Growth in Transition: What Are the Lessons Learned?
Mark De Broeck
The initial years of transition from a planned to a market-based economy in Central Europe, Russia, and other countries of the former Soviet Union witnessed deep and long declines in output and an upsurge in inflation. As of 2000, at the end of the first decade of the transition process in these countries, growth had resumed and, with the exception of some slow reformers, inflation had dropped to moderate or low levels. Only a handful of countries, however, have returned to output levels at or exceeding those at the start of the transition. These developments have prompted extensive research on the forces driving the behavior of output in the transition economies and on the appropriate policies for achieving macroeconomic stability and sustained growth. This survey reviews recent research conducted at the IMF in these areas.1
The unexpectedly deep and lengthy output declines early in the transition, and pronounced differences in output performance across the transition economies, have been the motivation for a number of quantitative studies of the determinants of output in transition.2 These studies typically pool cross-section and time-series data and generally adopt a single-equation specification that relates a measure of output performance to a set of explanatory variables, including those representing initial conditions, structural reforms, and macroeconomic policies. While the size of the data sample, the degree of econometric sophistication, and the range of explanatory variables differ across the studies, their broad conclusions are the same—they support the view that slow progress in macroeconomic stabilization and structural reform plus unfavorable initial conditions (for example, overindustrialization or distorted trade patterns) were associated with weaker output performance. There are some differences, however, in the answers to more specific questions regarding the relative importance of these explanatory variables and the dynamic patterns of their effects.
A number of papers attempt to offer a more comprehensive explanation of output developments in transition. They include Hernández-Catá (1997), Fischer, Sahay, and Végh (1998a), Havrylyshyn, Izvorksi, and van Rooden (1998), and Berg and others (1999).3 All four papers confirm that macroeconomic stabilization, progress in structural reform, and initial conditions—in the early years of transition in particular—are key factors explaining output performance. But there are some significant differences among the more specific findings. For instance, Fischer, Sahay, and Végh conclude that a fixed exchange rate regime affected growth positively. The other papers do not find strong evidence supporting this view. As Hernández-Catá suggests, this may reflect the fact that they include inflation as a separate variable, which may capture the indirect impact of the fixed exchange rate regime on growth by helping to bring inflation down. Berg and others find that structural reform has positive output effects already from the beginning of the transition, while Havrylyshyn, Izvorksi, and van Rooden and also Hernández-Catá argue that reform first had a negative effect. All four papers also establish the significance of an aggregate measure of progress in structural reform, but obtain different results when individual components that constitute the index are included separately.
Other papers adopt the same panel growth regression approach, but focus on the contributions of specific variables.4 Cristoffersen and Doyle (1998) examine the role of export market growth and find this variable to be strongly associated with output growth in transition. They also provide evidence that inflation above a threshold rate in the low teens affected growth negatively, but that disinflation did not have a negative impact, even in the moderate inflation range. Mercer-Blackman and Unigovskaya (2000) find countries that implemented IMF programs early on and completed them successfully tended to have a better growth performance. The program implementation indicator, however, is highly correlated with an aggregate structural reform measure, suggesting that both indices could be related to an underlying, unobserved factor such as commitment to reform.
Reflecting the increasing importance attached to building effective institutions in support of the market, Havrylyshyn and van Rooden (2000) add a separate set of variables measuring institutional development to a benchmark panel growth regression. Their results retain the significance of the core macroeconomic stabilization, structural reform, and initial conditions variables, and show the institutional variables to have a significant, although small, additional effect. Abed and Davoodi (2000) focus on the role of a corruption index in explaining key measures of economic performance, including growth, in the transition economies. They argue that corruption is largely a symptom of underlying structural and institutional weaknesses and show that a structural reform variable has more explanatory power empirically than the corruption index.
