INTRODUCTION

Transition countries in the European and Eurasian region all experienced a significant decline in output throughout the 1990s.¹ While the extent of the decline is sometimes disputed on the basis of incomparability between Soviet period and market regime measures of gross domestic product (GDP), there is no doubt all countries went through what Kornai labelled a 'transitional recession'. It is also agreed that the recovery came relatively early for Central Europe and the Baltics, which hit bottom in the period 1992–1994. Others, and in particular many in the Commonwealth of Independent States (CIS), witnessed a continued decline through much of the 1990s, but starting about 2000, the latter not only finally began to recover, they experienced a surge of growth with annual GDP growth rates between 6% and 10% and sometimes more.

In the second half of the 1990s, many econometric studies of growth in transition were undertaken to explain both the sharp decline and the recovery where it had begun. Traditional factor inputs were generally found to be unimportant, and a broad consensus emerged that for better performance stabilisation was a sine qua non, liberalising reforms generally promote growth as does institutional development and that unfavourable initial conditions retard it at least in early years of transition.² These explanations fit well the broad facts of Central Europe and to a lesser extent the Baltics, experiencing less decline and earlier recovery of GDP. One implication of this literature was that CIS did not face good prospects for growth because CIS countries by the end of the decade still lagged far behind Central Europe and the Baltics on all reforms, except perhaps stabilisation. It was therefore somewhat of a puzzle why they experienced a strong surge in GDP growth from 2000, with rates far higher than Central Europe had then or at the time of their recovery. The most popular alternative explanation was the oil price boom. The first objective of this paper is to show that this alternative explanation was far from the whole story: the oil boom contributed to high growth, but the underlying reason for recovery at this time remains the same as for Central Europe and the Baltics: they reached a minimum sufficient threshold of stabilisation, liberalisation and institutions. As the last have become increasingly important in the literature, a second and more modest objective of the paper is to trace out the sequencing of reforms in transition countries, and note two important facts: in all transition countries institutions lagged far behind liberalisation, and reached very low levels prior to the rowth recovery. The paper does not propose new econometric evidence on long-term determinants of growth rates, or the much-debated issue of the relative role of institutions and liberalisation. Instead, it poses a new question: was there a minimum threshold of reforms sufficient to re-start growth?

The paper is structured as follows: The next section reviews the empirical literature on determinants of recovery in transition, including in particular macrostabilisation, liberalisation, initial conditions and institutional development. The further section considers various explanations of the post-2000 surge of growth in the CIS and notes its apparent inconsistency with the earlier literature. The subsequent section proposes a reconciliation of this inconsistency by showing that for both Central Europe and the Baltics and CIS, growth restarted once a minimum threshold of transition reforms was reached. The penultimate section then turns to the less explored dimension of institutional development, and inquires in particular about its sequencing vis-à-vis stabilisation and liberalisation. Finally, the last section summarises what is known and what remains unclear on the determinants of growth in transition economies.

A REVIEW OF THE LITERATURE ON GROWTH IN TRANSITION

Several recent surveys have reviewed the large number of econometric studies explaining growth in the transition countries: Havrylyshyn (2001), Campos and Coricelli (2002). There appears to be a good measure of consensus on the following findings: standard factor input variables are not important; prior financial stabilisation and inflation control is virtually a sine qua non; broad measures of market liberalisation and structural reforms are statistically significant; and good institutions do matter. Some controversy remains: does budget control matter directly or as a factor in controlling inflation? Does privatisation alone have a positive effect or does it also need adequate institutional change? While it is agreed that unfavourable initial conditions affect growth prospects negativelyseveral studies show that this effect declines with time; it is widely agreed that good institutions are important but how they matter, which ones are most important and to what extent they should precede or follow the other policy reforms remain in dispute. Consider briefly the five main explanations: factor inputs, stabilisation, liberalisation, initial conditions and institutions.

Factor inputs continue to show statistical significance in econometric studies even as other explanatory variables have been added by the new growth economics (Barro and Sala-i-Martin, 1995) to account for Solow's well-known finding about productivity. In transition economies, as Havrylyshyn et al. (1999) noted, the dynamics in this period are different: it is not a matter of moving the economy to a higher production-possibility-frontier (PPF) (see Figure 1) rather it means first correcting the large inefficiencies of the communist period and by moving from inside the PPF (CPU) to the PPF allocation point reflecting international comparative advantage (M1), thereby capturing substantial efficiency gains without new inputs. In the isoquant diagram, first growth results as a move from an inefficient point XI inside isoquant Q1 to a point on the isoquant (X1^*) and to the optimal allocation (BCP^*). Eventually, factor inputs and technical improvements cause movement outward to higher-level isoquants, and correspondingly an outward shift of the PPF. Even the most optimistic views of transition in 1989 would have expected this adjustment to take several years, and it would appear that in many countries it is not complete even today.³ Therefore, it is not surprising that available econometric studies that do not go beyond the late 1990s show insignificant and often negative results for the factor inputs, capital and labour.⁴ Polanec (2004) suggests further that the uniqueness of transition recovery means that reform measures swamp traditional variables such as inputs.

Figure 1.

Growth in the transition-theoretical framework

Full figure and legend (17K)

That financial stabilisation was important for growth is not a surprising result, nor indeed was it a controversial issue; even critics of the Washington Consensus agreed on the need for stabilisation. Some observers argued for the use of exchange rate anchors as the centrepiece of any stabilisation strategy, and the econometric evidence on the effectiveness of an anchor tells different stories.⁵ Some countries achieved stabilisation using currency boards like Estonia, Lithuania and Bulgaria. But a large number of Central European and later most CIS and Southeast Europe (SEE) countries achieved it without this anchor, although some had crawling/adjustable pegs, Poland, and some maintained a de facto proximity to a peg Croatia. Russia arguably had a peg until 1998 (see Owen and Robinson, 2003; Esanov et al., 2005), with limited success in stabilisation. Ukraine's stabilisation came without a clearly defined regime.

