of Economics. University of Maryland. College Park. Yair Eilat i� a PhD. candidate in the Department of Economics at Harvard University. Jeffrey Sachs is the director of the Center for International Development at the Kennedy School of Government and the Galen L. Stone Professor of lmemationnl Trade at Harvard Unive rs ity. This research was funded under USAID Task Order PCE-Q-00-95--000 16-00and grants from the Harvard Institute for International Development and the Center for International Development, both at Harvard University. This paper was submitted to IMF Staff Papers in November 2000. We would like to thank Luis L6pez-Calva. Lawrence Camp. Oleh Havrylyshyn. Orest Koropccky. Charles Mann. and Janos Szyrmer, and express appreciation for the tireless research assistance of Georgi Kadiev. Hongyu Liu. Chris Saccardi, and Irina Tratch. 'The other two are Sachs, Zinnes, and Eilat (2000a), which develops an inilial conditions typology of countries in transition and creates indicators for gauging progress in reforms and performance over the lransilion period, and Zinnes. Eilat, and Sachs (200 I). which benchmarks competitiveness of transition countries in the years 1997-98.

146

©International Monetary Fund. Not for Redistribution THE GAINS FROM PRIVATIZATION IN TRANSITION ECONOMIES literature on transition by seeking to clarify what factors contributed, at the macro level, to the gains from privatization in transition economies over the past decade. In doing so, our goal is to point the way to a revised paradigm for privatization policy in transition economies. We first summarize the paradigm debate and show how the issues of privatiza tion play a central role. We find, as reflected in the original "Washington Consensus," that there has been a tendency to equate change-of-title with privatiza tion, with the consequence of change-of-title becoming the policy imperative. Based on a review of the literature on the gains from privatization, however, we identify the importance of additional factors. These include institutions to address agency issues, hardening of budget constraints, market competition, and depo)jtization of flrm objectives, as well as developing institutions and a regulatory framework to support them. In this paper we examine the empirical evidence across 24 countries to determine whether change-of-title alone has been sufficient to achieve economic performance gains or whether these other prerequisites found in the literatme (wbkh we refer to as "OBCA" reforms, see definition in Section il) are important. We then introduce two key elements of our approach. These are the impor tance of initial conditions for economic performance and the significance of the transformational cycle of transition. For our econometric analysis below, we then introduce several indicators, which we developed in Sachs, Zinnes, and EiJat (2000a), to captme the degree of change-of-title, agency-related issues, the progress in other reforms, and alternative measures of economic performance. We then proceed to examine econometrically the central concerns of the paper. We ft rst show that privatization involving change-of-title alone is not enough to generate economic performance improvements. This result is robust to several alternative measures of economic performance that we utilize, including GOP recovery, foreign direct investment, and exports. We then introduce our OBCA indicator to capture the reforms directed at prudential regulation, corporate gover nance, hardening of enterprise budget constraints, and management objectives. We show that, while this measure on its own contributes to economic performance improvements, the real gains to privatization come from complementing (combining) change-of-title reforms with OBCA reforms. As Pistor (1999) under scores, it is onJy when the legal and regulatory institutions supporting ownership are in place and functioning that owners can exercise their prerogatives conferred by a change-of-title to pressure fiTms to improve their productivity and prof itability. Only then will tl1e economic performance of the country improve. We can quantify this result in the following way: the higher the level of OBCA, the more positive the economic performance impact from an increase in change-of-title privatization. In particular, where change-of-title has a positive impact, the impact will be even more positive the higher is the level of OBCA; where change-of-title has a negatjve impact, the impact will be less negative the higher is tlle level of OBCA. A corollary to this result is that there is a threshold level of OBCA for change-of-title privatization to have a positive economic performance response. Thus, if complementary OBCA reforms are not sufficiently developed, change-of-title privatization may have a negative performance impact. An explanation for tlle cases of worsening overall economic performance from

147

©International Monetary Fund. Not for Redistribution Clifford Zinnes, Yair Eilat. and Jeffrey Sachs change-of-title privatization is that transfer of ownership without the institutional structures in place for owners to exercise their authority simply replaces poor government control of management with weak or no private-sector control. The paper indicates the countries and years that did not exceed these thresholds. This result also suggests that one size (policy) does not fit all; privatization policies must be tailored to the level of complementary reforms in place. We close by cautioning that our results are hardly definitive. While we have made every effort to use the latest and best data-including a 25-country survey2 especially conducted for this purpose-the amount of structural change occurring is enormous, the number of observations too few, and the data still too noisy to claim unconditional success. In addition, we believe that research at the macro level can be seen only as a supplement, not a substitute, to research at the firm level. Nevertheless, given that the results are in line with those predicted by agency theory and given that we have utilized a number of alternative economic performance measures and a variety of econometric specifications, we feel that future investigations will broadly support our central conclusions.

I. A Paradigm in Flux It has not been unusual historically, during a time of major economic crisis, for policymakers to base key and often radical actions in a region upon a set of tenets. Sometimes the exact nature and underlying assumptions of the tenets are not even clear until weU after the chaotic events. The twentieth century had its share of examples, including Lenin's "New Economic Program," Roosevelt's "New Deal," and the Marshall Plan for Europe. It is fair to say that the first decade of transition to a market economy also has been based on a series of tenets or, as we shall refer to them here, a "paradigm." So well known did this paradigm become that it is often referred to as the "Washington Consensus" since it became the mantra of the donor community centered around Washington, D.C. Since a description and analysis of this consensus may be fo und elsewhere (WiUiamson, 1990, 1993, 1997; Kolodko 1998; Aziz and Wescott, 1997), we only summarily mention that its key tenets included fast privatization, i1mnediate macrostabilization, quick liberalization and sustaining of financial discipline, and opening of the economy to foreign trade and investment. ln the realm of privatization, we may identify a further set of assumptions underlying the paradigm. First and foremost was the idea that the linchpin of tran sition was to transfer ownership of the firms in the economic sectors to private hands-and to do so as fast as possible. Once in private hands, a series of self reinforcing, virtuous, though self-interested, forces would emerge to demand the creation of all the institutions required for private ownership, thereby locking in the market economy. Moreover, the new shareholder class would demand corpo-

2While this survey (see Sachs. Zinnes. and Eilat, 2000a) includes Albania, we have dropped it from the analysis in the present paper owing 10 lack of data for a number of key variables.

148

©International Monetary Fund. Not for Redistribution THE GAINS FROM PRIVATIZATION IN TRANSITION ECONOMIES rate governance regulation to insure their ability to exert oversight on enterprise managers. These tenets led to a debate of greater and greater vehemence over the past decade (Balcerowicz, 1993; Nellis, 1999; Dabrowski, 1996; Stiglitz, 1998), even while the obsessions with macro stabilization, privatization, and structural adjust ment have given way to a fourth ingredient: systemic transformation (Aslund,l994; Kornai, 1994; Sachs, 1996). Now that a decade has passed, enough data has become available to examine these concerns. In particular, this paper asks whether privatization has led to better economic performance, and what are lbe preconditions necessary for privatization to generate gains in economic performance. A common though implicit thread underlying these questions is the degree to which supporting institutions are necessary in order to achieve the full gains from privatization (Pistor, 2001).Such institutions might include, inter alia, those responsible for shareholder protection, banking adequacy, creditor protection and bankruptcy courts, capital market supervision, and commercial code enforcement. In the present paper, we focus on the supporting role institutions have in bringing out the full potential of privatiza tion. We argue that policymakers should pursue "deep privatization"-that is, both change-of-title reform and a strengthening of supporting institutions.

II. The Theory of the Gains from Privatization With excellent surveys already available (e.g., Havrylyshyn and McGettigan, 1998; Sheshinski and L6pez-CaJva, 1999), we only highlight here those aspects pertinent to the motivation of our theoretical framework. We start by discussing the relevant theoretical literature and then move on to review the empirical evidence. A principal reason for privatization has been the existence of information asymmetries and incomplete contracting problems, leading to severe incentive problems and therefore serious efficiency losses from public ownership. This incentive-efficiency link has been called the "agency" problem and, within the context of privatization, has two threads. The managerial view (Vickers and Yarrow, 1990) concerns the inability of the state to monitor enterprise managers. This inability stems from the lack of a market to price and instiU discipUne on firms through the threat of takeover or bankruptcy. The political view (Shapiro and Willig, 1990; Shleifer and Vishny, 1994, 1996) concerns the temptation of polit ical interference to distott manager objectives away from profit maximization and toward others such as employment maximization. Moreover, this interference can also result in the perception among finn managers of a "soft" budget constraint (Kornai, 1986), in which they expect ex post subsidies or writeoffs to cover enter prise losses due to production inefficiencies. What the agency view points out is that the gains from change of ownership (referred to below as change-of-title) will likely depend on how a country's legal, regulatory, and institutional environment addresses agency-related issues. For the purposes of the empirical work below, we classify these issues into three types. The first relates to the firm's objective (0) fu nction and how closely it reflects

149

©International Monetary Fund. Not for Redistribution Clifford Zinnes, Yair Eilat, and Jeffrey Sachs

profit maximization. The second relates to the hardness of the firm's budget constraint (BC). The third relates to the legal and institutional framework through which firm owners are able to monitor and control enterprise managers, the so ca1led principal-agent (A) problem. For simplicity we combine the letters in paren theses to name this class of issues OBCA. On the issue of the implementation of privatization, Havrylyshyn and McGettigan (1998) identify two schools of thought. The first school of thought stresses the importance of the competitive environment and market structure over ownership (Nellis, 1999). For transition economies, the creation of a competitive environment would occur through the hardening of enterprise budget constraints rather than a rush into privatization. This was thought to occur, according to Frydman and others (1997), as a result of pressures from macroeconorruc stabi lization on firms to restructure or go out of business. The second school of thought stresses the need for a headlong rush into privatization, though the need to even tually follow up with the development of supporting institutions is sometimes noted. Both these views underscore the insights from the preceding discussion regarding the importance of the hardness of the firms' budget constraints, as well as the likely importance of establishing a multitude of market institutions. Much of the empirical literature on the impact of privatization on economic performance was inspired by Boardman and Vining (1989) and Megginson, Nash, and Van Randenburgh (1994) whose work is in the nontransition country context. Trus literature is of two types: case studies of a small sample of firms (Earle, Frydman, and Rapaczynski, 1993) and cross-industry econometric studies, either country-specific (Barberis and others, 1996) or multicountry (Frydman and others, 1998; Pohl and others, 1997). Based on either firm-level surveys or data on publicly traded firms, these studies are essentially microeconomic in nature and primarily analyze the effects on labor productivity, level of employment, enter prise revenues, and sometimes even profitability. These studies find privatization to have positive effects across these measures) With the exception of Claessens and Dj ankov (1998),4 this literature does not examine econometricaJly the contri bution of the legal or institutional regime to enterprise performance. While these studies are quite revealing, they can only provide a partial picture, mainly because even the largest of them covers only seven countries. There are currently over two dozen transition economies, but none of these papers deals with both Central and Eastern Europe and countries of the former Soviet Union. Part of the reason for this is the high cost of firm survey data collection in so many coun tries. Even where such firm-specific data exist they are hard to analyze, since little uniformity or consensus exists regarding the way to define, classify, collect, or treat such data, especially in the case of transition. A natural, if imperfect, alternative to complement the firm-level studies would be to consider macroeconomic econometric evidence of gains from privatization. Turning first, however, to the macroeconomic theory literature on the subject, we

3 See Havrylyshyn and McGettigan (1998) for a summary. 4 Controlling for institutional differences, they test several propositions of Shleifer and Vi shny ( 1994) regarding how privatization and stabilization affect firm behavior.

150

©International Monetary Fund. Not for Redistribution THE GAINS FROM PRIVATIZATION IN TRANSITION ECONOMIES

find that Literature is much less developed (Blanchard, 1997) than on the micro economic side. Still the literature carries two implications for our work. First, it suggests that gains from privatization at the level of macroeconomic performance depend on complementary policies, and not just those related to appropriate institutions, as we described above (Aziz and We scott, 1997). While privatization means the ending of subsidies, which drain state finances, privatiza tion also means the state will lose its share of enterprise profits unless comple mentary reforms create an adequate tax code and administration. The potential for efficiency gains from privatization requires price and wage liberalization in order to create a price system that reflects economic scarcity. Finally, unless privatiza tion is accompanied by reforms to liberalize the current and capital accounts, the newly privatized domestic firms may not be able to gain access to foreign skills, markets, and financing necessary for their success. Second, privatization may have opposite short-term and long-term economic performance impacts (Aghion and Blanchard, 1993; Roland, 1994). For example, unemployment may increase over and above what would be expected from the resource reallocation associated with enterprise restructuring suggested by the microeconomic perspective. This may occur if privatization leads to employment shedding as managers are freed from political interference and return to profit maximization as their principal objective. A healthy macroeconomic empirical literature does exist on the determinants of transition paths (de Melo, Denizer, and GeJb, 1995; Fischer, Sahay, and Vegh, 1996; Havrylyshyn, Izvorski, and van Rooden, 1998), often with real GOP growth as an explanatory variable. Sheshinski and L6pez-Calva (1999), however, indicate that little macroeconomic econometric evidence exists on the effects of privatiza tion. It is precisely this gap that we aim to redress in the present paper. For this purpose, we make use of the indicators created in Sachs, Zinnes, and Eilat (2000a). The indicator approach is predicated on the assumption that economic concepts can be captured-especially when data are poor or intermittent-by aggregating several imperfectly reported data series, in order to "put the law of large numbers to work."S

Ill. Data and Empirical Approach

The first element of our framework is the use of an initial conditions cluster typology. As explained in Sachs, Zinnes, and Eilat (2000a), we assign countries to groups based on similarities in variables at the start of transition. These variables represent various aspects that may be relevant for a country's prospects of transi tion perfonnance.6 The clustering exercise resulted in seven clusters of transition countries, as listed in Table 1.

ssee Zinnes. Eilat. and Sachs (2001) for a discussion of the indicator approach. 6For this purpose, we used a computer program that assigns countries to clusters in a way that mini mizes intracluster differences and maximizes inter-cluster differences according to a chosen list of vari ables. The categories of initial condition variables we used include physical geography, macroeconomics. demographics and health, trade and trade orientation, infrastructure, industriali7.ation, wealth, human capital. market memory. physical capital. culture. and political situation.

151

©International Monetary Fund. Not for Redistribution Clifford Zinnes, Yair Eilat, and Jeffrey Sachs

Ta ble 1. Summary of the Initial Conditions-Based Typology

Cluster Name (Number)* Country Membership

EU border states (1) Croatia, Czech Republic, Hungary. Poland, Slovakia, Slovenia The Balkans (2) Bulgaria, Macedonia, Romania The Baltics (3) Estonia, Latvia. Lithuani a Albania (4) Albania Western FSU (5) Belarus, Moldova, Russia, Ukraine Caucasus (6) Armenia, Azerbaijan, Georgia Central Asia (7) Kazakhstan. Kyrgyz Republic, Tajikistan, Thrkmenistan, Uzbekistan

*Figures I and 2 use these numbers ro refer to the clusters.

The cluster approach, by considering groups of countries based on their initial conditions, permits a more controlled basis for comparing "successful" and "failed" policies implemented during transition. By using cluster-fixed effects we can analyze within-cluster differences and thereby assess policy effectiveness.? A second element of our approach is acknowledging that the important factor in the time domain is the elapsed time since transition and not calendar time. Our hypoth esis, which we base on Sachs (1996) and Kornai (1994), is that each country, regardless of the actual calendar date, passed through a sequence of recessions, typically first from macrostabilization and then from restructuring. We capture these in our regressions through the use of transition year dummy variablesfor each year of transition.& We also use dummy variables to take explicit account of the effects of macrostabilization on economic perfonnance (Sachs, 1997). In this work we take advantage of a unique panel dataset of indicators for the period 1990-98 developed in Sachs, Zinnes, and Eilat (2000a).9 The dataset includes a series of indicators representing the components of the depth of privati zation and progress in transition. These indicators were constructed using two types of sources. We first used virtually all published data sources available at the time for which substantial country coverage existed for the transition countries. Second, we developed and administered a 100-question survey to research institutes in all 25 transition countries. The goal of the survey was to augment published sources with data sources not reported by international collection agencies.

7An alternative approach to control for initial conditions is to explicitly include initial condition vari ables in the regression. See, for example, de Melo, Oenizer. and Gelb (1995), who use principal compo nems analysis to cluster (i.e., reduce the number of) variables rather than countries. syear of beginning of transition: 1990: Bulgaria, Czech Republic. Hungary. Poland. Romania, Slovakia: 1991: Croatia. Macedonia, Slovenia; 1992: Armenia, Azerbaijan. Belarus, Estonia, Georgia, Kazakhstan, Kyrgyz Republic, Lithuania, Latvia, Moldova. Russia, Taj ikistan, Turkmenistan, Ukraine. Uzbekistan. The values of these dummies are available from the authors. Our analysis shows that most of the cluster and transition year dummies are statistically significantly differem from one another. providing support for their inclusion in the regressions. See Sachs, Zinnes, and Eilat (2000b) for a detailed discus sion of the methodology and interpretation of results. 9A1J indicators are scaled to have a mean of zero and a variance of unity across the 25 countries and years 1990-98 of transition.