The panel growth regression approach gives robust conclusions on the overall importance of a core set of variables, but also suffers from limitations and methodological weaknesses. As discussed above, studies based on this approach come to different conclusions regarding more specific issues. Also, panel growth regressions do not necessarily perform well in explaining the growth performance of individual countries. Focusing on the case of Uzbekistan—a slow reformer with a better-than-average output performance—Zettelmeyer (1998) shows that a standard panel estimation systematically underpredicts the country’s growth during the period 1992–96.5 Better results are obtained when country-specific factors are taken into account.
From a more methodological point of view, most panel studies tend to ignore the fact that policy variables are not exogenous but, rather, depend on the economic environment. They also do not take into account that some factors, initial conditions for instance, can influence growth both directly and, by affecting policy capacities and choices, indirectly. These shortcomings reflect the underlying weakness that the regression specifications generally involve a single equation and are not based on a structural model of growth in transition. Hernández-Catá (1997) is the only example in this literature of a study that derives the estimated equation from a formal model. His model focuses on the reallocation of capital from a less productive, old sector to a more productive, new sector, and shows that stronger reform efforts may result in a steeper output fall early on, but a stronger recovery later.6
The panel regression approach has been the most common framework to empirically study growth in transition, but other methodologies have been applied as well. Fischer, Sahay, and Végh (1998a) use the coefficients estimated in the benchmark Barro (1991) and Levine and Renelt (1992) growth regressions for market economies to assess the longer-run growth prospects of the transition economies.7 Fischer, Sahay, and Végh (1998b) use the same approach but focus on a smaller sample of EU accession candidates, and estimate that it could take the most advanced Central and Eastern European countries less than two decades to catch up with low-income EU countries.8 This paper also tries to quantify the income losses incurred under central planning, on average about one generation of growth.
De Broeck and Kostial (1998) and De Broeck and Koen (2000a and 2000b) use a growth accounting methodology to explain the output performance of Poland, Russia, and other countries of the former Soviet Union.9 These authors conclude that the output fall and subsequent recovery in the early years of the transition were accounted for mainly by a sharp V-shaped movement in total factor productivity rates. Based on the experience of Poland—the country with the longest record of sustained positive growth—factor accumulation and sectoral reallocation to higher productivity sectors take on increasing importance as growth becomes more robust. These insights are in line with the results obtained from panel growth regressions, which indicate that factor input movements do not have important explanatory power during the contraction and initial recovery phases. Keane and Prasad (2000) provide a political economy interpretation of the relationship between inequality, transfers, and growth in transition economies.10 Using a cross-section analysis, they confirm that progress toward establishing a market economy enhances growth, but also find that, conditional upon the degree of structural reform, policies that maintain a greater degree of equality are more conducive to growth.
In recent years, the issue of corruption—generally defined as the abuse of public office for private gain—has attracted renewed interest among both academics and policymakers. In growing recognition of the adverse impact of corruption on economic performance and its ensuing impact on the success of IMF programs, research within the IMF has contributed to the burgeoning literature in this area by highlighting the impact of corruption on economic efficiency, equity, and growth, while also providing insights into its origins, effects, and possible remedies.
The theoretical literature on the economic consequences of corruption has focused on the detrimental impact of corruption on economic growth, efficiency, equity, and welfare.1 Corruption reduces economic growth by lowering incentives to invest and can be expected to lower the quality of public infrastructure and services, reduce public revenues, misallocate talent to rent-seeking activities, and distort the composition of government expenditures and of tax revenues. Empirical evidence based on cross-country comparisons does indeed suggest that corruption has large, adverse effects on private investment and growth. Mauro (1995), in the first study of its kind, used subjective indices of corruption produced by private rating agencies to show that a country that improves its standing on the corruption index from 6 to 8—a rating of 0 being the most corrupt and 10 being the least—will experience something on the order of a 4 percentage point increase in its investment rate and a 0.5 percentage point increase in its annual per capita GDP growth rate.2 Tanzi and Davoodi (forthcoming) discuss how corruption affects growth through its impact on small- and medium-sized enterprises, which are often seen as the engine of development.3
One specific way in which corruption may hamper economic performance is by distorting the composition of government expenditures. Corrupt governments may shift spending away from productive activities (such as health and education or high-quality physical infrastructure) and toward the construction of “white elephant” projects or lower quality investments in infrastructure that are not actually needed. This conjecture has been confirmed in a large number of cross-section studies undertaken at the IMF.4 Empirical evidence suggests that corruption reduces spending on government operations and maintenance, and results in lower spending on health care and education services, such as medicine and textbooks. Higher levels of corruption also tends to be associated with rising military spending.