152

©International Monetary Fund. Not for Redistribution THE GAINS FROM PRIVATIZATION IN TRANSITION ECONOMIES

To capture the depth of privatization, we follow d1e theoretical framework presented above and break "depth" into two major components. The first we call change-of-title. This indicator consists of tile Ew-opean Bank for Reconstruction and Development (EBRD) large-scale and small-scale privatization indices, the private sector share of GOP, tile percentage of state firms privatized, and the private sector share of employment. The second we call OBCA (see above) whkh aims to capture the firm management objectives, the hardness of its budget constraint and the quality of corporate governance, and shareholder protection regulation. The indicator consists of the share of tax anears in GDP, the ratio of budget subsidies to average GDP over tile period, the share of bad loans to total loans, the electricity tariff collection ratio, the likelihood of a government bailout of a midsized private-sector f1rm, the existence of bankruptcy courts, and tile EBRD restructuring and legal system indices. Figures 1 and 2 present the progress in change-of-title and OBCA over tlle transition cycle averaged by cluster (see Table 1) as well as the scores of the different countries in 1998. The Appendix provides the "recipes" used for constructing these variables. In addition to the change-of-title and OBCA indicators, we also developed an aggregate reform indicator (REF) of the otller reforms under way. We use REF as a control to ensure that our privatization variables do not proxy for other reforms. REF comprises several components. The social safety net component captw-es several aspects of d1e government's attempt to soften tlle negative social impacts of transition. The price liberalization component comprises goods and services prices, as well as wages and the degree of competition in tile price formation process. The capital markets component comprises subindicators for the stock market, secw-ities market, and the nonbank financial institutions. The tax reform component includes improvements not just in the tax code but also in its adminis tration. The banking sector component focuses on d1e degree of competition in the sector and the degree to which the sector is providing economic agents with adequate credit and services. Finally, tile land refom1 component concentrates on measuring the degree that land markets function in a way consistent with the needs of a market economy. For measuring economywide, macroeconomic performance, we have chosen four measures. These include real GOP per capita, foreign direct investment (FDI) per capita, FDI per unit GDP in 1989, and exports per unit GDP in 1989. We now look at these measures in more detail. The first economic performance measure, IGDP, describes domestic output. Using GDP growtil rates from EBRD (1999, p. 73), we construct an index of real GDP relative to 1989 (so tilat the value for each country is lOO in J 989). The index tilerefore indicates the degree of economic recovery by showing the percent of pretransition output attained in a given year. This approach facilitates the compar ison of performance across countries witil vastly different initial per capita figw-es.10 We also test the logarithmic transformation of dlis variable (Log/GDP) as a dependent variable.

I!Yfhis approach differs from other papers in the literature-for example, de Melo, Denizer, and Gelb (1995); Havrylyshyn, Izvorski. and van Rooden (1998): and Fischer. Sahay, and Vegh (1996). who use the annual growth rate of real gross domestic product as a dependent variable and not growth since 1989.

15J

©International Monetary Fund. Not for Redistribution Clifford Zinnes, Ya ir Eilat, and Jeffrey Sachs

Figure 1. Inter- and Intra-Cluster Variation of COT Indicator of Privatization over the Transition Cycle and for 1998, Respectively

I. EU Border S ta tes 2. The Balkans 3. The Baltics 1.5 1.5

0 0 .... 0 -I -1 ;;;0 ;a c �"'"-7...... ,.�--'-:�-:-'-::-' -1.5 2 3 4 567 f:... Q � 5. Western FSU 6. The Caucasus 7. Central Asia 1.5 1.5 1.5

0 0 0

-1 -1 -1

-1.5 '----�-�----l -1.5 2 3 4 5 6 7 2 3 4 5 6 7 2 3 4 6 7 Years since u·ansition

2 f-H UN - . , SVK

L I RUS i KG� MKD LVA GEO KA 00 I 0\ IHR .= AlM 0\ � ROM MPA [ .: 0 ;;; �r 0 l11� 0 1- AZE :0 .:: h TJ Q u -I I- TKM BLR

-2 �------�'------�'------�'�------�'�------L-'------L-J' 2 3 4 6 7 Years since transition

Source: Sachs, Zinnes. and Eilat (2000a). Notes: Upper panel: Year transition began: 1990: Bulgaria. Czech Republic, Hungary. Poland Romania. Slovakia. 1991: Croatia. Macedonia, Slovenia. 1992: Am1enia, Azerbaijan, Belarus, Estonia, Georgia, Kazakhstan, Kyrgyz Republic, Lithuania, Latvia. Moldova. Russia. Taj ikistan. Turkmenistan. Ukraine, Uzbekistan. Dato Exclusions: Macedonia was excluded from the Balkans for the figure above due to lack of data on COT. EU border is missing transition year 9 since observations for Croatia and Slovenia arc missing. Lower panel: Symbols: Hollow square: average of the cluster for 1998. Horizontal line: average of the entire sample for 1998. Country codes: ALB-Aalbania, ARM-Armenia, AZE-Azcrbaij an, BGR-Bulgaria. BLR-Belarus, CZE-Czecb Republic, EST-Estonia, GEO-Georgia. HUN-Hungary. HRV-Croatia, KAZ-Kazakhstan. KGZ-Kyrgyz Republic. LAV-Latvia, LTU-Lithuania, MDA-Moldova. MKD-Macedonia. POL-Poland. ROM-Romania, RUS-Russia, SVK-slovakia. SVN-Slovenia. TJK-Taj ikistan. TKM-Turkmenistan. UKR-Ulcraine, UZB-Uzbekisran. Cluster numbers: See Table I.

154

©International Monetary Fund. Not for Redistribution THE GAINS FROM PRIVATIZATION IN TRANSITION ECONOMIES

Figure 2. Inter- and Intra-cluster Va riation of OBCA (Firm Incentive) Indicator of Privatization over the Transition Cycle and for 1998, Respectively

I. EU Border States 2. The Balkans 3. The Baltics 2 2 2 � 0 0 0

9"' -I -I -I () :0 c: -2 -2 2 2 3 4 5 6 7 8 2 3 4 5 6 7 8 2 3 4 5 6 7 E-. 0 (.) 5. Western FSU 6. 111e Caucasus 7. Central Asia 2 2 2

0 0 - 0 � -I -I � -I

-2 -2 2 2 3 4 5 6 7 2 3 4 5 6 7 2 3 4 5 6 7 Years sinceI transition

HUN 00 0\ 0\ 2 i.: 0 H V

<:;() S N BGR :0 c: SVK � !s MKD (.) 0 '�

-2 �------�------�------L------��------�------� 2

Source: Sachs, Zinnes. and Eilat (2000a). Notes: Upper panel: Year transition bega11: 1990: Bulgaria. Czech Republic. Hungary. Poland Romania. Slovakia. 1991: Croaua. Macedonia, Slovenia. 1992: Annenia, Azerbaijan. Belarus. Estonia, Georgia, Kazakhstan. Kyrgyz Republic. Lithuania. Latvia. Moldova. Russia. Tajikistan, Turkmenistan. Ukraine, Uzbekistan. Dala Exclusions: EU border is missing transition year 9 since observations for Croatia and Slovenia are missing. The Balkans are missing transition years I and 9 since Macedonia gained independence only in its second year of transition. Lower panel: Symbols: Hollow square: average of the cluster for 1998. Horizontal line: average of the entire sample for 1998. Country codes: ALB-Albania, ARM-Am1cnia. AZE--Azerbaijan. BGR-Bulgaria, BLR-Belarus. CZE-Czech Republic. EST-Estonia. GEO-Georgia, HUN-Hungary. HRV-Croatia. KAZ-Kazakhstan, KGZ-Kyrgyz Republic. LAV-Latvia. LT U-Lithuania, MDA-Moldova. MKD-Macedonia, POL-Poland. ROM-Romania, RUS-Russia, SVK-Siovakia. SVN-Slovenia, TJK-Tajikis ta n. TKM-Turkmenistan. UKR-Ukraine. UZB-Uzbekisum. Cl uster mtmbers: See Table I.

155

©International Monetary Fund. Not for Redistribution Clifford Zinnes, Yair Eilat, and Jeffrey Sachs

The second economic performance measure is an indicator of foreign revealed preference on the quaHty of a country's environment for economic activity. Here we construct two related measures. The first, FD!pop, is constructed by dividing foreign direct investment by total population, both from EBRD (1999). The second, FD!re/, has the same numerator but uses pretransition GOP in 1989 at purchasing power parity (from de Melo, Oenizer, and Gelb, 1995) as the denomi nator. We deflate by population and by 1989 GOP to provide two perspectives on what might be comparable indicators of foreign investor activity across countries. In the FD!pop regressions we include !NCpc89 (de Melo, Oenlzer, and Gelb's per capita income in 1989 at purchasing power parity) as an additional explanatory variable to reflect the fact that higher income countries generally attract more foreign direct investment. The last economic performance measure, EXPrel, refers to exports (as reported by the balance of payments statistics) and proxies a country's interna tional competitiveness. This has been deflated by GOP in 1989 (again, de Melo, Oenizer, and Gelb, 1995). In these regressions we use LogPOP (log of population) to capture the fact that small countries are more export intensive than big countries. While each of these performance measures is a highly imperfect measure of a country's true economic performance, our hope is that taken together, they provide a more realistic window into what is actually happening in these countries.

IV. Is "Change of Title" Enough?

Perhaps the most straightforward test of the Wa shington Consensus-that change-of-title per se yields economic performance gains-is to place change-of title (COT) in regressions with performance measures as dependent variables. Consider the equation:

PERFi.t = 81 COTi.t + 82 COTi.t- l + 83 REFi.r

(I) where the i and t subscripts are for country and year, respectively, the 8 parame ters are to be estimated, and "(;,1 is the regression's error term. PERF stands for our five performance measures described in Section ill, namely, IGDP, Log!GDP, FD!pop, FD!rel, and EXPrel. REF measures other reforms. The k, j, and m are summation indexes over six clusters, eight transition periods, and three macrosta bilization periods, respectively. CLUST(k); are dummy variables for each of the clusters. For example, CLUST(k); is equal to one if country i belongs to cluster k and it is zero otherwise. These capture our beliefs about the importance of initial conditions. TrYEAR(j);,, are dummy variables for years since the start of transi tion. For example, TrYEAR(});,, is equal to one if t is the jth year of transition for country i, and it is zero otherwise. These capture our belief that systemic trans-

156

©International Monetary Fund. Not for Redistribution THE GAINS FROM PRIVATIZATION IN TRANSITION ECONOMIES

formation, population expectations, learning-by-doing, and other factors cause countries to fo 1low an adjustment process linked to the years since transition began. STAB(m);., comprises three dummy variables that capture the impact of macrostabilization. STAB(!);., is one for country i during the flrst two years of macrostabilization and it is zero before and after; STAB(2);,1 is one for years three through flve after macrostabflization and zero before and after; STAB(3)u is one for the sixth year and beyond of macrostabilization and zero otherwise.11 Z repre sents other variables we use as controls, as described in the previous section. Table 2 provides estimates of the regressions for the alternative specifications implied by equation (1) for the panel of 24 countries from the start of transition through 1998. Regardless of performance measure used, the results are similar. We find that the level of reforms contributes to recovery and performance across most specifications, though this effect is somewhat muted once stabilization dummies are included in the regression. 12 The main result here, however, is a negative one: change-of-title does not seem to have a significant impact. This suggests that change-of-title alone is not enough to generate economic perfor mance gains.IJ

V. Complementary Reforms to Deepen Privatization Gains

Given the tenor of the paradigm debate as described at the start ofthis paper, the results of the previous regressions may come as no surprise. The literature suggests institutions, in the broad sense, as the leading candidates for the missing elements. These include those institutions related to prudential, regulatory, and budgetary authorities. For this reason and as described in the introduction, we created our OBCA variable to capture the firm's management obj ective, corporate governance, shareholder protection, and the hardness of the firm's budget constraint. To test the importance-of-institutions hypothesis, we add OBCA to equation (1) to get

PERFu = !1 COT;,r + f2 OBCA ;,r + /3 REF + ... + Yi.r , (2)

where, as before, PERF refers to each of the five performance measures and the

"... " refers to Z aod the dummy variables CLUST(k), TrYear(j ) , and STAB(m). The fsdenote parameters to be estimated. Table 3 provides the regression estimates for alternative specifications of equation (2). Two conclusions can be inferred from these regressions. First, regardless of the performance measure, the introduction of OBCA does not change the fundamental result that change-of-title has little effect on achieving privatiza tion gains. Second, for most specifications, OBCA has a weak positive effect on

IIAn alternative is to use log of inflation as a proxy for macro stabilization (see Fischer, Sahay, and vegh, 1996). 12When we decompose rherefonn indicators imo its components, we find that capital market devel opment has the strongest positive effect on perfonnance. 13Though not reported here, these results do not change when we replace contemporaneous COT by COT lagged for one or two years.

157

©International Monetary Fund. Not for Redistribution � Table 2. Does Change-of-Trtle Alone Generate Gains from Privatization? (J1 ()) Regressoni il b c d e I g h j

De{H'IIdenr W1riable JGDP JGDP JGDP Log/GDP FD/pop FDlpop FD/rel FDlrel EXPrtl EXPrtl

COT 0.714 0.306 1.481 0.009 -{).016 -2.197 0.048 -{).227 -{).227 -{).235 (0.-143) (0. 181) (0.899} (0.345) ( -{).003) (-{}.427J (0.07) (-{}.339) (-{}.302) (-{}.303)

REF 5.122** 4.917** 4.43 1** 0.082** 15.83** 10.91 2.104** 1.482 -0.426 -0.628 (2.422) (2. 253} (2.04) (2489) (2. 243) (1.107) (2.248) ( 1.1) (-{). 364) (-0.502)

INCpc89 0.001** 0.008*** O.OOl*** (2.109) (3.161J (2. 757) Log POP -3.033*** -3 .�··· (-4.331) (4.413)

Cluster dummies yes yes yes yes yes yes yes yes yes yes Tra nsition year dummies yes yes yes yes yes yes yes yes yes yes Slabllizalioo years dummies �es yes yes yes !:!) 0 Number of observations 179 179 179 119 178 178 171 171 166 166 :+ 0.709 0.728 0.717 0.711 0.426 0.448 0.370 0.393 0.659 0.673 0 Adjusted R2 ::J 0. Not.es: The numbers in paremhesis represent robust t-stalislics after White conection. *. **. and ••• represent 10. 5. and I percent significance. respectively. < 0, (2000a). 1992-95. (!) COT: Change of ti tle indic-atoc. Units: mean variance I. Source: Sachs, Ziones, and Eilat E'l:clusions: Macedonia: :::;: REF: Indicator of progress in reforms. including tax. price/wage liberalization, social safety net, capital markets, banking. Units: mean 0, \'ariance I. S<>urce:

Regre.ssion a b c d e f g h j

Depe11denJ \'arioble IGDP IGDP IGDP l»gJGDP FD/pap FD/pop FD/rel FD/rel EXPrel EXPrel

COT 0.314 -0.136 1.103 0.001 -1.831 -3.892 -{).206 -{).475 -0.292 -{).284 (0.189) (�. 078} (0. 652) (0.059) (�.367} (�. 741) HJ.297J (�.685) (�388) (�.365)

OBCA 1.871 1.891 1.714 0.033 7.365* 7.063* 0.982 0.94* 0.135 0.1-16 :r (1.4} (1.411} (1.3Tl} ( 1.431) (1.619) (1.66/) (1.595) ( 1.625) (0.187) (0.187) m G) REF 5.274** 5.361** .t.679** 0.084** 13.42 9.02 1.745* 1.235 -0.26 -0.419 )> (2.399) (2.34) (2.112) (2.471} ( 1.721, (0.813) (1.656) (0.813) (�.187) (�.284) z (J) 'Tl INCpc89 0.001*"' 0.008*** 0.007** ;v (2.035) (3.089) (2.679) 0 Log POP -3.08*** -3. 124**" � "t:J (-4.376) (-4..114) ;v Cluster dummies yes yes yes yes yes yes yes yes yes yes � '-l Transition year dummie:. yes yes yes yes yes yes yes yes yes yes Slabilization year dummies yes yes yes � yes 0 Number observations 173 173 173 173 172 of 172 165 165 162 162 z Adjusted 1(2 0.713 0.738 0.72::! 0.712 0.426 0.#7 0.367 0.389 0.655 0.669 z -t Notes: The numbers in parentheses represent robust /-statistics after White correction. *. **. and **" represenl 10, 5. and I percent significance, respectively. ;o )> COT: Change-of-title indicator. Unit:.: mean 0. variance 1. Source: Sac�. Zinnes, and Eilar (2000a). Exclusions: Macedonia: 1992-95. z OBCA: Indicator for degreebageocy" issuesunder control. including management objeche function, hardness of budget con:.ttaint. abilit} of owners to control U> t ::::t and monitor managemenL Units: mean 0. variance I. Source: Sachs. Zinnes. and Eilat (2000a). Exclusions. Ukr:tine. Armenia: 1992: Georgia, Tajikistan: 1992-93. 0 REF: Indicator of progress in reforms, including tax, price/wage liberalization. social safety net. capital markets. banking. Units: mean 0. variance 1. Source: z Sachs,Ziones. and Eilnt (2000a). m 0 IGDP: Index: of real GOP. 1989= 100. Source:(19 EBRD 99). 0 LogJGDP: log transformation of IGDP. z FDlpop: FDI in 1995 U.S.S I population. Source: EBRD ( 1999). 0 FDJr el: FDI in 1995 U.S.$ I purchasing power parity-adjustedi nc ome in 1989. Units: Percent. Source: EBRD (1999) andde Melo. Denizer, andGelb (1995). � m IMF (1999). Exclusions:Azerbai jan all years. (J) EXPre/: E.xpons (from balance ofpayments} in 1995 U.S.$ x 1000 I purchasing power parity-adjusted inoome in 1989. Source: WDJ (1999) and de Melo. Denizer. and Gelb (1995). Exclusions: Bulgaria. Czech Repu blic, Hungary, Poland: 1990� Russia, Slovakia: 1992-93: Romania: 1991;Ta jikistan, Turkmenistan; Uzbekistan: 1992. JNCpc89: Nati onal income percapi ta in 1989 at purchasing power parit}'. Units: 1989 U.S.$. Source: de Melo. Denizer, and Gelb ( 1995). Log POP: Log of population. Source: EBRD ( 1999). Ouster, transit ion year, and stabilization yeardummies: � explanation in text

©International Monetary Fund. Not for Redistribution Clifford Zinnes, Yair Eliot. and Jeffrey Sachs generating performance gains from privatization. These results suggest that while the effect of OBCA alone on economic performance is supportive, we should look further to substantiate its theoretical importance. Though not reported here, these results do not change when we replace contemporaneous COT and OBCA by COT and OBCA lagged by one or two years. Note that these results do not imply that privatization, "deep" or otherwise, has little impact on economic performance. Rather, they indicate that change of title or agency-related regulations each taken on its own has at best a limjted effect on economic performance. What we want to check to test the "new paradigm" is whether economic performance gains require simultaneous improvements in both COT and OBCA . We refer to such a simultaneous improvement as the "deep priva tization" effect. We can examine what other policy reforms deepen privatization impacts on economic performance by adding an interaction term to our model as fol lows:

PERFu = ht COT;,1 + h2 OBCA ;,1 + h3 REFu

+ h4 COT;/ OBCA i.r + ...+ 'Yi.r, (3)

where, as before, PERF refers to each of the five performance measures and the " ..." refers to Z and the dummy variables CLUST(k), TrYear(j), and STAB(m). The hs denote parameters to be estimated. Table 4 presents the estimation results of alternative specifications of equation (3). The strongest conclusion of these regressions is the powerful role of OBCA in support of COT economk performance improvements. Thjs synergistic effect is captured in the COT*OBCA interaction term, whkh is significantly positive across all regression specifications. To check the robustness of this result we repeated the regressions for various specifications and methods. These include using random effects and OLS models, inclusion of other quadratic terms (i.e., COT squared, OBCA squared, and COT multiplied by REF), dividing the san1ple into subsarnples by period and by geography, and replacing the cluster dummies by country dummies and the transition year dummies by calendar year dummies. The results of these exercises for the case where JGDP is used as the performance measure are shown in Ta ble 5. A sintilar exercise was done using our other performance measures and did not yield significantly different results. We also verify that these results do not change when we replace contemporaneous COT, OBCA, and COT*OBCA by one- or two-year-lagged COT, OBCA, and COT*OBCA. In all these cases the coefficient on COT*OBCA remains very significantly positive.'4 The interpretation of this strong result is that the higher the OBCA level a country bas, the more positive is the impact of an increase in change-of-title on economic performance. That is, if the change-of-title impact is positive, it will be even stronger when OBCA is higher, and if the change-of-title impact is negative,

14 lL is especially interesting to see that in the regressions where more than one quadratic term is included, not only does COT*OBCA maintain its significance, but the other quadratic tem1s do not prove to be significant.

160

©International Monetary Fund. Not for Redistribution Ta ble 4. Synerglsflc Effect of the Interaction Between COT and OBCA on Economic Perfo rmance

Regression a b c d e f g II I j

DqJendenJ Variable IGDP IGDP IGDP Log/GDP FD/pop FD/pop FDJrel FD/rel EXPrel EXPrel

COT o.n 0.175 1.525 0.007 �.385 -2.348 0.010 �-261 �.057 �-129 (0.492} (0.106) (0.955} (0.291) (--0.076) (--0.43/) (0.015) (-0.372) (--0.074) (--0. 164)

OBCA 3377-- 2855** 3.201** 0.54 1** 12.52*** 11.74-*'* 1.63 1** 1.489** 0.72 0.543 --! (2.494) (2.162) (2.484) (2.373) (2.65) (2.672) (2326) (2.270) (0.926) (0.659) I m REF 4.597** 4.729** 4.032* 0.075** 11.223 5.808 1 ...149 0.821 �.757 �.834 (j) (2.192) (2.11) ( 1.910) (2.169) (1.522) (0.561) (1.457) (0.573} (--0.543) (-- 0.567) )> z CJ) COT*OBCA 3.343*** 2.J47U 3.293*** 0.045*** 11.45*** 10.308*** 1.49*** 1.251** 1.325*** 0.956* , (4.09) (2.043) (4./11) (3.378) (3.012) (1.892) (2. 749) (2.465) (3.145) (}.8) 0;:o Tl'ICpc89 0.001** 0.008*** 0.007*** � (2.033) (3.029) (2.687) "0 ;:o Log POP -3.001*** -3.051*** � (-4.107) (-4.274) --!

Ousterdummi es yes yes yes yes yes yes yes yes yes yes � Transition year dummies yes yes yes yes yes yes yes yes yes yes 0 Stabilization year dummies yes yes yes yes z Number 173 173 173 173 172 165 165 162 162 z of observations 172 --! Adjusted R2 0.733 0.744 0.74 1 0.725 0.-152 0.463 0.391 OA02 0.669 0.674 ;:o )> z Notes: 1be numbersin parenthesis representrobust 1-statistics after White correction. *. **. and *** represent 10, 5. and I percent significanre.ly. respective CJ) =i COT: Change-of--title indicator. Units: mean 0. �·ariance I. Source: Sachs, Zinnes, and Eilat (2000a). Exclusions: Macedonia: 1992-95. 0 Indicator for degree wagency" issues control, including management objective of budget con strainl, abilit} of ownersto control OBCA: under function. hardness z and monitor rnanagemenL Units: mean 0, variance 1. Source: Sachs, Zinnes, andEilat (2000a).E."( ClU£ions: Ukraine. Armenia: 1992; Georgia. Tajikistan: 1992-93. m COT*OBCA: COT multiplied by OBCA. 0 REF: Indicator of reforms, including price/wage liberalization, social safety capital markets, banking. Units: mean variance Source: progress in talC, net, 0. I. z Sachs, Zinnes,and EJat(2000a ). 0 IGDP: Index of real GOP, 1989=I 00. Source: EBRD f 1999). � m LogiGDP: log transformation ofIGDP. (/) FD/pop : fDI in 1995 U.S.S I population. Source: EBRD (1999). FD/rel: fDI in 1995 U.S.S I purchasing power parity-adjusted adjustedincome in 1989. Units: PercenL Source: EBRD (1999)and de Melo, Deniz.er. and Gelb (1995), IMF(1 999). Exclusions: Azerbaij an all years. EXPrel: Exports (from balance of payments) in 1995 U.S.S x 1000 I purchasing power parity-adjusr.ed income in 1989. Source: WDI (1999) and de Melo, o-...... Denizer, andGelb (1995). Exclusions: Bulgaria, Czech Republic, Hungary. Poland: 1990: Russia. Slovakia: 1992-93; Romania: 1991:Ta jikistan. Thrtanenistan, Uzbekistan: 1992. lNCpc89: National income per capita in 1989 at purchasing power parity. Units: 1989 U.S.S. Source: deMelo. Denizer, and Gelb (1995). LogPOP: Log of population. Source: EBRD(1999). Cluster. transitio year. and stabilization year dummies: see n ©International Monetaryexplanation inFund. texL Not for Redistribution o- Ta ble 5. Additional Regressions for Estimating the Effect of the Interaction Between COT and OBCA rv on Economic Performance

Regression a b c d I! f g h I j k

Depe11dem \'ilriable IGDP IGDP IGDP IGDP IGDP IGDP IGDP IGDP IGDP IGDP IGDP

car -0.961 1.149 1.138 0.534 0.874 -1.151 0.82 0395 :!.448 -1.516 1.102 (0.357) (0..186) (-0. 594) (0.357) (0.274) (1.467) (-0.64) (0.716) (-0.602) (0 .729} (0. 7! 0 OBCA 6.146*** 3.357*** 2.908* 3.254** -1.693 4.368*** -5.422 5.492*** 3.67 1 *** 10.03*** 4.851*** (4..115) (2.655} (/.941) (2449) (-0. 656} (2656) (-/.531) (3.565) (3.71) (5.331) (3.677) 3 ** *** ** REF 0.37 3.606 -U77** 4.692** 2.90-+ 9286 8.131 3.587 0.371 -3 .78 2.469 a - N (-2.05) ( 1.605) (2./69) (2255) (1.067) (2.66) (2. 105) ( 1.576) (1.2/2J (0. 195} (-1..164) 5' Car*OBCJ, 3.638*"* 2.26** 4.353*** 2.876** 3.507*** 4.264*** 4.324** 3. 179*** 3367*** 4.565*** 4.107*** ::J

Number of observ ations 173 173 173 173 74 99 96 101 173 173 173

R2 0.737 0.734 0.733 0.44 0.599 0.763 0.78 0.93 0.184 0.713

Notes: The numbersin parentheses represent robust l-51arlstic) forregressions b-lc and z statistics for regression a, afterWhite correction. * = 10 perce nt significant. ** = 5 peroent significant; *** = L percent significant. -! :r: COT: Changeof tide indicator. Units: mean 0, variance I. Source: Sachs, Ziooes, and Eilat (2000a).Exclusions: Macedonia: 1992-95. m OBCA: lndicarorfor degree �agency� issues under control. including management objective fu nction, hardness of budget constraint. ability of owners to control (i) and monitor managemenL Units: mean 0. variance I. Source: Sach , Ziooes, and Eilat (2000a).Exclusions: Ukraine. Annenia: 1992:Georgia, Taj ikistan: 1992-93. }> z REF: Indicator of progress in reforms. including tax. price and wage liberalization. social safety net. capital markeiS. banking. Unirs; mean 0. variance 1. Source: en Sachs. Zinnes. and Eilat (2000a). IGDP: lnde.� of real GOP, 1989=100. Source: EBRD (1999). ,., ;:v COT*OBCA: COT multipLied by OBCA. 0 COT"2: COT squared. s OBCA"2: OBCA squared. \) ;:v COT*REF: COTmultiplied by REF. Cluster. transition years. calendaryear, and country dummies: see explanation in text. � -! Abbreviations: RE: Random Effects. FSU: Russia. the Baltics, and other countries of the form er Soviet Union. CEE.: Central and Eastern Europe. � 0 z z -! }>;:v z :3en 0 z 0m 0 z 0 s m (/)

it will be Jess negative. This latter case, where the effect of a change-of-title increase is negative, can be explained by the fact that transfer of ownership without the institutional structures in place for owners to exercise their authority simply replaces poor government control of management with weak or no private sector control. To understand these effects, we differentiate equation (3) with respect to COT:

dPERF;,, I dCOT;,, = ht + h4 OBCA;,, . (4)

The above equation shows that if h4 is positive, the higher is OBCA, and the larger is the effect of a change in COT on performance. This equation also allows us to determine the level of OBCA needed to generate a positive performance effect of an increase in change-of-title. Similarly, to determine the level of change-of-title needed to generate a posi tive performance effect of an increase in OBCA, we differentiate equation (3) with respect to OBCA :

dPERFu l d0BCA;,1 = hz + h4 COT;,1• (5)

Note tl1at by construction the sample mean (across all countries and years) of OBCA and COT is zero. Consequently, in equation (4), since the coefficient on COT (h1) is not significantly different from zero, an average level of OBCA is not enough to ensure COT has a positive economic performance gain. In equation (5) for the case of OBCA, however, since the coefficient of OBCA (h2) is statistically significant and positive, an average level of COT is enough to ensure OBCA has a positive economic performance gain. To be more precise about the effect of change-of-title on performance, we can use direct statistical tests to determine the critical levels of OBCA above (below) which an increase in COT guarantees a positive (negative) effect on performance. We do this by performing one-sided F-tests using the coefficients estimated in regression a of Table 4. To find the upper (lower) critical value, we search for the minimum (maximum) value of OBCA for which the null hypothesis that dPERF/d COT in equation (4) is smaller (greater) than zero can be rejected for a chosen confidence level. We then repeat this exercise using dPERF/dOBCA in equation (5) to determine the critical levels of COT for which OBCA has a signif icant impact on performance. The results of these tests for confidence levels of 90 and 95 percent in the case when tl1e dependent variable is JGDP are shown in Figure 3. As an example, at the I 0-percent significance level, for any country with a level of OBCA below -1.0 (i.e., one standard deviation below the sample mean across all countries and years), any change-of-title increase will cause a loss in economic performance. Similarly, at the 5-percent significance level, for any country with a level of OBCA above 0.5, any change-of-title increase will cause a gain in economic performance. While only indicative, it is interesting to inquire what countries fall into these ranges. Table 6 shows what countries fall into the 95-percent confidence level for

164

Figure 3. Va lues of OBCA (COT) Required for an Increase in COT(OBCA) to Generate Economic Performance Gains or Losses in GDP

Levels of OBCA, for which:

COT reduces -I COT effect COT improves Ar 90% confide nce: 0.5 perfom1ancc I uncertain performance • I I )lo OBCA I 0 COT reduces COT effect I COT improves At 95% confidence: -I. I perfonnancc uncertain 0.7 performance

LEVELS OF COT, FOR WHICH

OBCA reduces OBCA etTect -0.4 OBCA improves Ar 90% confidence: -1.8 performance uncertain performance

• I I I )lo COT I I 0 OBCA reduces OBCA effect OBCA improves Ar 95% confidence: perfom1ance -2.1 uncertain -0.3 perfonnance

Source: Author's estimates, using one-sided F-tes1 for coefficients estimated in regression a on Table I.

a definitive response to a change in change-of-title. The table suggests that, with the exception of Bulgaria since 1997 (and Armenia for just 1997), only the EU border states and the Baltics have high enough levels of OBCA so that increases in change-of-title are likely to generate economic performance improvements. On the other hand, with the notable exception of the Czech Republic in 1990, no countries in the EU border states or the Baltics appear to have bad OBCA levels so low such that there would be a likely loss in their economic performance from a change-of-title increase. While we do not present here an analogous table for changes in OBCA, one should nonetheless note that no country in our sample had a change-of-title level low enough to generate negative performance impacts from an increase in OBCA, even at the 90-percent confidence level. That is, OBCA may not always generate improved perfonnance in the short run, but it has not proven to do any harm.

VI. Policy Implications The analysis in tluspaper supports some recommendations for policymakers. First and foremost, they should consider carefully when recommending quick privatization if the requisite OBCA-related, legal, and regulatory institutions are

165

Ta ble 6. Years in which OBCA levels Would Have Caused a COT Increase to Lead to a Gain (loss) in IGDP. at 95 Percent Confidence level

Country Ye ars of Ye ars of Cowttry Years of Ye ars of IGDP gain IGDP Joss IGDP gain IGDP loss

Anneoia 1997 Through 1994 Lithuania Through 1993. Never Since 1996 Azerbaijan Never 1992 Macedonia Never Never Belarus Never Since 1997 Moldova Never Never Bulgaria Since 1997 Never Poland Since 1994 Never Croatia Since 1996 Never Romania 1998 Through 1991 Czech Rep. Since 1994 1990 Rus$ia Never 1996-97 Estonia All years Never Slovakia Since 1994 Never Georgia Never Through 1994 Slovenia Since 1994 Never Hungary Since 1994 Never Tajikistan Never Through 1995 Kazakhstan Never 1994 Turkmenistan Never Never Kyrgyz Rep. Never Never Ukraine Never Through 1994 Latvia 1993, since 1996 Never Uzbekistan Never Never

Source: Author�· calculation�.

not sufficiently developed and functioning. Our analysis suggests that countries in the western FSU do not meet these conditions (with the Caucasus and Central Asia borderline). Economic perfom1ance gains come only from "deep" privatization, that is, where change-of-title reforms occur in the presence of high enough levels of OBCA. Moreover, as a resull of tbeiT different initial conditions, the economic performance responses of countries to the same policies are different. In the area of privatization, these responses depend on the level of complementary reforms and on OBCA-related reforms in particular. Thus, a corollary of our analysis is that, in the case of transition countries, one size (policy) does not fit all. Policy prescriptions, therefore, should be less ideological and more tailored to the country's institutional conditions and stage of transition. While ownership matters, institutions matter just as much.

166

CategOI) Definition Effect Weight Scoring Availability* Source

Enterprire privatization Indicator Pos MOV I 0-8 Computed -f I (car: Change of Title) Large-scale privatization iode.'\ Pos 0.2 l to 4.33 (I worst' 4--S EBRD {1994-99) rn Small-scale privatization index Pos 0.2 1 to 4.33 (I worst) 4-S EBRD (1994-99) 0 Percentage of small firms privatized Pos 0.2 Percent 0-8 SmYey, WB )> Private sector emp oyme t share 0.2 Perren! �7 EBRD, WB z l n Pos U> Private sectorGOP share Pos 0.2 Percent 0-8 EBRD ( 1994-99) -n 0;;o OBCA (Priwlization Indicator Pos MOV I 0-8 Computed s: performance incentives} -o Budget constraint Indicator Pos 0.4 MOVI �7 Computed ;;o Tax arrears/average GOP Neg 0.2 Pen:ent 0--6 WB. EBRD � -f Budget subsidies/average GOP Neg 0.3 Pen:em 1-7 EBRD Bad loan�otal loans Neg 0.2 Percent 0-8 EBRD (1994-99) � Electricity tariff collection ratio Pos 0. 1 Percent 4-7 EBRD (1994-99) 5 z Likelihood of mid-sized pri\·ate Neg 0.2 0 = vel) unlikely 0-8 Survey

firmbeiDg bailed out to 4= very likely z -f Agency problems/ lndicator Pos 0.6 MOVl 4-S Computed )>;;o managemem objecti\'es Ellistence of bankruptcy courts Pos 0.1 1 =Yes.O=No 0-8 Survey z Go\emance/restru cturiog index Pos 0.6 1 to .t33 (I worsll 4-8 EBRD ( 1994-99) U> :::{ Legal system for investment index Pos 0.3 I to 4.33 (I worst) 5-8 EBRD (1994-99) 5 z n(){e the Notes: The OBCA indicator comprises the su b-indicators . ..budget' and �agency/objectives." In order to interpret the subindicator tables. first that all m categories and sub-categories of the table ha\'e weights listed in the column ··weight" and the direction of the impactof the variable on reform progress listed in the 0 0 column '"Effect."' Thesec ompr ise hi erarc hical levels. Fora given level the weights add up to unity 1. For example, in lheOBCA indicator, the weights for hardness z of -Budget constraint" 0.4 and "J\gency problems/management objectives·· 0.6 add to I. as do lhe weights of the five and three variables used within each of these 0 two subcategories. ·'Ava ilabilit} ·· summarizes years forwhich data is a\'ai lable. For example, -2" is 1992. Abbre\'iations: EBRD: Europeanior Bank Reconstruction s: and De\'elopm.eot. MOVI: Mean zero, variance I. Sur\ey: Harvard lnstitute of International Development TransitionSurvey of Foreign Institutes see Sachs, Zinnes, m (/) and Eilat (2000a). WB: World Bank Enterprise Reform and Privatization Database.