Corruption can also adversely affect the provision and the quality of publicly provided social services by reducing government revenue.5 Reduced quality may discourage individuals from using these services and make them less willing to pay for them.
Individuals would then engage more in tax evasion and firms would have greater incentive to participate in the underground economy, which shrinks the tax base and further diminishes the government’s ability to provide quality public services. Because corruption exacerbates an unequal distribution of wealth and unequal access to education, and other means to increase human capital, corruption can be expected to increase income inequalities and poverty.6 Hindriks, Keen, and Muthoo (1999) find that distributional effects of tax evasion and corruption are unambiguously regressive.7
The existing literature on the consequences of corruption leaves unanswered the question of why, given the high costs of corruption, governments do not eliminate it. A possible answer is that if corruption is systemic, the likelihood of detection and punishment decreases and incentives are created for corruption to increase further. Individuals at the highest levels of government may, themselves, have no incentives to control corruption or to refrain from taking part in rent-seeking activities (discussed below). At the same time, eradication of corruption may be costlier in countries where the level of institutional development and the general economic environment is weak. Dabla-Norris and Freeman (1999) develop a model of the interconnectedness of underdevelopment and corruption in which pervasive corruption discourages market activity, thus reducing the incentives to allocate resources for eliminating it.8 The mutual importance of corruption and economic activity is illustrated by the model’s multiplicity of equilibria: one with low corruption and a high level of economic activity, and the other with widespread corruption that discourages economic activity.
The multiplicity of equilibria provides an explanation not only for the persistence of corruption, but also for the observation that the incidence of corruption varies across countries, even for similar activities. The existence of two stable equilibria in corruption suggests that, once corruption becomes ingrained, it may be very difficult to eradicate. This leads to an important policy implication: ad hoc, anticorruption campaigns will have little effect in eliminating corruption. For anticorruption policy to be effective it must be sustained. An important question is what characteristics make countries more likely to fall into a high-corruption, low-growth trap in the first place.
Corruption always results from a combination of opportunities and incentives. Opportunities for the abuse of power are prevalent in areas where restrictions and government intervention lead to the availability of rents, such as complex tax and customs systems, exchange rate controls and financial market regulations, trade restrictions, windfall gains from natural resource wealth privatization decisions, and discretionary public spending.9 Empirical evidence suggests that the rule of law, particularly effective anticorruption legislation, the availability of natural resources, the economy’s degree of competition and trade openness and the country’s industrial policy affect the breadth and scope of corruption.10
Incentives to engage in corrupt behavior are shaped, among other things, by the nature of meritocratic recruitment and promotion in civil service, public sector salaries, the quality and effectiveness of legal enforcement, the degree of transparency in government operations. Several studies at the IMF demonstrate the link between low public sector salaries and high levels of corruption in economies where monitoring is costly or ineffective and penalties for engaging in corruption are low. Ul Haque and Sahay (1996) present a model that shows there is an optimal level of wages that maximizes the government’s net revenues—that is, raising wages may more than pay for itself in terms of increased tax receipts.11 Low public sector wages, by attracting low-skilled human capital to the government sector, reduce tax collections. The implication for the IMF’s policy advice is that wage cuts can undermine public sector efficiency, and prescriptions for raising statutory tax rates, alone, may not increase revenue collection. Dabla-Norris (2000) shows that, in the absence of appropriate institutional controls, governments can economize on the costs of providing efficiency wages by allowing corruption to take place.12 Empirical evidence points to a fairly robust link between civil service pay and corruption across countries, suggesting that raising civil service pay could be a necessary, albeit not sufficient, condition for preventing corruption.13
Much of the more recent literature on this topic examines the political economy of corruption. In a study of corruption in a representative democracy, Pani (2000) shows that corruption reduces the level of taxes and public expenditures in equilibrium if citizens respond to corruption by changing their votes and policy preferences.14 Other studies argue that corruption patterns may be endogenous to political structures, with institutionalized corruption often associated with kleptocratic states.15 In this respect, recent empirical evidence on transition economies suggests that initial conditions such as the past political history of a country and its propensity to embark on structural reforms are important determinants of corruption.16
The article, “Inflation and Growth,” in the June 2000 issue of the Research Bulletin surveys evidence on the relationship between these two variables, including in transition economies.