REFERENCES

Aziz, Jahanger, and Robert Wescotl, 1997, "Policy Complementarities and the Washington Consensus," IMF Working Paper 97 I 118 (Washington: International Monetary Fund).

Aghion, Philippe, and Olivier Blanchard, 1993. "On the Speed of Transition in Central Europe," Wo rking Paper 6 (London: European Bank for Reconstruction and Development).

Aslund, Anders, 1994, "Lessons of the First Four Years of Systemic Change in Eastern Europe," Journal of Comparative Economics, Vol. 19, No. I, pp. 22-38.

Balcerowicz, L., 1993, "Common Fallacies in the Debate on the Economic Transition in Central and Eastern Europe," Working Paper 11 (London: European Bank for Reconstruction and Development).

Barberis, Nicholas, Maxim Boycko, Andrei Shleifer, and Natalia Tsukanova, 1996, "How Does Privatization Work? Evidence from the Russian Shops," Jo urnalof Political Economy, Vol. 104, pp. 764-790.

Blanchard, Olivier, 1997, The Economics of TransiTion in Eastern Europe (Oxford: Clarendon Press).

Boardman, Anthony, and Aiden Vining, 1989, "Ownership and Performance in Competitive Environments," Journal of Lawand Economics, Vol. 32, No. I, pp. 1-33.

Claessens, Stijn, and Simeon Djankov, 1998, "Politicians and Firn1s in Seven Central and Eastern European Countries." Working Paper 1954 (Washington: World Bank).

Dabrowski, Marek, 1996, "Different Strategies of Transition to a Market Economy: How Do They Work in Practice?" Wo rking Paper 1579 (Washington: World Bank). de Melo, Martha, Cevdet Denizer, and Alan Gelb, 1995, "From Plan to Market: The Patterns of Tra nsition" (unpublished; Washington: Transition Economics Division; Policy Research Department, World Bank).

European Bank for Reconstruction and Development, 1999, Tr ansition Report (London: European Bank for Reconstruction and Development).

Earle, John. Roman Frydman, and Andrzey Rapaczynski, 1993. Privatization in the Transition to a Market Economy (Budapest: Central European University Press).

Fischer, Stanley, Ratna Sahay, and Carlos A. Vegh, 1996, "Stabilization and Growth in Transition Economies: The Early Experience;· Journal of Economic Perspectives, Vol. 10, No. 2, pp. 45-66.

Frydman, Roman, Cheryl Gray, Marek Hessel, Andrzey Rapaczynski, 1997, "Private Ownership and Corporate Performance: Evidence from Transition Economies," Research Report 97-128 (New York: C.V. Starr Center for Applied Economics, New York University).

---, 1998, "When Does Privatization Work?" (unpublished; Privatization Project).

Havrylyshyn, Oleh, Igor lzvorski, and Ron van Rooden, 1998, "Recovery and Growth in Transition Economies 1990-97: A Stylized Regression Analysis," IMF Working Paper (Washington: International Monetary Fund).

Havrylyshyn, Oleh, and Donald McGettigan, 1998, "Privatization in Transition Countries: A Sampling of the Literature," IMF Wo rking Paper 99/6 (Washington: International Monetary Fund).

International Monetary Fund, 1999, International Financial Statistics Ye arbook (Washington: International Monetary Fund).

168

Kolodko, Grzegorz, 1999, "Ten Years of Post-Socialist Transition: The Lessons for Policy Reforms," World Bank Development Economics Research Group Working Paper 2095 (Washington: World Bank).

Kornai, Janos. 1986, "The Soft Budget Constraint," Kyklos, Vol. 39, No. l, pp. 3-30.

---, 1994, "Transformational Recession: The Main Causes," Journal of Comparative Economics, Vol. 19, pp. 39-63.

Megginson, William, Robert Nash, and Matthias van Randenborgh, 1994, "The Financial and Operating Performance of Newly Privatized Firms: An International Empirical Analysis," Journal of Finance, Vol. 49, pp. 403-52.

Nellis, J., 1999, "Time to Rethink Privatization?" (unpublished; Washington: World Bank).

Pistor, Katharina, 1999. "Corporate Governance Issues: A Preliminary Response," private communication.

---, 200 I, "Law as a Determinant for Equity Marke t Development," in Assessing the Va lue of Law in Transition Economies, ed. by Peter Murre!J (Ann Arbor, Michigan: University of Michigan Press).

Poll!, Gerhardt, Robert Anderson, Stijn Claessons, and Simeon Djankov. 1997, "Privatization and Restructuring in Central and Eastern Europe," World Bank Technical Paper 368 (Washington: World Bank).

Roland, Gerhardt, 1994, "On the Speed and Sequencing of Privatization and Restructuri ng," Economic Journal, Vol. 104 (September), pp. 1158-68.

Sachs, Jeffrey, 1996. "The Transition at Mid Decade," American Economic Review. Papers and Proceedings, Vol. 86 (May), pp. 128-33.

---, 1997, "An Overview of Stabilization Issues Facing the Economies in Transition," in Economies in Transition: Comparing Asia and Europe, ed. by T.W. Woo, S. Parker, and J. Saclls (Cambridge, Massachusetts: MIT Press).

Sachs, Jeffrey, Clifford Zinnes, and Yair Eilat, 2000a, "Patterns of Economic Reform and its Determinants in Transition Economies: 1990-1998," CAER IT Discussion Paper 61 (Cambridge, Massachusetts: Harvard Institute for International Development).

---. 2000b, "The Gains from Privatization in Transition Economies: Is 'Change of Ownership' Enough?" CAER ll Discussion Paper 63 (Cambridge: Massachusetts: Harvard Institute for International Developme nt).

Shapiro, Carl. and Robert Willig, 1990, "Economic Rationales for the Scope of Privatization." in The Political Economy of Public Sector Refo rm and Privatization, ed. by Ezra Suleiman and John Waterbury (Boulder, Colorado: Westview Press).

Sheshinski, E, and L. L6pez-Calva, 1999, "Privatization and Its Benetits: Theory and Evidence," CAER 11 Discussion Paper 35 (Cambridge: Massachusetts: Harvard Institute for International Development).

Shleifer, Andrei, and Robert Vishny, 1994, "Politicians and Finns," Quarterly Journal of Economics, Vol. 109, pp. 995-1025.

---, 1996, "A Theory of Privatization," Economic Journal, Vol. 106, pp. 309-19.

Stiglitz, Joseph, 1998, "More Instruments and Broader Goals: Moving Tow ard the Post Washington Consensus," WIDER Annual Lectures 2 (Helsinki: United Nations University World Institute for Development Economics Research).

Vickers, John, and George Yarrow, 1990, Privatization: An Economic Analysis (Cambridge, Massachusetts: MlT Press).

169

Williamson, John, 1990, "What Washington Means by Policy Refonn," in Latin American Adjustment: How Much has Happened? ed.by John Williamson (Washington: Institute for International Economics).

---, 1993, "Democracy and the 'Washington Consensus'." Wo rld Development, Vol. 21, pp. 178-83.

---, 1997, "The Washington Consensus Revisited," in Economic and Social Development into the XXI Century, ed. by L. Emmerij (Washington: Inter-American Development Bank). World Bank, 1999, The World Development Indicators (Washington: World Bank). Zinnes, Clifford, Yair Eilat, and Jeffrey Sachs, 2001, "Benchmarking Competitiveness in Transition Economies." Economics of Transition, Vol. 9, No. 2, pp. 315-53.

170

Federalism With and Without Political Centralization: China Ve rsus Russia

OLIVIER BLANCHARD AND ANDREI SHLEIFER*

In China, local governments have actively contributed to the growth of newfirms. In Russia, local governments have typically stood in the way, be it through taxa tion, regulation, or corruption. We argue that the difference can be traced to lies in the degree of political centralization present in China, but not in Russia. In China the central government has been strong and disciplined enough to induce local governments to fa vor growth. In Russia, it has not. We agree, but with an important caveat. We believe the experience of Russia indicates that another ingredient is crucial, namely political centralization. [JEL P30, P50]

ver lhe past decade, China's GOP has grown at one of the highest rates in the world, Russia's at one of the lowest. The difference has come mostly from0 the growth of the new private sector. In China, the new private sector has thrived. In Russia, it has stagnated. Why this sharp divergence between private sector evolutions? In both coun tries, the evidence points to the importance of the behavior of local governments. In China, local governments have actively contributed to the growth of new finns (Oi, 1992; Qian and We ingast, 1997). In Russia, local governments have typically stood in the way, be it through taxaUon, regulation, or corruption (Shleifer, 1997; Johnson, Kaufmann, and Shleifer, 1997; McKinsey, 1999; and EBRD, 1999).1

*Olivier Blanchard is Professor of Economic� at the Massachusens Institute of Technology. Andrei Shleifer is Professor of Economics at Harvard University. They thank Stanley Fischer. Alan Gelb. Ed Glaeser, Simon Johnson, Ratna Sahay, Jim Snyder, and Daniel Treisman for commentS. They thank the National Science Foundation for financial assistance. 'The McKinsey study of I 0 sectors of the Russian economy, and of the specific obstacles faced by new firms in each sector, is particularly insrructive in this regard.

171

There are two main hypotheses for the attitudes of local governments in Russia. The first-can it "capture"-is that local governments have been captured by the initial rent holders, primarily by the old firms that dominated the Russian economy before the transition. In this view, local governments have worked both to generate tr�msfers to these firms, and to protect them from competition by new finns. In this first view, their hostile attitude vis-a-vis the new private sector has been deliberate. The second view-caU it "competition for rents"-is that the behavior of local governments has been instead the unintended result of adminis trative disorganization. Too many agencies have tried to extract rents from new private f�rms, making it unprofitable to create or run a private business, at least legally.2 These two lines of explanation are plausible and not mutually exclusive, but they raise the obvious question of why things have been different in China. Here again, there are two main hypotheses. The first is that the initial rent holders were weaker in China than in Russia. China started its transition from a very low level of economic development. Its agriculture did not rely on large collective fam1s, and its industry had relatively few large enterprises. Russia, in contrast, started its transition as a fully industri alized economy, dominated by large state firms and collective farms. According to this view, the potential for capture was simply more limited in China than in Russia. The second view points to the strength of the central government in China. Transition in China has taken place under the tight control of the Communist party. As a result, the central government has been in a strong position to either reward or punish local administrations, reducing both the risk of local capture and the scope of competition for rents (Huang, 1998). By contrast, transition in Russia has come with the emergence of a fledgling democracy. The central government has been neither strong enough to impose its views, nor strong enough to set clear rules about the sharing of the proceeds of growth (Shleifer and Treisman, 1999; Treisman, J999b). As a result, local governments have had few incentives either to resist capture or to rein in competition for rents. The aim of this paper is to explore this last argument, and more generally to explore the role of federalism in transition. The question is an important one. Based on the experience of China, a number of researchers have argued that feder alism could play a central role .in development (see in particular Qian and Weingast, 1997; Roland, 2000). Indeed, a new term, "market preserving feder alism," has been coined to emphasize the benefits of decentralization for Chinese growth. We agree, but with an important caveat. We believe the experience of Russia indicates that another ingredient is crucial, namely political centralization. In doing so, we echo a theme first developed by Riker (1964): for federalism to function and to endure, it must come with political centralization.

2Shleifer and Vi shny (1993) have shown how such disorganization and competition for rents leads to a much worse outcome than monopoly corruption from an organized govemmem. Based on a survey of shops in Russia and Poland. Frye and Shleifer ( 1997) have shown shops in Moscow are visited by a much larger number of inspectors and regulators than shops in Warsaw.

172

I. A Model of Federalism and Incentives

We start by writing down a modd of fe deralism and local government incen tives. The model is very simple, but it provides a convenient way to look at the facts and discuss the issues. This is what we do in the next two sections. Think of the government as having two levels: central and local (in other words, ignore for the moment the fact that there are at least three relevant levels of government in both Russia and China: central, regional, and local). Suppose each local government faces a simple choice. It can either foster growth, by limiting transfers of resources to state and former state firms and allowing new private firms to enter and to grow. Or it can kill growth, by transferring resources to old firms and/or preventing new firms from being created. Why might a local government choose the second option? Under the "capture" view, it may want to protect state or ex-state firms from competition. Under the "competition for rents" view, it may be simply unable to prevent bribes and corruption by local officials. Sorting out the relative importance of the two should be high on the research agenda but is not essential here. For our purposes, both have the same implication: no growth. Let y be the additional output under growth. With appropriate normalization, let y also stand for the additional amount of revenues available to the central and local governments under growth. Let b be the private benefits to the local govern ment of kj))ing growth. Under the capture interpretation, b may reflect the trans fers back from existing firms to the local government, in the form of bribes, cash, or in-kind payments. Under the competition for rents interpretation, b may reflect the cost to a local government of reducing or coordinating bribe taking by local officials. Now turn to the central government. Assume (an assumption to which we return later) that the central government wants to foster growth, and think of the central government as having two main tools, a carrot and a stick: • Revenue sharing (the carrot): The central government can choose the extent of revenue sharing with local governments. Let a be the share of revenues from additional growth going to local governments: if a local government chooses to foster growth. it gets ay in revenues. • Political centralization (the stick): The central government can affect the probability that the local government stays in power to enjoy either the revenues from growth or the private benefits from killing growth. Denote by Px the probability that the local government stays in power if it kills growth, and by Py the corresponding probability if it fosters growth. Define p = P/Px· The value of p clearly depends first on whether local officials are appointed or elected. If they are appointed, then presumably the central government can choose p freely and make it as high as it wants. If they are elected, the outcome depends on the ability of the central government to affect the outcome of the election, through endorsement and support of specific candi dates. If the center has little control over the outcome, and capture is impor tant, p may be less than one: the local government may be more likely to stay

17J

in power if it kills growth than if it fosters it. All it may take is for incumbent firms to be better organized politically than new entrepreneurs. Under these assumptions, the local government chooses growth if Py a y > Px b, or equivalently if

pay> b. (J)

The local government is more Likely to choose growth, the stronger the stick (the higher p), the bigger the carrot (the higher a), the larger the growth potential (the hlgher y), and the smaller the benefits of capture or the lower the costs of reining in competition for rents (the lower b). This fo rmula provides a convenient way of organizing the discussion of Russia versus China.

II. Growth, Tax Sharing, Political Centralization, and Other Issues

Before proceeding to look at the empirical evidence on the various parameters of the model, one may well want to challenge the assumption that the central government is pro-growth, or at least more pro-growth than local governments. Surely, both China and Russia provide numerous examples where the policies of the central government destroyed the economy. In the context of transition and change, however, the assumption that the central government is less likely to be captured by initial rent holders than local governments seems reasonable. Local governments are smaller relative to state and ex-state fir ms, more directly affected by the unemployment implications of closing a particular fL rm, and more likely to respond favorably to requests for transfers or protection. Central governments may be captured, as well, but not necessarily by groups opposed to growth. Capture by the "oligarchs," for example, may well lead to a massive redistribution of wealth in their favor but not necessarily to lower growth.3 Much previous research has focused on y. If growth prospects are very good, then letting new firms enter and fo stering growth is attractive. The example of Moscow, and of its mayor, Luzhkov, comes to mind here. But if growth prospects are dim anyway, the returns to allowing new business to enter and grow as opposed to protecting the old firms may be low, and y may be smal1.4 This is particularly Likely if the improvements from pro-growth policies take a long time to materi alize, and the incumbent politicians are unlikely to benefit from them. If y is small, there may be little the center can do to convince local govern ments to choose growth. Even large values of a and p may not change the inequality. A number of recent studies have provided some evidence on a, both for Russia and for China. In an econometric study of the fiscal relations between China's

3For more discussion, and a more agnostic view. of whether central or local governments are more likely to be captured, see the discussion in Bardhan and Mookherjee (2000). 4Qur normalization that government revenues move with output may not be innocuous here: y, inter preted as government revenues, may be low not because additional output is low, but because the govern ment-local or central-cannot successfully tax incremental output, which is hidden in the unofficial economy or through transfer pricing.

174

regional and central governments, Jin, Qian, and Weingast ( 1999) have found a high value of marginal a, about 0.8. (In describing the evidence, we need to distin guish between the three levels of government: central, regional, and local.) No corresponding study exists for the relations between China's local and regional governments. Wong (I 997) suggests that the nature of the contracts between local and reg ional governments, and therefore the outcomes, may be similar to those for the relation between central and regional governments. In an econometric study of the fiscal relations between Russia's local and regional govemments, Zhmavskaya (2000) has found that marginal a is only about 0. 1, giving local governments only weak incentives to increase the tax base. No corresponding study exists for the relations between Russia's regional and central government. In personal correspondence, however. Treisman has reported that estimates of a obtained from regressions with a specification similar to that of Jin, Qian, and Weingast (1999) are not lower for Russia than they are for China. In sum, the evidence is somewhat murky. At the local level, a might be some what higher in China than in Russia. It seems difficult to conclude, however, based on the available evidence, d1at differences in a are enough to explain the differ ences in the behavior of local govemments in Russia and China. Turn fi nally top, the relative probability of staying in power if pmsuing pro growth policies. ln China. the Commw1ist party has the power to appoint and dismiss governors, and it has exercised this power both to support the governors whose regions have performed well economically and to discipline governors who have followed anti-growth policies (see Huang, 1998). Perhaps as an ultimate prize, the governors whose regions perform well have been brought into the national government in Beijing. It is clear that, in China, p is a large number-if the power of the Communist party is viewed as absolute, then p is close to infinity. In Russia, governors are now elected, not appointed. The ability of the national government to reward or penalize governors through administrative and electoral support has been limited. For this reason, p in Russia is much lower than in China; it is arguably less than one. This difference in political control, rather than the difference in revenue sharing arrangements, may therefore be the reason why inequality (I) holds for China and not for Russia. For a high enough p, even a low a may sustain pro growth policies. The Chinese central govemment has allowed a substantial share of tax revenues, as well as spending responsibilities, to stay with the regions, but, given its power of appointment, it might have gotten away with a lower value of a. For the Russian government, on the other hand, there may have been no value of a that would lead local governments to foster growth. As a resulr, there may have been little incentive for the central government to maintain a high value of a anyway. lt is interesting to look at the evolution of the relation between central and regional governments in Russia in the 1990s in the light of this model. In the early 1990s, Russia's central government relied on the use of a > I for most of the reg ions. [t did this by taxing a few oil-producing regions and by using deficit finance to compensate for the resulting lack of net revenues at the center

175

(Treisman, 1999a). During that period, particularly in 1992-93, Ye ltsin also had administrative control over the governors, so p was higher than afterward. As Treisman (l999a) shows, this policy worked tolerably weU for a while and bought the center peace with the regions. In the mid- to late 1990s, that equilibrium feU apart. First, stabilization poli cies forced the central government to reduce a for most regions, as the central deficit had to be reduced and hence large transfers to the oblasts became unaf fordable. Second, political decentralization and party-free gubernatorial elections reduced p significantly. Third, the continued recession reduced at least the expec tation of y. In Russia's federal structure today, equation (I) fails, in part because a might be low but also, and more importantly, because p is low. To return to Riker (1964), the peripheral.ized federalism that characterizes Russia today may simply not be sustainable. Our focus on the role of political parties in Russia achieving-or not achieving-political centralization may be too narrow. In Russia, two forces have become at least partial substitutes. The first is the national media. Media compa nies in Russia are private and controlled by interests closely tied to political move ments, particularly the government and the center-left opposition. The media groups used television and newspapers aggressively to get their preferred candi dates elected in the 1996 presidential election (when both supported Ye Jtsin) and especially in the 1999 Duma elections. The second centralizing force has been the energy monopolies, especially Gazprom and United Energy Systems. The first holds monopoly over the supply of gas in Russia; the second controls the elec tricity grid. Both companies, while nominally private, have been close to the government. Both have been used by the government to provide cheap energy, as well as energy without payment, to cooperative regions. In this way, both have been used to make the conduct of regional governments more responsive to the needs of the center. How efficient or desirable these substitutes have been, however, is an open question.

Ill. Discussion and Implications Our analysis has a number of implications for China, for Russia, and for the economic theory of federalism. With respect to China, our analysis implies that, to the extent that fe deralism has played a helpful role in promoting China's economk growth, such federalism relied crucially on the centralizing role of the Communist party.s If the Communist party, as it yields power in the future, is not replaced or supplemented by other national parties that influence the appointment or the elec tion prospects of governors, p will fall, leading to greater rent seeking and lower efficiency in its federal an·aogements. The message of our analysis for China is clear: the competitive benefits of "market preserving fe deralism" emphasized by China scholars depend very much on political centralization.

SEven in China, control by lhe center is not absolute. Young (2000) argues, for example, that political centralization has not prevented regional governments from crccling lrade barriers between provinces.

176

With respect to Russia, our analysis suggests that federalism has failed precisely because of political decentralization. There is no question that carefully designed tax and other fiscal policies can raise a in Russia. These policies would require a clearer division of tax bases between the central and the regional govern ments, as well as a division of tax collection and spending responsibilities that does not exist today.6 Nevertheless, given the low level of political centralization, such fiscal measures may not be enough to induce local governments to foster growth. Will centralization come, and if so what form will it take? In principle, central ization in Russia could come through the creation of national parties that exercise influence over the governors needing their support in elections. Centralization may alternatively take the form of greater administrative control over governors through more aggressive bargaining over issues that bear on their regions and remain under the control of the center, such as the allocation of electricity and gas. Centralization may also involve the suspension of the democratic process. Presumably, a turn to a competitive national party system is more compatible with political freedom than are the alternatives. Yet some form of centralization is prob ably necessary for the federal equilibrium in Russia to change-for a switch in the sign of inequality (1).7 From this perspective, Mexico provides a very instructive, though not in every way appealing, example. In the 1920s and 1930s, the Mexican economy presented a far more extreme version of peripheraJized federalism than Russia presents today (Diaz-Cayeros, 1997). Following the revolution, the Mexican states were each run by their own dictator, or cacique, who controlled the regional sources of military power, collected the regional taxes without remitting them to the center, and erected trade and other barriers against other Mexican states. The result was fiscal disorder and economic stagnation. In 1938, President Calles, with the support of the military, transformed the Pa1tido Revolucionario Institucional (PRJ) into a national hegemonic party. He convinced the regional leaders to join the party and to adhere to its national policies with an offer of a carrot and a stick. The carrot was a promise of long, secure, and profitable careers under the protective wing of the PRl, which would gain control over the nomination (and effectively the election) of governors. The stick was a threat of personal violence against the caciques who declined to join. Nearly aUjoined a few who did not were kmed. As part of this deal, the central government obtained fu ll centralized control over tax collection in Mexico, leaving the states to rely on transfers from the center, as well as control over trade and regulatory policies that turned Mexico into a common market. The economic benefits of the 1938 arrangement were large, but they did come with tremendous political centralization. It may have been "market preserving federalism," but the political market was not the one that was preserved.

6See Shleifer and Treisman ( 1999) for one such proposal. 7This paper was written in December 1999, before President Putin came to power. His regional poli cies, including those toward governors, have been indeed very much focused on raising p.

177

The last example leads us to the implication of our analysis for federalism in general. This implication is not new, and draws on Riker (1964), yet it has been neglected in the recent discussions of China praising the decentralization benefits of fe deralism. As best we can tell, the economic benefits of decentralization obtained from fe deralism rely crucially on some form of political centralization. Without such centralization, the incentives to pursue regionalist policies are too high, and cannot be eliminated solely through clever economic and fiscal arrange ments. It is possible that a federal country can "muddle through" without political centralization, as Russia bas done in the 1990s and Brazil and India have done for longer, but some political system of aligning the interests of national and regional politicians is needed to get beyond "muddling through."

REFERENCES

Bardhan, P., D. and D. Mookherjee, 2000, "Capture and Governance at Local and National Levels," American. Economic Review, Papers and Proceedings, Vol. 90 (May), pp. 135-39. Diaz-Cayeros, A., 1997, "Political Responses to Regional Lnequality: Taxation and Distribution in Mexico" (Ph.D. dissertation; Durham, North Carolina: Duke University). European Bank for Reconstruction and Development, 1999, Transition Report 1999: Ten Ye ars of Tr ansition (London). Frye, Tomothy, and Andrew Shleifer. 1997, "The Invisible Hand and the Grabbing Hand," American Economic Review: Papers and Proceedings, Vol. 87 (May), pp. 354-58. Huang. Y., 1998. "The Lndustrial Organization of Chinese Government," Working Paper 99-{)76 (Cambridge, Massachusetts: Harvard Business School). Jin, H .. Y. Qian, and B. Weingast. 1999, "Regional Decentralization and Fiscal Lncentives: Federalism, Chinese Style" (unpublished; College Park, Maryland: University of Maryland, Nobel Symposium on Transition). Johnson, Simon, Daniel Kaufmann, and Andrei Shleifer, 1997, "The Unofficial Economy in Transition," Brookings Papers on Economic Activity: 2, Brookings Institution. pp. 159-239. McKinsey Global Institute, 1999, Russia 's Economic Performance (Moscow: McKinsey). Oi, J., 1992, "Fiscal Reform and the Economic Foundations of Local State Corporatism in China," World Politics, Vol. 45, No. l (October) pp. 99- 126. Qian, Y., and B. Weingast, 1997, "Federalism as a Commitment to Preserving Market Incentives," Journal of Economic Perspectives, Vol. II (Fall), pp. 83-92. Riker, W., 1964, Federalism: Origins, Operation, and Significance (Boston, Massachusetts: Little, Brown). Roland, G., 2000, Transition and Economics: Politics, Markets, and Firms (Cambridge, Massachusetts: MIT Press). Shleifer, Andrei, 1997, "Schumpeter Lecture: Government in Transition.'' European Economic Review, Vol. 41, No. 3, pp. 385-410.

---, and Daniel Treisman, 1999, Without a Map: Political Ta ctics and Economic Refo rm in Russia (Cambridge, Massachusetts: MIT Press).

Shleifer, Andrei, and Robert W. Vishny. 1993, "Corruption,'' Quanerly Journal of Economics, Vol. 108 (August). pp. 599-6 17.

178

Treisman, Daniel, 1999a, After the Deluge: Regional Crisses and Political Consolidation in Russia (Ann Arbor, Michigan: Uruversity of Michigan Press).

---, 1999b, "Decentralization, Tax Evasion, and the Underground Economy: A Model with Evidence from Russia" (unpublished; Los Angeles, Califorrua: Department of Economics. University of California). Wong, C.. ed. 1997, Financing Local Government in the People's Republic of China (Hong Kong SAR: Oxford University Press). Young, A., 2000, "The Razor's Edge: Distortions. Incremental Refonn and the Theory of the Second Best in the People's Republic of China," Quarterly Joumal of Economics, Vol. 115 (November), pp. I 091-135. Zhuravskaya. E., 2000, "Incentives to Provide Local Public Goods: Fiscal Federalism. Russian Style," Journal of Public Economics, Vol. 76. No. 3. pp. 337-68.

179

Falling Tax Compliance and the Rise of the Virtual Budget in Russia

BRIAN AITKEN*

The decline in cash revenue in Russia has been the key macroeconomic policy fa ilure of the transition. During 1994-98, a shatp deterioration in cash compli ance was offset by a rise in non-cash revenue, as the government increasingly financed its spending through mutual arrears write-ojjs. This paper argues that the fall in cash compliance emerged when money printing was replaced with a method of budget financing that did not, in the short run, compromise the government's goals of low inflation, a stable exchange rate, and low interest races, but which ultimately has led the government into a low cash revenue trap. [JEL: H26, H30, E62, E65, 023]

n ussia's macroeconomic policy has been dogged for years by the federal I �government's fa ilure to collect enough cash revenue. Cash collection has fa llen short of budget targets as an annual event. and every year the large budget gaps have been filled by spending arrears and additional short-term borrowing. The combination of growing debt and shrinking revenues naturally alarmed cred itors, and it is not surprising that confidence ultimately gave way in the summer of 1998 with disastrous consequences for the government.

What is surprising though, given what was at stake, is that the government did not act more forcefully to reverse the revenue decline. The government clearly • understood that its major achievements-low inflation and a predictable exchange rate-were increasingly jeopardized by chronically low revenue. Why then was it

*Brian Aitken is a Senior Economist at the International Monetary Fund. The author thanks John Anderson, Tom Richardson, Tapio Saavalainen. Ratna Sahay, and Jeromin Zettelmeyer for comments on a previous draft. and participants of the December 1998 European D Department seminar.

180

©International Monetary Fund. Not for Redistribution FA LLING TAX COMPLIANCE AND THE RISE OF THE VIRTUAL BUDGET IN RUSSIA

seemingly impotent to reverse the revenue decline? Did technical barriers keep revenue low, or was collecting taxes merely a question of "political will"? If the reasons were not technical, why would the government have allowed political considerations to dictate a policy that was clearly selfdestructive? In this paper, I argue that the revenue decline was brought about by a budget practice that was actually quite sensible from a short-run perspective-namely, the virtual budget, or the practice of offsetting tax and expenditure arrears. The rise of the virtual budget, however, triggered a fundamental shift in taxpayer compliance as enterprises discovered that they could benefit from tax offsets by accumulating tax arrears. Falling cash compliance further aggravated the government's need for financing, which, ironically, made offsets appear even more attractive as a short run solution. This set in motion a vicious circle where the expectation that tax arrears would be offset became a self-fulfllling prophesy. With the practice of offsets firmly entrenched, expectations could be changed, and compliance improved, only through a long and painful campaign against tax delinquents with a great short-run loss in non-cash budget financing. In this sense, the government was caught in a low cash revenue trap. By rnid-1998, it was simply too late for fresh revenue raising efforts to prevent devaluation and default. The key event setting this process in motion was the elimination of the infla tion tax, when the government stopped printing money but did not support this with adequate cuts in expenditure. Instead of more drastic expenditure cuts. the government sought to replace money printing with budget financing that would not in the short run compromise its goals of low inflation, a stable exchange rate, and low interest rates. The revenue decline is described in several stages. First, a fu ndamental change in taxpayer behavior took place beginning in 1995 as cash compliance, which had been more or less constant during 1992-94, feU sharply. Second, the elimination of the inflation tax led to the rise in the virtual budget. I argue that the virtual budget reduced the incentives for enterprises to pay taxes as well as for the govern ment to punish tax delinquents. Finally, a simple model of tax compliance is devel oped to show how these developments can be characterized as a low revenue trap. In considering the arguments in this paper, it is useful to keep in mind at least one possible alternative explanation for the revenue decline based loosely on the Gaddy and Ickes (1998) model of the virtual economy.1 In this alternative, enter prises don't choose cash compliance, except perhaps at the margin. Instead, they can be thought to pay in cash only "what they can." If during high inflation, for example, seigniorage was transferred from households to enterprises in the form of budget subsidies, then eliminating the inflation tax might have reduced the pool of real resources from which enterprises could pay taxes. Cash revenue might be expected to fall even if the measured tax liability, which is a reflection of virtual economic activity, did not. While I do not reject this alternative, I discuss some reasons why it is not an entirely persuasive explanation of the revenue decline. In addition, I discuss what implications it might have for reversing the decline.

I Gaddy and Ickes ( 1998). In their paper. Gaddy and lckes do not model tax compliance djrectly, but argue that cash revenue collection can be increased only by redirecting enterprises' cash proceeds away from other expenses such as wages.

181

I. Federal Revenue Developments

ln this section, I fo cus on tax revenue developments of the federal govern ment. For data reasons, I set aside questions regarding local government revenues (comprised of 89 separate regions) and payroll tax revenues accruing to the pension fund, social insurance fund, and other extrabudgetary funds.2 Federal cash revenue has declined sharply since 1992 (Table 1). Much of the decline during the period 1992-94 is explained by explicit changes in tax policy, such as a reduction in the VAT and profits tax rates and a phasing out of export duties. These changes are also reflected in a decline in the assessed tax liability, which can be estimated quite simply using data on tax arrears as assessed by the state tax service and confirmed by enterprises, including deferrals but excluding fr nes and interest penalties.3 Since tax poucy changes during 1992-94 are discussed elsewhere, I will not focus on them in this paper.4 After 1994, however, the revenue decline took on a very different character. During this period, cash collections fe ll by as much as 2lh percent of GOP despite a modest increase in the assessed tax liabiuty (see also Figure 1 ); the share of the assessed liability paid in cash-or cash compliance-fell from 89 percent in 1994 to 65 percent in 1996. Despite this fall, total revenue including non-cash receipts has been steady since 1994. The figures for non-cash revenue simply reflect the quantity of tax arrears cleared through one of the government-sponsored offset schemes, which I will discuss in more detail below.

II. Implications for Common Explanations of the Poor Revenue Performance

Before tuming to the reasons for the deterioration in cash compliance and the rise in non-cash revenue, it is important to note what the stability of the tax liability implies about many of the commonly cited explanations for the poor

2Data on local government and extrabudgetary revenues are especially difficult to interpret. Local governments are widely believed to systematically underreport their revenue, mainly because the di stri bution of federal transfers is in pan determined by local revenue shortfalls. In addition, an unknown share of revenues and expenditures reponed by local governments and some extrabudgetary funds take place through non-cash transactions such as through mutual arrears clearing operations. While the share of these operations is believed to be large and growing, the absence of a Lime series prevents cash and non-cash revenue developments from being analyzed separately. Aside from these concerns, however, there is no fundamental reason why the arguments in this paper cannot be extended to local and extrabudgetary revenues. 3Total revenue is based on official Russian treasury figures excluding privatization proceeds and proceeds from sales of stocks of precious metals. and is composed of cash revenue and non-cash revenue (tax offsets). Tax offsets are recorded as revenue in the treasury figures when the offset transaction takes place. rather than when the tax liability was incurred. Data on changes in the stock of tax arrears come from the state tax service, and exclude fines and penalties. The total tax liability as assessed by the state tax service is estimated by adding total revenue to the change in the stock of tax arrears. An •'adjusted tax liability." shown in Table I, is estimated by adjusting for differences in timing between when a taxable transaction actually took place and when a tax liability was later forn1ally incurred as part of an offset operation. 4See for example, Chapter 5 of Citrin and Lahiri ( 1995).