For a more in-depth survey, which also covers research outside the IMF, see Oleh Havrylyshyn, “Recovery and Growth in Transition: A Decade of Evidence” (IMF, forthcoming).
Ernesto Hernández-Catá, “Liberalization and the Behavior of Output During the Transition from Plan to Market,” IMF Working Paper 97/53, 1997; Stanley Fischer, Ratna Sahay, and Carlos Végh, “From Transition to Market: Evidence and Growth Prospects,” IMF Working Paper 98/52, 1998a; Oleh Havrylyshyn, Ivailo Izvorksi, and Ron van Rooden, “Recovery and Growth in Transition Economies 199097: A Stylized Regression Approach,” IMF Working Paper 98/141, 1998; Andrew Berg, Eduardo Borensztein, Ratna Sahay, and Jeromin Zettelmeyer, “The Evolution of Output in Transition Economies: Explaining the Differences,” IMF Working Paper 99/73, 1999. For more survey-oriented analyses, see Oleh Havrylyshyn, Thomas Wolf, Julian Berengaut, Marta Castello-Branco, Ron van Rooden, and Valerie Mercer-Blackman, “Growth Experience in Transition Countries, 199098,” IMF Occasional Paper No. 184, 1999; Stanley Fischer and Ratna Sahay, “The Transition Economies after Ten Years,” IMF Working Paper 00/30, 2000.
Peter Christoffersen and Peter Doyle, “From Inflation to Growth: Eight Years of Transition,” IMF Working Paper 98/100, 1998; Valerie Mercer-Blackman and Anna Unigovskaya, “Compliance with IMF Program Indicators and Growth in Transition Economies,” IMF Working Paper 00/47, 2000; Oleh Havrylyshyn and Ron van Rooden, “Institutions Matter in Transition, but So Do Policies,” IMF Working Paper 00/70, 2000; George Abed and Hamed Davoodi, “Corruption, Structural Reforms, and Economic Performance in the Transition Economies,” IMF Working Paper 00/132, 2000.
Jeromin Zettelmeyer, “The Uzbek Growth Puzzle,” IMF Working Paper 98/133, 1998; see also Günther Taube and Jeromin Zettelmeyer, “Output Decline and Recovery in Uzbekistan: Past Performance and Future Prospects,” IMF Working Paper 98/132, 1998.
The basic idea that successful transition involves the reallocation of resources to a more productive new sector also underlies the theoretical analysis in Zuzana Brixiova, Wendi Li, and Tarik Yousef, “Skill Acquisition and Firm Creation in Transition Economies,” IMF Working Paper 99/130, 1999.
Robert Barro, “Economic Growth in a Cross Section of Countries,” Quarterly Journal of Economics, May 1991; Ross Levine and David Renelt, “A Sensitivity Analysis of Cross-Country Growth Regressions,” American Economic Review, September 1992.
Stanley Fischer, Ratna Sahay, and Carlos Végh, “How Far is Eastern Europe from Brussels?” IMF Working Paper 98/53, 1998b.
Mark De Broeck and Kristina Kostial, “Output Decline in Transition: the Case of Kazakhstan,” IMF Working Paper 98/45, 1998; Mark De Broeck and Vincent Koen, “The ‘Soaring Eagle’: Anatomy of the Polish Take-Off in the 1990s,” IMF Working Paper 00/06, 2000a, and “The Great Contractions in Russia, the Baltics, and the Other Countries of the former Soviet Union: A View from the Supply Side,” IMF Working Paper 00/32, 2000b.