182

©International Monetary Fund. Not for Redistribution 1. Ta ble Russia: .Federal Revenue Developments, 1992-98 � (Percentof GDP) r r- 1992 1993 1994 1995 1996 1997 1998 z (i) Jan.-Jun. Annualized I � X Assessed taxliability (A+B) 16.6 15.3 12.9 13.4 13.6 13.9 13.5 13.4 () 0 A Revenue 16.6 13.7 11.8 12.2 11.9 12.6 11.4 10.6 s: Cash 16.6 13.7 11.4 10.8 8.9 9.-t 9.9 9.9 r--o Noo-casb (tax offsets)2 0.0 0.0 0.4 l..f 3.0 3.1 1.4 0.7 5> B. Change in stock oflaX arrearsl 1.6 1.1 1.3 1.7 1.3 2.1 2.8 ()z m Memornndum items: )> Cash as a share of assessed liabilitJ z (in percent) 90 89 80 65 68 74 74 0 � Adjusted tax liability4 16.6 15.6 13.4 13.9 13.7 13.8 I m Stock of federal tax arrears 0.0 1.6 1.5 1.8 3.0 3.9 4.8 ;;o GOP (billions of rubles) 19.2 172 61 1 1630 2256 2675 1242 2700 U5 m Sources: Russian Treasury. state service, and author's calculations. 0 tax ,., 1Based on developments. seasonally adj usted. through June 1998. � ?fax offsets are reflected in full when the operation is conducted, rather than according to tbe budget year in v.thich they are booked. lberefo re. total fe\'enue m will differ from lMF program definitions andfrom RussianMinistJy of Frn ance definitions. < lExcluding fines and penalties. � 4Adj usted tocorrect for the bias in measured tax liabilities caused b:,. ne"' laX liabilities arising fromoffset operations. � r- CD c 0 (j) !2:1 z ;;o �c );:

Figure 1. Assessed Federal Tax Liability and Cash Revenue (Percent of GOP)

All Taxes 18.0

16.0

14.0

12.0

10.0

8.0 Cash Revenue

6.0 1992 94 96 Jan-Ju n 98•

VAT Profit Tax 9.0 4.0 8.0 3.5 7.0 3.0 6.0 2.5 5.0 2.0 4.0 1.5 3.0 Cash Revenue 2.0 1.0 1.0 0.5 0 0 1992 94 96 Jun-Jun 1992 94 96 Jan-Jun 98* 98*

Excise Taxes Trade Taxes 3.0 3.5

2.5 3.0 2.5 2.0 2.0 1.5 l.S l.O 1.0

0.5 0.5

0.0 0 1992 94 96 Jan-Juo 1992 94 96 Jan-Juo 98* 98•

Sources: Russian Treasury, state tax service. and author's calculations.

revenue performance in Russia. For this, it is fir st useful to decompose cash revenue in the following way.

cash tax cash tax observed cash revenue = liability x compliance = rate x tax base x compliance

Here cash compliance is taken to mean, by defiilltion, the share of the tax liability as assessed by the tax service that is paid in cash. Changes in cash revenue will reflect changes in tax rates, the observed base (as measured by the tax service), or cash compliance. Anything that affects tax rates or the observed base

184

©International Monetary Fund. Not for Redistribution FALLING TAX COMPLIANCE AND THE RISE OF THE VIRTUAL BUDGET IN RUSSIA wilJ also affect the assessed tax LiabiJjty. For example, a reduction in the value added tax (VAT) rate, an increase in VAT exemptions, or an increase in under reporting of econornic activity will all lead to a decHne in the assessed VAT liabiHty. If we see that the assessed VAT liabiljty has been stable, then none of these factors could cause a decline in overall cash revenue. In fact, there is a long list of cormnonJy cited reasons for the poor revenue performance in Russia that fail to explain the recent revenue decline. Many of the reasons focus on a shrinking of the measured tax base, either due to exemptions and tax concessions or through tax evasion related to underreporting of production and the growth of the informal economy.5 But, in order for these factors to explain the revenue decline after 1994, they would require a corresponding fall in the measured tax liability. Likewise, changes in tax policy also frul to explain the revenue decline after 1994. As is clear fTom Figw·e 1, for example, the Liability for some taxes, such as export taxes and the profits tax, declined after 1994. These changes, however, were offset by an increase in the liability for other taxes such as excises and VAT. More important for the decline in overall cash revenue, cash compliance for each of the major taxes worsened beginning in 1995. Finally, widely recognized tax administration problems such as the complexity of the tax system and a dysfunctional state tax service could have played a role in the revenue decline only insofar as they explrun why assessed taxes were not collected. A corrupt, poorly focused, poorly trained, and poorly funded tax service, an adrrurustratively complex VAT and profits tax, a lack of established audit procedures, and a proliferation of transfer pricing might have hindered correct tax assessment and contributed to tax evasion, but revenue devel opments show that problems assessing tax liabilities cannot be driving the revenue decline. To summarize, there are many factors that might be used to argue why revenue was not higher than it in fact was, or why Russia has failed to achieve its full revenue potential. But the stable tax liability implies that the fall in taxpayer compliance must be the driving force behind the revenue decline after 1994.

Ill. The Rise in the Virtual Budget One of the notable features of the decline in cash compliance is that it was more or less offset by a rise in non-cash revenue, but this does not mean that the government simply began allowing enterprises the choice of paying taxes in cash or in kind. In fact, I argue that the decline in cash compliance was caused by the government's desire to mruntain expenditure following the elimination of the inflation tax. I make this argument in two stages. First, in this section, I show that the rise in non-cash revenue, and correspondingly non-cash expenditure, emerged when the government substituted one form of soft financing-money printing for another-tax offsets. Second, I argue that the virtual budget, or the practice of

5See. for example, OECD (1997), Ickes. Murrell, and Ryterman (1997).

185

carrying out expenditures and collecting revenues in non-cash fonn, created an environment in which enterprises benefited greatly by reducing their cash compHance. Throughout the period 1992-97, cash spending fe ll well short of the budgeted level (Table 2). Since spending commitments were made based on the budgeted limits, the shortfall in cash spending led directly to an increase in federal govern ment spending arrears. Pdor to 1995, these arrears were mostly cleared in cash within the spending limits of the following year. At that time, clearing arrears was not especially difficult since (1) high inflation significantly reduced the cost to the budget in real terms of clearing arrears in the fo llowing year; and (2) spending in excess of identified financing was generously covered by inflationary money creation. By the end of 1993, for example, the federal government had run Rub 4 billion in spending arrears, or about 2.3 percent of 1993 GOP, which were to be cleared within the spending limits of the 1994 budget. These arrears amounted to Jess than ¥3 of I percent of GOP in 1994. High inflationary financing aiJowed the government to run arrears each year and at the same time keep the real stock of arrears from growing. Budget practices shifted in 1995, when the government strengthened its efforts to control inflation by sharply reducing monetary financing of the budget (Table 2).

Table 2. Russia: Shortfall in Federal Government Expenditure Relative to Budget, 1992-97 (Percent of GDP)

1992 1993 1994 1995 1996 1997

Non-interest expenditure

Budget 20.6 22.7 14.9 16.3 16.5 Cash execution 26.0 18.3 20.8 12.6 11.2 11.7

Cash shortfall relative to budget 2.3 1.9 2.3 5.1 4.8

Memorandum ir.eros: Monetary financing of the federal budget 6.5 9.7 2.7 2.3 1.3 Annual average inflation (in percent) 1735 875 307 197 48 15

Sources: Russian Treasury, state tax service, and author's calculations.

Keeping inflation down meant that spending arrears could no longer be cleared by printing money. To break with this practice the federal government chose to clear its stock of 1994 arrears by issuing special treasury obligations (KOs) to suppliers. The cash burden of repaying these obligations was minimized by giving suppliers the option of using them to clear outstanding tax arrears, which when cleared were booked as non-cash revenue. As a one-time measme, issuing special treasmy obligations was a convenient way of clearing past arrears. However, spending arrears continued to mount in l995; even though budgeted expenditures were reduced sharply in that year, cash spending fell short of the budgeted level by nearly 211z percent of GOP. In addition

186

to the measured shortfall, unbudgeted spending related to the war in Chechnya, the run-up to parliamentary elections in late 1995, and presidential elections in mid- 1 996 added to the cash shortfall. As spending arrears mounted, the federal government came under heavy pres sure to "clear the books" once again by offsetting these arrears against the tax arrears of enterprises. The result was a succession of mutual arrears clearance schemes, or tax offsets. The mechanics of each offset scheme differed, but in a1l cases the federal budget cleared its arrears to suppliers by issuing a claim on the governmentthat could be used to pay tax arrears.6 Each scheme was intended only as a bookkeeping exercise to clear past-and not current-arrears, but the sequen tial and ongoing nature of the schemes meant that they de facto financed current expenditures (Table 3). By 1996, over one-fourth of total federal revenue took the form of non-cash tax offsets.

Table 3. Russia: Federal Government Expenditures Financed Through Offsets, 1992-97 (Percent of GDP)

1992 1993 1994 1995 1996 1997

Expenditures financed through offsets by budget year 0.0 0.0 1.6 3.2 3.9 2.2 by date of offset transaction 0.0 0.0 0.4 1.6 3.0 3.1

Memorandum Item: Total federal non-interest expenditures 26.3 l8.3 21.2 L4.6 14.6

Sources: Russian Treasury, slate tax service, and author's calculations.

IV. The Virtual Budget and the Fall in Cash Compliance

The emergence of the virtual budget triggered a fall in cash compliance for two reasons: (1) tax offsets increased the benefits to enterprises of running tax arrears; and (2) tax offsets reduced the government's incentive to penalize tax delinquents.

Impact of Offsets on Ta xpayer Incentives

As I have noted, each of the offset schemes was intended to clear only past tax arrears and did not relieve taxpayers of the need to pay current tax liabilities in cash. Repeated episodes of "one-time," "final" offset schemes confirmed enter prises' expectation, however, that, as long as the budget was running arrears, there would be another offset scheme in the future. To participate in a future offset, an enterprise needed tax arrears.

6Sce Crony and others ( 1997) for a complete description of the various offset schemes.

187

It is clear why enterprises supplying state orders might prefer to settle their tax !labilities effectively "in kind" through mutual arrears clearing rather than in cash. Many of the enterprises supplying the government, such as in the defense sector, have few alternative markets for their goods. The prices these enterprises charge the government are commonly believed to be well in excess of the cash value of the goods in a competitive market. Selling goods to the government on credit, and then using this credit to write off tax liabilities, was one way for enterpri ses to realize profits on their less marketable goods. Even profitable enterprises, such as Gazprom, also benefited from trade in kind with the government. Pipeline constraints effectively limited the supply of natural gas to the export market where Gazprom's profits were greatest and customers paid in cash. In contrast, the cash value of supplying an additional unit of gas to the domestic market was quite low, far below the contract price that Gazprom was able to charge for supplying gas to government agencies. With low marginal costs of production, Gazprom had a clear incentive to build up tax arrears on its exports, and then clear those arrears by offsetting them against its sales to the government. While it is not surprising that enterprises that were owed money by the budget wouLd run up tax arrears, a crucial feature of the various offset schemes is that they extended the potential benefits of tax arrears to all enterprises whether or not enterprises had any direct link to the federal budget. In all the offset schemes, the enterprise holding the claim on the budget could, whether or not it was explicitly allowed, trade this claim to another enterprise wishing to use the claim to clear tax arrears. In some schemes the claims were securitized, and an active and highly lucrative secondary market quickly developed. Other schemes required chains of debtors and creditors, with the budget on one end and the tax delinquent on the other. In these cases, claims were effectively traded through bogus post-dated contracts and side agreements between firms. The main point is that offsets reduced the incentive to pay taxes in cash for all enterprises in the economy. Offsets were particularly valuable to the more politi cally exposed large enterprises. The Karpov Commission found that in their sample of the 210 largest tax delinquents, less than 8 percent of payments to the federal budget during 1996 and the first half of 1997 took the form of cash.7 Likewise, the potential gains from offsets were particularly Large for enterprises with profitable cash markets but no direct claim on the budget. As such, it is no surprise that the tax arrears of the oil sector rose sharply in advance of the widely anticipated offset exercise that took place in late 1997 (Table 4).

The Government's Response to Falling Tax Compliance

One might expect that the government, having lost the inflation tax as a source of financing, would do everything in its power to improve tax compliance. After all, the government knew precisely which enterprises owed what taxes, and the list of the largest tax delinquents was publicly known. Why then did the government tolerate the decline in compliance? This is somewhat of a paradox and critical in

7Report of the Interdepartmental Balance Commission ( 1997).

188

Ta ble 4. Russian Federation: Federal Tax Arrears of the 100 Largest Tax Debtors, by Sector (End of period stock in billions of rubles)

December July Oc1ober April Sepember December April Seprember 1995 1996 1996 1997 1997 1997 1998 1998

Auto 1.8 4.0 4.4 4.7 5.2 5.9 5.7 5.3 Coal 0.1 0.1 0.2 0.6 0.6 1.0 0.7 Energy 2.3 2.6 2.3 4.0 5.7 5.7 5.0 6.7 Gas 1.9 6.2 6.8 7.9 S.J 8.0 10.5 16.7 OiJ 5.6 10.4 12.3 JO.S 13.1 9.1 4.6 5.3 Rail 2.6 2.4 2.6 2.2 3.6 3.9 3.3 4.9 Other 0.4 0.2 0.4 1.0 0.3 0.3 0.7 3.8 TotaJ 15.6 27.2 29.6 32.7 35.5 35.6 33.5 43.2 Memorandum item: Total fcderaJ tax arrears 30 58 74 82 96 104 120 146

Sources: Russian Treasury. state taxservice, and author's caJcu1ations.

explaining why cash compliance actually fe ll. The answer is that in the short run the government found tax offsets to be a less costly alternative, from both the political and fm ancial points of view, to taking steps against tax delinquents. Enforcing cash compliance is politically costly to the government. Sanctions against tax delinquents such as seizing assets and bankrupting enterprises can be effective only if they inflict pain on owners and managers, either by stripping ownership or shutting down the operations of the enterprise. By their nature then, sanctions will involve some short-run political cost to the government; owners and managers are often politically powerful., particularly when closely allied with regional governments, and displaced workers are not likely to be sympathetic to the fe deral government's cause. But if threats against tax delinquents are to be credible, taxpayers need to perceive the government as wilJjng to carry out these threats to their conclusion despite the political costs. At the same time, the sanctions against taxpayers rarely provide any direct short-run financial benefit to the government. One potentially cost-effective sanc tion would be seizing the liqwd assets of delinquent enterprises. However, enter prises have increasingly resorted to barter and use of money surrogates, such as veksels and arrears, and as a result rarely have any liquid assets to seize.& In the Karpov Commission's sample of Large tax delinquents, for example, some 57 percent of the enterprises received less than 20 percent of their total revenue in cash form. Based on this sample, the Commission concluded that " ...it makes no sense to apply fiscal punishment measures directed at finding and arresting accounts of the majority of 'classical' industrial enterprises. Indeed, there is

8The state tax service has the legal authority to block a tax delinquent's bank accounts and force aJI current revenues into a single setLlement account that is applied to the payment of tax debts. Hendley, Jckes, and Ryterman ( 1998) discuss the increasing use of barter and other non-cash transactions as a way of evading seizure of revenues through the single settlement account system.

189

nothing there to find."9 I set aside the question of why these enterprises are cash starved, and there certainly is evidence to suggest that they are cash starved by design. In any case, seizing liquid assets of tax delinquents is rarely an effective option. As an alternative, the tax aud1orities on occasion seize non-liquid productive as sets of enterprises. For legal and political reasons, however, an enterprise's assets can only be seized piece by piece, rather than as a coherent whole. IO Moreover, until the federal debt center was established in mid-1998 d1ere was not even a centralized market outlet for the sale of seized property. For these reasons, seized assets have had very little cash value. Indeed, seizing physical assets has been such a logistical headache that the tax authorities have been known to declare assets "seized" without bothering to remove them from enterprise premises. The result has been that, regardless of the book value of the assets seized, the cash proceeds from seiZttre and sale of productive assets have been a negligible fraction of the volume of tax arrears cleared through offsets. Tax offsets, on the other hand, offered an alternative to sanctions that made much more sense to the government from a short-run perspective. Offsets were a way of capturing the entire nomJnal value of outstanding tax arrears and chan neling it toward the clearance of the already large stock of expenditure arrears of the budget. At the same time, offsets avoided the political costs associated with sanctions. From a narrow, short-run point of view, they appeared to all parties to be a rational solution to a complex set of problems. Enterprises understood this and, as a result, did not take seriously government threats that painful sanctions would be taken against tax delinquents. This explains why a number of extraordinary steps taken by the government to improve taxpayer dJscipline generally failed. In late 1996 the government estab lished the Emergency Tax Commission (VChK) to review in a high profile manner the tax liabilities of some of the largest tax debtors. Only in a few cases, however, did the commission take actions to effectively penalize the enterprises for failing to clear tax debts. In most cases, consequences such as asset seizure and bankruptcy were avoided through agreements whereby the enterprises would stay current on new tax liabilities and restructure old tax debts. When declarations of tough action were made by the VChK, they were rarely carried out to their conclu sion. In fact, the high profile nature of the YC11K's failure to follow through on its threats could well have contributed furilier to the decline in cash compliance (Crotty, and others, 1997). Several other high profile efforts also proved ineffective in the face of strong political opposition. In mld-1998, the tax service launched a public campaign to seize assets of Gazprom subsidiaries, which had been accumulating tax arrears at

9[nterdepartmental Balance Commission ( 1997). IO[n December 1997. the Emergency Tax Commission took a path-breaking decision to seize assets of several major tax deli.nquentS in the fonn of seizure of their entire "property complex." This decision, by preserving an enterprise's value-producing operation in tact. would have greatly increased the benefits to the government of seizure as well as increasing the costs to the enterprise managers and owners. For these reasons. the decision mobilized the direct opposition of powerful political lobby groups and was very soon reversed on legal grounds.

190

©International Monetary Fund. Not for Redistribution FA LLING TAX COMPLIANCE AND THE RISE OF THE VIRTUAL BUDGET IN RUSSIA

a rapid pace since the end of 1997 (see Table 4). However, the campaign mobilized widespread political opposition against the government, and in the end an agree ment was reached whereby Gazprom would pay only a part of its current tax liability in cash. As a result, Gazprom's tax arrears continued to mount in the second half of 1998. Likewise, an attempt by the government in mid-1998 to restrict export access for oil companies with growing tax arrears ran directly against the interests of powerful lobby groups, and the measure was largely circumvented in the end. The failure to enforce tax collection is also reflected in the daily deliberations regarding lower profile tax delinquents. Of over 1 ,400 tax delinquent enterprises reviewed by the regional lnterbranch Balancing Commissions during the first six months of 1998, in only 5 percent of the cases was a decision taken to seize assets. Although bankruptcy was initiated in about 19 percent of the cases, the poor state of the bankruptcy law and the quality of the courts prevent bankruptcy from being regarded as a serious threat. In contrast, in about 40 percent of the cases the tax debts were either deferred or restructured, or the balancing commission postponed a decision. It

Reasons for the Fall in Cash Compliance To summarize, the fall in cash compliance was the product of (1) expenditure arrears of the budget and (2) a shift in the method of financing these arrears from money printing to tax offsets. The method of financing contributed to the fall in compliance by rewarding tax delinquents, while the government's short-run need to finance expenditure arrears assured tax delinquents that their behavior would be tolerated. Since both factors contributed to the decline in cash compliance, simply declaring an end to tax offsets should not by itself lead to a recovery in cash revenue. So long as the government is spending in arrears, enterprises are likely to continue running tax arrears on the expectation that some form of mutual arrears clearing exercise will need to take place.

V. The Low Revenue Trap in a Model of Cash Compliance In this section, I bring together the factors that 1 have argued led to the fall of cash compliance into a coherent framework. I describe a simple game between the government-by which I am referring to the fiscal authorities-and enterprises in which cash compliance is induced by the threat that the government will seize assets (see Appendix for a complete description). ln deciding how to respond to nonpayment the government determines a seizure policy balancing the cash revenue gained from seizing assets against the alternative of using unpaid tax liabilities as an imperfect substitute for cash revenue as a means of financing expenditure needs. This model serves two modest but useful purposes. First, it describes clearly in a static framework the main argument of the paper-that a loss of cash

IIReport of the Interdepartmental Balancing Commission ( 1998).

191

financing paradoxically increases the government's willingness to tolerate lower cash compliance. The second pw·pose is to illustrate how, over time, the govern ment might find itself caught in a low cash revenue trap. Tn other words, the government's short-run goal of financing its expenditures from any means might outweigh the longer-run desire to finance expenditures from current cash revenue, leading it to postpone the costly investment of establishing a reputation for not tolerating tax delinquency. If enterprises understand this, they will challenge the government's commitment to enforce cash collection and reduce their compliance, which will in turn fwther aggravate the government's short-run financing concerns. While the model is fairly simple, it nonetheless provides a basic frame work for a richer model to examine these dynamic aspects more thoroughly.

The Government's Preferences

The model begins with the assumption that the government (again, taken to mean the fiscal authorities) wishes to finance as much as possible of a given target expenditure leveL In making this assumption, the larger question of how this level is detemlined in the first place is deliberately overlooked. This is an important question, and no attempt is made to answer it here. For the moment, one can assume that the fiscal authorities are essentially given the expenditure target, which itself is determined by larger political imperatives. Assume expenditures can be financed from three sources: (1) cash revenue collected from enterprises; (2) taxes collected in non-cash form through offsets; and (3) other sources such as seigniorage or borrowing, which are taken to be exogenous. By treating them as exogenous, I am assunling that these sources are detennined by factors outside the model, such as a given inflation target in the case of seigniorage or, in the case of borrowing, the desire to keep domestic real interest rates low and limit the growth of government debt. As with the target expenditure level, I do not intend to argue that these sources of financing are outside the control of the government but are simply linlited by other policy goals, whlch are taken to be given for the purposes of the model. The model will then allow us to consider the effects of changes in other sources of financing, for a given target expenditure level, on cash and non-cash revenue collection. A key aspect of the model's setup is the value the government places on cash revenue compared with other sources of fm ancing. All else equal, the government would prefer as much fm ancing as possible to come from cash revenue. From a long-run perspective the solvency of the government depends on its ability to finance expenditures with revenue. as opposed to borrowing or unsustainable money printing. With regard to revenue itself, the government has a clear prefer ence for cash revenue over non-cash revenue. One reason is that some expendi tures, such as wage payments and debt service, simply cannot be paid in kind or through offsets. Another is that cash is more fungible and does not distort govern ment expenditure priorities toward spending on goods that tax delinquents happen to supply. Whlle the government prefers cash revenue, it nonetheless has a desire to finance its target expenditure level, regardless of the source of financing. In other

192

words, the government would not be happy with a relatively high level of cash revenue collection if at the same time spending fell well short of the target. Likewise, it is not enough to meet the spending target if spending is financed entirely through borrow ing and non-cash revenue. One useful way to understand these preferences is in terms of the government's short-run and long-run concerns: the government bas a short-run desire to meet its spending target, but the long-run health of the budget would require 'fi nancing this spending through a greater share of cash revenue. In this model, these preferences are characterized in terms of the government's indifference curves, shown in Figure 2 (see also Appendix Figure A2). Of course in reality the government is a composite of many competing preferences, but indif ference curves are nonetheless a useful shorthand device for characterizing the trade-off faced by the government between cash revenue and total revenue. As this trade-off is modeled, the government would be indifferent to accepting lower cash revenue only if it were compensated by sufficiently higher total revenue. An important assumption is that the non-cash revenue needed to compensate the government for a lost ruble of cash revenue increases as cash revenue reaches lower levels. As a technical matter, since cash revenue is also a source of total revenue onJy the upper half of the figure is relevant for the model. With these pref erences in mind, how tax enforcement policies of the government interact with enterprise behavior to determine cash revenue will now be described. Cash revenue will depend on the level of cash compliance chosen by enter prises. The government can influence the level of cash compjjance by threatening to seize assets of tax delinquents, although seizure can be tmderstood to include any

Figure 2. Government Indifference Curves

Total revenue (Cash and non-cash)

lllifit)' = Li

utility = 11

Cash revenue

19:3

action that is painful to the enterprise managers and disrupts an enterprise's opera tions. Strictly as a practical matter, cash revenue is defined to include any cash proceeds arising directly from the sale of seized assets. How aggressive the govern ment is in enforcing compliance is reflected in its seizure policy, which is revealed to enterprises as a first step in the model. Enterprises then choose their optimal compliance based on this policy. The government fully anticipates enterprises' response when formulating its policy in the first step. As a final step, the govern ment carries out its policy based on the actual compliance chosen by enterprises.

Government Pre-Commits to a Seizure Policy

The basic framework is first developed assuming that the government can pre commit to carrying out an announced seizure policy. The main advantage to the government of pre-committing to a tougher seizure policy is that by inducing greater compliance the government reduces the number of seizures that are actu ally needed to meet a cash revenue target. That is to say, the policy announcement has a signal effect on taxpayer compliance. (Later I consider what happens if the government cannot pre-commit.) In the case of pre-commitment, enterprises choose their optimal level of cash compliance in response to the announced seizure policy. On the one hand, lower cash compliance exposes enterprises to a higher risk of having their assets seized by the government, an outcome they regard as costly. On the other hand, by paying less of their tax liability in cash, enterprises that escape seizure can settle a higher share of their tax liability in kind or through mutual arrears clearing arrangements such as offsets. Consistent with the arguments made in Section IV, it is assumed that enter prises would rather settle through offsets than in cash. Based on these considera tions, overall cash compliance chosen by enterprises will be greater the more aggressive the government's announced seizure policy (Appendix equation 4). Given its preferences, the government faces a trade-off when setting its seizure policy; in response to non-payment the government can either seize assets or use unpaid tax liabilities as an imperfect substitute to cash for financing expenditure needs. One of the assumptions driving the model's results is that the amount of cash the government can get directly from seizure-for example, from the subse quent sale of assets-is less than the amount of the enterprise's unpaid tax liability. When setting .its seizure policy, the government weighs the benefits of a more aggressive poticy yielding greater cash proceeds-both directly from seized assets as well as from the greater cash compliance that the policy will induce-against the loss of potential non-cash revenue (Appendix equation 10). The seizure policy the government will choose can be shown graphically by modifying the indifference curves in Figure 2. As noted above, enterprises will respond to a more aggressive seizure policy by paying more cash revenue. Since the government fully anticipates this response, its choice of seizure policy is effec tively a choice of the cash revenue it wishes to target. This allows Figure 2 to be recast in terms of the trade-off between total revenue and the seizure policy, where the indifference curves of this trade-off have the same properties as the curves in Figure 2. In addition. the government faces a "budget constraint," where more

194

©International Monetary Fund. Not for Redistribution FA LLING TAX COMPLIANCE AND THE RISE OF THE VIRTUAL BUDGET IN RUSSIA

seizures decrease total revenue by reducing non-cash revenue. Since the loss of non-cash revenue is partially offset by cash proceeds arising directly from seizure, the slope of this constraint will be flatter the higher the proceeds are from seizure (Appendix Figure A3). The seizure policy chosen by the government is repre sented by point a in Figure 3. An increase in cash proceeds arising directly from seizure would increase the efficiency of seizure. If the substitution effects of a more efficient seizure policy were greater than the income effects, the government would respond by announcing a more aggressive seizure policy. This is shown as a shift from point a to point b. Enterprises will react by increasing cash compliance. Within this basic framework, one result of eliminating the inflation tax would be a fall in cash compliance. The logic is straightforward. Given the government's preferences, a loss of non-revenue financing will increase the government's wilJ ingness to forego cash revenue for a higher level of total revenue, effectively flat tening the government's indifference curves. As a result, the government will adopt a seizure policy that targets a lower level of cash proceeds and greater total revenue, which is shown in Figure 4 as a shift from point a to point b. Enterprises, faced with a reduced risk of seizure, will lower their cash compliance, and cash revenue will decline.

Government Cannot Pre-Commit to a Seizure Policy

Given that seizing assets reduces total revenue, it is clearly in the govern ment's interest to threaten enterprises by announcing an aggressive seizure policy and then, once enterprises have determined their cash compliance, relax the seizure policy and convert a greater share of tax arrears into non-cash revenue for financing the budget. If the government cannot pre-commit, enterprises will

Figure 3. The Government's Seizure Policy

Total revenue (Cash + non-cash)

Frequency of seizure

195

Figure 4. Decline in Non-Revenue Financing

Total revenue (Cash + non-cash)

Frequency of seizure

understand that the real threat of seizure is less than the stated policy and will chal lenge this policy by choosing a lower initial level of compliance. I compare this result with the pre-commitment equilibrium in Figure 5. A detailed description of the comparison is included in the Appendix, but the intuition is straightforward. Whether or not the government pre-commits, enter prises will choose compliance based on the seizure policy they expect to be carried out, with compliance increasing with a higher perceived probability of seizure. This is represented by the upward-sloping curve. In the pre-commitment case, the

Figure 5. Pre-Commitment Versus No Pre-Commitment

Frequency of seizure

Enterprises· cash cnmpliance

�<;:,------+-- Government policy- p* with precommitment

c . p** ·········-····················

!...... ! b pb ...... Government policy no prccommiunem

I , Cash compliance

196

government chooses a seizure policy corresponding to point a. If the pre-commit ment is credible, enterprises will respond to this policy by choosing cash compli ance of ¢*. Ex post, however, the government would prefer to deviate from its announced seizure policy by choosing a less aggressive seizure policy corre sponding to point b. If the pre-commitment is not credible, enterprises will antic ipate the government's ex post policy preference, and will select a lower compliance corresponding to point c. Based on this compliance, the government will carry out seizures until the marginal benefit of the cash proceeds arising directly from sales of seized assets equals the marginal cost of lost financing (p**). This equilibrium is characterized by both lower compliance and fewer seizures than in the pre-commitment case.

Tax Enforcement in a Dynamic Setting and the Low Revenue Trap

The above comparison is useful in illustrating how, in a static framework, a low return to enforcing tax compliance in the short run might cause enterprises to question the credibility of the government's enforcement policy. But is this true in a dynamic framework? That is to say, what if the government regards this as a repeated game? I11 that case, the government might want to carry out its pre announced seizure policy even if this period's cash revenue benefits are very low. ln a repeated game the government would take into account the consequences of its actions this period on tax compUance next period. In other words, the govern ment might view the costs of enforcing tax compliance as a worthwhile investment in establishing its reputation, given the future benefits of a credible tax enforce ment policy. Although I have chosen for the sake of simplicity to model a one-time game, the framework can nonetheless be used to illustrate how short-run and long-run considerations might compete. If the government takes a long view, it will suffer the short-run costs of establishing a reputation. If the reverse is true, the govern ment could find itself caught in a low cash revenue trap. As shown in the Appendix, a fall in non-revenue financing, such as through the elimination of the inflation tax, increases the premium the governmentplaces on its short-run goal of financing its target expenditures by non-cash means (Appendix, Section ill). The temptation to deviate from an announced seizure policy, therefore, could increase the point that the pre-commitment itself loses credibility. At this point, enterprises would begin cha1lenging the stated seizme policy, causing compliance to fall by much more than would be the case under pre-commitment. Given the govern ment's need to finance short-run expenditures, the fall in compliance would further increase the short-run costs of reestablishing credibility. This equilibrium can be characterized as a low cash revenue trap. Enterprises pay Uttle in cash because they know the government's short-run desire to finance its expenditure needs through offsets will outweigh its long-run desire to improve cash compliance. The government. responding to low cash revenue and mounting expenditure arrears, will confirm enterprises' expectations. Attempts to break out of this trap by establishing a reputation for tax enforcement will not be credible as long as the government's expenditure needs are unchanged.

197

VI. Implications

Perhaps the key implication of the model is that the government can only break out of the revenue trap by cutting expenditures. In this way, the govern ment would eliminate its reliance on non-cash revenue as a source of fm ancing and would be free to enforce cash compliance regardless of the lost financing that enforcement would imply. Enterprises would understand this and would respond by increasing cash compliance. As the government establishes a reputa tion for enforcing cash compliance, cash revenue will increase and expenditures can be raised. Ironically, expenditures must first be cut to a lower level than would be the case if the government could pre-commit to a seizure policy from the beginning. This result is closely analogous to the reduction in real balances required to reduce inflation during a stabilization. If wages and prices are set in a forward looking manner, the government could reduce inflation without a fall in real balances if it were able to pre-commit to a reduction in the rate of money growth. The pre-commitment would translate into lower inflationary expectations, and slower wage and price growth, with real balances actually rising as the lower inflation makes money more attractive. If the policy of lower money growth is not credible, wages and prices will continue to be set based on higher expected inflation. In this case, a period of lower real balances and higher real interest rates would be required before inflation falls to the new rate of money growth. As in the case of expenditures and cash compliance, real balances must first be brought down to a level lower than is consistent with the lower inflation in the long run. A second implication of the model is that decreasing the payoffto the govern ment of seizure makes seizure less credible and increases the amount of seizure required to enforce a given level of compliance. If the substitution effects exceed the income effects (as shown in Figure 3), less effective seizure would compel the government to opt for a weaker policy of tax enforcement and a reduction in the frequency of seizure. Tllis second result is important for understanding the rise in barter and other non-cash transactions between enterprises. As discussed earlier, barter reduces the reliance on cash, making seizure a less attractive option for the government in response to non-compliance. In the context of the model, the tise in barter bas the dual effect of emboldening enterprises to reduce compliance and weakening the government's determination to fight non-compliance. It is no surprise then that the sharp rise in non-cash transactions has coincided with the fa ll in cash compliance. 12 As discussed, a possible alternative explanation for the revenue decline, based loosely on the Gaddy and Ickes (1998) model of the virtual economy, is that enter prises don't choose cash compliance, but pay in cash as much as they can out of their available resources. In this case, cash compliance can be increased only at the

•2Table 5 in Hendley, Ickes, and Rytem1an ( 1998), for example, shows a positive correlation between the severity of an enterprise's tax arrears and its reliance on baner and promissory notes in its transactions.

198

expense of some other cash requirement, such as wages. Supposing this to be true, one must still account for why cash revenue fell suddenly beginning in 1995. If cash revenue is limited only by the true resources available to enterprises, what would have caused these resources to have declined so suddenly? While any number of stylized models can be contrived to link the vittual economy to the revenue decline, one plausible explanation relates to the reduction in cross-subsidization accompanying the fall in the inflation tax. Since households hold the bulk of base money, they also paid most of the inflation tax during the period of high inflation. Part of the seigniorage is likely to have been used in one form or another to subsidize loss-making enterprises, either in the form of govern ment expenditures or through central bank cfu·ected credits to industry. In this case, the elimination of the inflation tax would represent an effective tax reduction for households, and a conesponding reduction in subsidies to loss-making enter prises. Despite the loss of subsidies, these enterprises might continue to operate as value-adding on paper through a tangle of barter anangements and mutual non payments. As such, their measured tax Liability would remain constant, even while the real resources available to pay cash revenue has fallen. Such an explanation certainly cannot be rejected outrighL and might possibly account for part of the decline in cash revenue. The key question is, which is quan titatively more important for explaining the revenue decline: (1) a reduction in cross-subsidization to loss-making enterprises; or (2) a reduction in the share of true value added paid to the government? The answer is critical for determining how to reverse the revenue decline. In the second case, as I have suggested, expen diture would need to be cut to the point that the government is not expected to rely on non-cash forms of financing, and can pursue an enforcement policy aimed at the long-run goal of raising compliance regardless of the short-run consequences. In the first case, cutting expenditure, seizing assets, and establishing a reputation would do nothing in the short run to raise cash revenue. ln fact, expenditure cuts to the extent that expenditures are a form of subsidy-might actually reduce compliance in the short run, until the loss-making enterprises are shut down and the resources they use channeled into value-adding activities. ln this regard, the case of the automobile producer AvtoVAZ is revealing. Although AvtoVAZ produces a good that enjoys a large cash market it has for some time retained the status of the single largest tax debtor in the Russian economy. And, while the enterprise is no doubt inefficient, it would be difficult to describe it as value-subtracting. AvtoVAZ pays a small fraction of its tax liability not because there are no true profits to be made in production and sales of auto mobiles, but because the enterprise has not been left with any cash flow. Enterprise managers are widely believed to have used barter and money surrogates to channel all the cash flow from the manufacture and sales of automobiles to the legally separate retail arm LogoVAZ, while at the same time using transfer pricing to shift the bulk of the tax liability to AvtoVAZ. Since LogoVAZ no doubt pays close to its full share of federal tax liability, enforcing cash compliance in the current legal environment would require shutting down AvtoVAZ. The extent to which the economy is represented by enterprises such as AvtoVAZ or by, say, value subtracting defense manufacturers is an open question.