Michael Keane and Eswar Prasad, “Inequality, Transfers and Growth: New Evidence from the Economic Transition in Poland,” IMF Working Paper 00/117, 2000.
See for instance, the essays collected in Vito Tanzi, Policies, Institutions and the Dark Side of Economics, (Northampton, Massachusetts: Edward Elgar, 2000).
Paolo Mauro, “Corruption and Growth,” Quarterly Journal of Economics, August 1995.
Vito Tanzi and Hamid Davoodi, “Corruption, Growth and Public Finances,” IMF Working Paper (forthcoming).
Vito Tanzi and Hamid Davoodi, “Corruption, Public Investment and Growth,” IMF Working Paper 97/139, 1997; Paolo Mauro, “Corruption and the Composition of Government Expenditure,” IMF Working Paper 9/98, 1998, Journal of Public Economics, August 1998; Sanjeev Gupta, Luizde Mello, and Rajii Saran, “Corruption and Military Spending,” IMF Working Paper 00/23, 2000, and forthcoming in the European Journal of Political Economy; Sanjeev Gupta, Hamid Davoodi, and Irwin Tinogson, “Corruption and the Provision of Health Care and Education Services,” IMF Working Paper 00/116, 2000.
Vito Tanzi and Hamid Davoodi, “Corruption, Public Investment and Growth,” IMF Working Paper 97/139, 1997; Sanjeev Gupta, Hamid Davoodi, and Irwin Tinogson, “Corruption and the Provision of Health Care and Education Services,” IMF Working Paper 00/116, 2000.
Sanjeev Gupta, Hamid Davoodi, and Rosa Alonso-Terme, “Does Corruption Affect Income Inequality and Poverty?” IMF Working Paper 98/76, 1998.
Jean Hindriks, Michael Keen, and Abhinay Muthoo, “Corruption, Extortion, and Evasion,” Journal of Public Economics, December 1999.
Era Dabla-Norris and Scott Freeman, “The Enforcement of Property Rights and Underdevelopment,” IMF Working Paper 99/127, 1999.
Vito Tanzi, “Corruption Around the World: Causes, Consequences, Scope and Cures,” IMF Staff Papers, December 1998; George Abed and Hamid Davoodi, “Corruption, Structural Reforms, and Economic Performance in the Transition Economies,” IMF Working Paper 00/132, 2000.
Carlos Leite and Jens Weidmann, “Does Mother Nature Corrupt?: Natural Resources, Corruption, and Economic Growth,” IMF Working Paper 99/85, 1999.
Nadeem Haque and Ratna Sahay, “Do Government Wage Cuts Close Budget Deficits? Costs of Corruption?,” IMF Staff Papers, December 1996.
Era Dabla-Norris, “A Game-Theoretic Model of Corruption in Bureaucracies,” IMF Working Paper 00/106, 2000.
Caroline Van Rijckeghem and Beatrice Weder, “Corruption and the Rate of Temptation: Do Low Wages in the Civil Service Cause Corruption?” IMF Working Paper 97/73, 1997.
Marco Pani, “Corruption and Political Equilibrium,” IMF Working Paper (forthcoming).
Joshua Charap and Christian Harm, “Institutionalized Corruption and the Kleptocratic State,” IMF Working Paper 99/91, 1999; Era Dabla-Norris, “A Game-Theoretic Model of Corruption in Bureaucracies,” IMF Working Paper 00/106, 2000.
George Abed and Hamid Davoodi, “Corruption, Structural Reforms, and Economic Performance in the Transition Economies,” IMF Working Paper 00/132, 2000; Emine Gurgen and Thomas Wolf, “Improving Governance and Fighting Corruption in the Baltics and CIS Countries: The Role of the IMF,” IMF Working Paper 00/1, 2000.