199

Appendix I. A Model of Cash Compliance

This Appendix describes a simple game between the government and enter prises in which cash compliance is induced by the threat that the government will seize assets. In response to non-payment, the government chooses its optimal seizure policy balancing the revenue gained from seizing assets against the alter native of using unpaid tax liabilities as an imperfect substitute to cash for financing a chosen expenditure target. In this game, a decrease in non-revenue financing paradoxically increases the government's willingness to tolerate lower cash compliance when formulating its seizure policy. It increases the likelihood, moreover, that enterprises will challenge government's commitment to its policy of seizure, resulting in a further worsening of cash compliance. First I consider the outcome assuming the government can pre-commit to a seizure policy. I then compare this to the case where the government cannot pre commit, but rather adjusts its policy in response to enterprise behavior.

I. The Government Can Pre-Commit to a Seizure Policy The game can be thought to take place sequentially. In the first step, the government determines and then pre-commits to a seizure policy. As an alternative to seizing assets, the government considers allowing enterprises to use unpaid tax liabilities as an imperfect substitute for cash to finance its expenditure target, in effect allowing enterprises to pay these taxes in kind. In the second step, each enterprise determines its optimal level of compliance so as to minimize its costs, balancing the cost of paying taxes against the likelihood that the government will seize assets if it does not pay. When formulating its seizure policy in the first step, the government anticipates the optimal response of enterprises to its policy in the second step. A key assumption of the model is that the cash value of the seized assets to the government is less in ruble terms than the value of the unpaid tax liabiJity, if the government instead uses it to finance expenditures.

Enterprises ' optimal tax compliance

To solve the model, it is useful to ft rst consider enterprises' optimal tax compliance for a given seizure policy and then to solve for the optimal seizure policy taking into account enterprises' optimal response. Each enterprise i chooses a level of cash compliance to minimize its costs:

rnin¢i¢;t; + v(l - ¢;)!; + p;C;. (1)

An enterprise's total costs equal its cash compliance (¢;) times the enterprise's total tax liability (t;), plus the cost of payiilg taxes in kind (v < 1) times the unpaid tax liability, plus the probability of seizure (p;) multiplied by the cost to the enter prise of having its assets seized (C;). A crucial element of the game is how enterprises perceive the probability of seizure. One can think of the government as announcing its intention to seize

200

assets of a specified portion of the worst tax delinquents. After enterprises have chosen their tax compliance, the government will rank enterprises according to tax compliance plus some random adjustment factor, unknown to the enterprises, which might reflect factors such as the enterprise's lobby power, social impor tance, etc. The government then intends to seize assets from lowest to highest until it reaches its target number of enterprises. Although I don't explicitly model the mechanics of such a seizure policy, l capture its essence in the following formulation. Each enterprise perceives the probability of seizure to be

p; = j(p, ¢;), (2)

where p is the share of enterprises to be targeted by the government. An enterprise understands that the higher this share, the greater the chances that, for a given level of compliance, its assets will be seized. The functionf(p, ¢;) is assumed to take a form such that the probability of seizure can be charted as Figure Al. This func tional form implies that, as an enterprise's compliance falls, the chances of seizure become increasingly larger. The enterprise solves its cost minimizing equation (I) taking into account the probability of seizure for a given government policy p. From this equation, the cost minimizing

t;(l-v) = -f2 (p, ¢;*) C; (3)

Net marginal cost Marginal value of reducing = of tax payments the probability of seizure.

Enterprises will increase compliance until the net marginal cost of paying taxes in cash is equal to the marginal value of reducing the probability of seizure. The optimal compliance will be a function of the cost of paying taxes in kind, the government's seizure policy, and the cost of seizme. As a simplifying convention, enterprises are assumed to be identical. This allows the industry average cash compliance to be derived directly from the enter prise's optimization in equation (3), which is expressed as follows:

Figure Al

j{p2,$,) �J(p •• $;) .______

201

¢* = ¢{p, C, v) (4) ¢1 > 0,¢2 > 0, ¢>:! > 0, where average compliance is an increasing function of the government's seizure activity, the costs of seizure, and the cost to the enterprise of tax payments in kind. One noteworthy, although not sw-prising, implication of the enterprise's opti mization is that for a given seizure policy, overaU compliance will be lower to the extent that random factors, such as lobby power and social importance, factor into the government's decision to seize assets. In terms of the model, a greater random ness to the seizure process would be reflected in the limited ability of the enter prise to reduce its seizure ri sk by raising compliance, or in a lower value for ¢2 in equation (4). All else equal, tllis would have the effect of reducing the marginal benefit of raising compliance, and, to restore equilibrium, enterprises would reduce compliance until the marginal benefits equaled the marginal costs.

The government's optimal seizure policy In deriving the government's optimal seizure policy, start by assuming that the government (taken here to mean the fiscal authorities) wishes to finance a given expenditure target, net of borrowing and seigniorage. This net target is assumed to be exogenous in the model. It can finance this net target either by collecting cash from enterprises or by collecting taxes in non-cash form. Non-cash tax payments can be understood to take place either directly in kind or through mutual arrears clearing arrangements such as offsets. All else equal, the government is assumed to prefer cash both because it is more fungible and because some expenditures, such as wage payments and debt service, simply cannot be paid in kind or offset. The government's preferences, then, can be characterized by the fo llowing utility function:

U(E -A, R) (5)

u1 < Ou11 < 0 (6) u2 > Ou22 < 0, where E is the expenditure target net of borrowing and seigniorage, which is assumed to be exogenous, R is cash revenue, and A is total cash and non-cash revenue. Based on this utility function, the government will be indifferent between cash revenue and total revenue as charted in Figure A2. For a given expenditure target, government would be indifferent to having lower cash revenue if it had sufficiently higher total revenue. At lower levels of cash revenue, however, the amount of increased total revenue needed to leave the government indifferent is greater. The government determines its seizure policy to maximize utility:

maxp U(E- A, R) (7)

202

subject to the constraints:

A = (1 - p)T + pQ (8) R = ¢T + pQ,

where p is, as before, the probability of seizure, Tis the total tax liability, and Q is the cash proceeds the government recovers from seizing assets, which is assumed to be exogenous for the purposes of this model. Total revenue A is comprised of cash revenue paid voluntarily by enterprises and non-cash revenue for those enterprises from which the government chooses to accept payment in kind (the sum of which is (l - p) 1), plus cash proceeds from seizure. Cash revenue is determined by cash compliance plus cash proceeds from seizure.

Figure A2 A

.______R

The government sets its seizure policy anticipating the cost-minimizing reac tion of enterprises in equation (4). Combining this reaction with equations (7) and (8) and, to simplify, setting the total tax liability T equal to one, the government's optimization becomes

maxpU(E - (1 - p) -pQ, ¢(p, C, v) + p Q). (9)

The utility maximizingp (denoted p*) is that which solves the first-order condition:

u2(p*)(¢1(p*, C, v) + Q) = -ut(p*)(l - Q) ( 10)

Marginal utility of Marginal disutility of cash from seizure = net loss of total policy revenue from seizure.

The government will set a seizure policy at that point where the marginal utility of cash, both arising directly from seized assets as well as from greater compliance, equals the marginal disutility that arises from a net loss of total revenue. The component ¢1-the change in compliance with respect to a change in frequency of seizw·e---can be understood to represent the signal effect of a stated government seizure policy on cash compliance.

203

Figure A3 A

To understand the effects of the exogenous variables on the government's optimal seizure policy, it is useful to show the optimization problem graphically in terms of the tradeoff between total revenue and seizure policy p, where the policy is in effect set to target a level of cash revenue. Cash revenue R is substituted from the constraints in equation (8), combined with the enterprise reaction function in equation (4) directly into the utility function in equation (7). UtiHty can now be expressed as a function of total revenue A and seizure policy p and is maximized subject to the constraint on total revenue in equation (8). The optimal seizure policy p* that solves equation (10) is shown as point a in Figure A3. Ao increase in cash receipts from seizure (Q) will increase the efficiency of seizure. If the substitution effects of this efficiency gain exceed the income effects of higher cash revenue for a given seizure policy, an increase in Q will result in an increased effort by the government to seize, shown in the first panel of Figure A4 as a shift from point a to point b. In terms of equation (10), an increase in Q increases the marginal benefit of seizure while at the same time reducing the marginal cost in terms of lost total revenue. Increasing cash revenue by increasing the probability of seizure p will decrease the government's marginal utility of cash (u2), restoring equality in equation (10). By equation (4), cash compliance wiU increase. An increase in the government's expenditure target or a decrease in non revenue financing (an increase in E) wilJ reduce the government's willingness to seize assets. This is shown in the second panel of Figure A4. As E increases, so does the marginal disutility of the financing shortfall (u1). To restore balance to equation (10), the government will forgo cash revenue by reducing p to tl1e point where the higher marginal utility of cash revenue offsets the higher dis utility of the financing shortfall. With a less aggressive seizure policy, cash compliance of enterprises will be correspondingly lower. Finally, when either the cost to enterprises of seizure (C) or the cost of paying taxes in kind (v) increases, enterprises react to a given seizme policy by increasing compliance (equation (4)). Although cash revenue unambiguously increases, the effect on seizure policy is ambiguous, depending on the impact of these changes on the signal value of seizure. On the one band. the increase in cash revenue will reduce the marginal utility from seizure, and, all else equal, the government will

204

©International Monetary Fund. Not for Redistribution FA LLING TAX COMPLIANCE AND THE RISE OF THE VIRTUAL BUDGET IN RUSSIA

Figure A4 A A Increase in Q Increase in E

Pa Pb Pa want to reduce seizures and increase total revenue. On the other hand, if increases in C or v increase the effectiveness of seizure policy on compliance, then this will lead the government to want to increase its seizure activity. The net effect will depend on the relative size of these two effects.

II. The Government Cannot Pre-Commit to a Seizure Policy The case where the government cannot pre-commit to a seizure policy will now be considered. In the pre-commitment case, the government first determined a seizure policy as in equation (10), and then, in the second step, enterprises determined their optimal level of compliance and paid taxes according to this level. Once enterprises paid, the government then carried out its pre-announced policy. In this sequence, however, the government's behavior is not time consis tent; once enterprises chose compliance and then pay, the government no longer has an incentive to carry out its seizure policy. In terms of the model, the govern ment's optimal p. which it determines ex ante, is greater than optimal p, which it would prefer to carry out ex post. To see this, the government's optimal seizure policy is solved by taking the level of cash compliance as given. In this case, as before, the government solves

maxpU(E -A,¢ + pQ) ( 11) subject to the constraint

A == I - (l - Q)p ( 12) but now taking ¢ as exogenous. The optimal ex post seizure policy p (denoted p**) solves the following first-order condWon:

(13)

205

Comparing this to the optimal seizure policy p* given in equation (10), it is clear that since ¢1 > 0 it must be the case that p** < p* for any value of ¢. Once enterprises choose their tax compliance and taxes are paid, the government can no longer influence compliance through its seizure policy, and the marginal gain to seizure is only the additional cash proceeds from seizing assets. At this point, the government will prefer to deviate from its stated seizure policy by increasing the use of tax liabilities to reduce its financing shortfall. As can be seen from equation (14), moreover, lhe incentive to deviate from the stated seizure policy increases with compliance. As compliance increases, the marginal utility from cash revenue declines, inducing the government to seize less and finance more expenditures through non-cash revenue. The ex post optimal seizure policy, therefore, will be a declining function of compliance:

p** =p(¢) (14) p' < 0.

Enterprises, of course, can be expected to take into account the government's ex post behavior when deciding their optimal cash compliance, in effect challenging the government's stated intention to seize assets. In this case, the inability of the government to pre-commit leads enterprises to lower cash compliance further. To determine optimal compliance, enterprises now minimize costs taking into account the government's optimal ex post seizure policy determined in equation (16). As before, each enterprise minimizes

min¢;¢;!; + v( 1 - ¢;)!; + p;C; · (15)

where its own probability of seizure is now expressed as

p; = j(p**(¢),¢;). (16)

The value of ¢; (denoted ¢;**) which maximizes each enterprise's utility is that which solves

t;( l - v) = -fi (p(¢), ¢;*'.)C;. (17)

In contrast to the pre-commitment case, enterprises now determine their own compliance based on their expectation of overall compliance. While each enter prise is assumed to perceive itself as too small for its own behavior to affect overall compliance, it nonetheless knows that all other enterprises are determining their own compliance in a similar manner. Since all enterprises are identical, the equi librium overall compliance will be the point in equation (17) where ¢** = ;** = ¢. The resulting equilibrium can be compared to the pre-commitment equilib rium graphically. As in equation (4), enterprises will respond to a higher proba bility of seizure by increasing compliance. This is represented by the upward-sloping curve ¢{p) in Figure A5. In the pre-commitment case, the govern-

206

Figure AS

p ${p)

p* 1<::------0,1---- Precommitment

'· ························ · No precommitmem r i p($) ell** ell*

ment can choose a seizure policy p = p*, and the equiUbrium compliance

Ill. The Cost of Pre-Commitment

If the government views the process of seizure as a repeated game, it might be willing to endure some short -run costs of pre-commiunent because it is worried about moral hazard and is seeking to establish a reputation. However, the govern ment's cost of adhering to an announced seizure policy is itself a function of the expenditure target net of non-revenue financing. In this case, an increase in the expenditure target net of non-revenue financing might increase costs to the point that the pre-commitment itself loses credibility. As a result, compliance would fall by much more than would be the case under pre-commitment as enterprises began challenging the stated seizure policy. From equations (10) and (15), the net marginal utility lost from adhering to a pre-committed seizure policy will be

(18)

As discussed above, there is an ex post incentive for the government to forgo seizure and divert the unpaid tax liability toward "fi nancing its expenditure target. If the expenditure target (E) is small, then the pre-conunitment equilibrium will be

207

* characterized by a large p and low marginal utility levels -u1(pjand u2(p*). In this case, the marginal utility gained by deviating ex post from a seizure policy is accordingly small. If, on the other hand, E is large, the equilibrium will be char * acterized by a small p and high marginal utility levels -u1(p*) and u2(p*). Therefore, equation (18) shows that an increase in the expenditure target or a decrease in non-revenue financing would increase the marginal utility gained by deviating ex post from a seizure policy. This has the important implication that an increase in the expenditure target or a decrease in non-revenue financing can lead to a collapse in compliance and, at the same time, a reduction in the number of seizures, both by reducing the government's willingness to engage in a policy of seizure and by reducing the government's ex post commitment to such a policy.

REFERENCES

Cillin, Daniel A., and Ashok K. Labiri, 1995, Policy Experiences and Issues in the Baltics, Russia, and Other Countries of the Fo rmer Soviet Union, IMF Occasional Paper No. 133 (Washington: International Monetary Fund). Crotty, John, Katherine Baer, Richardson, Thomas and Richard Highfield, 1997, "Russian Federation: Addressing the Decline in Federal Cash Revenue Collections" (unpublished; Washington: Fiscal Affairs Department, International Monetary Fund). Gaddy, Clifford G., and Barry W. Ickes, 1998, "Russia's Virtual Economy," Fo reign Affairs, Vol. 77 (September-October). pp. 53-67. Hendley, Kathrynn, Barry Ickes, and Randy Ryterman, 1998, "Remonetizing the Russian Economy," in Russian Ente1prise Reform: Policies to Further the Tr ansition, ed. by Harry Broadman, Discussion Paper No. 400 (Washington: World Bank). Ickes, Barry W., Peter Murrell. and Randy Ryterman, 1997, "End of the Tunnel? The Effects of Financial Stabilization in Russia," Po st-Soviet Affairs, Vol. 13 (April-June), pp. 105-33. Lnterdepartmental Balancing Commission, 1997, "On the Causes of Low Tax Collection, General Causes of the Arrears Crisis, and Opportunities of Restoration of Solvency of the Russian Enterprise," Report of the Interdepartmental Balancing Commission (Moscow, Russia: Interdepartmental Balancing Commission).

---, 1998, "On the Work of the Interdepartmental Balancing Commission in 1998" (Moscow, Russia: Interdepartmental Balancing Commission).

OECD, 1997, OECD Economic Surveys, Russia.

208

dots ( ... ) indicate that the data are not available;

a dash (-) indicates that the figure is zero or less than half the final digit shown, or that the item does not exist:

a single dot (.) indicates decimals;

a comma (,) separates thousands and millions;

"billion" means a thousand million: and '·tri llion" means a thousand billion;

a short dash (-) is used between years or months (for example. 1998-99 or January-June) to indicate a total of the years or months inclusive of the beginning and ending years or months;

a slash (/) is used between years (for example. 1998/99) to indicate a fiscal year or a crop year; and

components of tables may not add to totals shown because of rounding.

The tem1 "country," as used in this publication, may not refer to a territorial entity that is a state as understood by international law and practice; the term may also cover some territorial entities that are not states but for which statistical data are maintai ned and provided internationally on a separate and independem basis.

Design: Luisa Menjivar-Macdonald and Sanaa Elaroussi