Liberalisation of markets and related structural reforms also show up as one of the main determinants of growth during the transition (see Bakanova et al., 2004; Vinhas de Souza, 2004). This is consistent with the argument that the main driver of early growth is the reallocation of resources to more efficient uses, in reaction to the new price signals and incentives of a market economy. But strong statistical significance is generally found for broader measures of market reforms, such as the European Bank for Reconstruction and Development (EBRD) transition index, less so for individual components. Thus, price liberalisation alone is significant in only a few studies; privatisation also comes out insignificant in most but occasionally significant in a few specifications. This suggests that it is the combined effect of several policies that matters in creating new opportunities for efficient resource allocation and rational decisions by the new private sector. A clear consensus exists on one point: transfer of ownership alone has at most some small positive effects, but significant benefits come only with the complementary development of competitive market institutions.⁶

Initial conditions have been measured variously as degree of over-industrialisation, share of defence industry, years under communism (a proxy for market memory or 'mental' distance from capitalism), distance from Europe, war or civil conflict, etc. Because the possible number of measures of initial conditions is so large, the results not surprisingly vary according to the choice of variable, choice of period and econometric specification. A strong role of initial conditions is found by De Melo et al. (1997). However, Havrylyshyn et al. (1999), using the same measures with additional years of data, point out that even if this was true in early years, the statistical significance of the initial conditions declines over time (Bakanova et al. (2004), find the same results). In the same spirit, Zinnes et al. (2001) distinguish immutable initial conditions such as geography and history from changeable ones such as degree of industrialisation, share of defence, etc. and also find the latter matter little after a short period of time.

It is widely agreed that institutions are important for sustained growth, although Johnson and Subramanian (2005) note that there remain some critical unanswered questions explored in the penultimate section. There are many writings on the role of institutions in non-transition economies, starting with the pioneering contributions of North (1993, 1995), and ending with the most recent revival of his ideas for developing countries. We note only a few points most pertinent to transition. Most studies do not include institutional quality as a variable, and in all cases there is no strong underlying model of how exactly institutions induce better growth performance. Moers (1999), Havrylyshyn and van Rooden (2003) show that institutional indicators are statistically significant in explaining growth, although the stabilisation and liberalisation variables remain significant as well. The latter study separates the pre-1995 and post -1995 period and finds that the coefficient of institutional variables increases in the latter period; the coefficient on initial conditions decreases. This suggests that liberalisation, stabilisation and initial conditions are the most important determinants in the early recovery, but continued and sustained growth depends less on initial conditions and much more on institutional development. Beck and Laeven (2006) show econometric results that attribute almost all the explanatory power to institutional quality alone for the entire period. This is a puzzling result for such a short period, and may reflect the underlying econometric problem that all institutional indicators are highly correlated with each other and with other reforms and performance measure.

Despite some differences, the few econometric studies relating growth and institutions in transition agree that there is a strong important link, and most emphasise that neither market liberalisation, stabilisation nor institutions alone have an overwhelming explanatory power, but rather all of them matter in a complementary fashion.

The above consensus is seriously questioned in a recent still-unpublished meta-survey of Babetskii and Campos (2007), who show that in 43 econometric studies, 'reform' coefficients are positive and significant only one-third of the time, negative-significant one-third and insignificant one-third. Their findings merit attention, but the authors do not claim that reforms have no positive effects, rather that econometric evidence for this is not strong enough. Further, they define reform only as structural measures, that is market liberalisation, and note importantly that 'controlling for institutions, initial conditions, stabilisation, is associated with a lower probability of finding positive and significant impact of reforms on growth' (p. 18). Such a conclusion does not contradict the emphasis this paper will place on three key dimensions of market-oriented reforms – stabilisation, liberalisation and institutional development. Here, I adopt a broader emerging consensus view of transition reforms, as defined by Roland (2001) as the evolutionary–institutionalist view, which recognises the importance of Washington Consensus emphasis on stabilisation and market liberalisation, but insists these must be 'grounded in adequate institutions ...to deliver successful outcomes' (p. 30). The approach in this paper is to consider 'transition reforms' to cover stabilisation, market liberalisation including private sector development, as well as institutional development. If 15 years of transition and academic debate have taught us anything, it is surely that all these are needed.

THE POST-2000 SURGE IN CIS GROWTH

The transition process that began in 1989 in Central Europe, about 1991–1992 for the Baltics but generally later for the CIS countries, has been accompanied by a transitional recession with a substantial decline in GDP, as can be seen in Figure 2. Just how huge this decline was and the portion of it attributable to continued socialist period distortions and the portion due to the inevitable adjustment costs of reforms is still very much in dispute.

Figure 2.

Index of GDP by country group 1989–2004, 1989=100

Full figure and legend (26K)

While this dispute is beyond the scope of the present paper, two brief points from Aslund (2001) and Havrylyshyn (2006) merit attention. First, since any significant reforms in the CIS countries could not even have begun before the late 1991 breakup of the USSR, the proper benchmark year for the index of GDP should be no earlier than 1992, which of course means the bottom that was reached about 1996–1997 would not be as low as 45%–60% of the transition start year but rather higher. Second, any estimated adjustments made to GDP values or index do not change the relative positions: Central Europe and the Baltics clearly saw earlier recovery than CIS and achieved a much higher level of GDP recovery by 2004.

The general pattern can be seen in Figure 2 with Central Europe and the Baltics starting to recover in the years 1992–1994, but the CIS countries – especially those with moderate reforms, described in Table 1 – continue to show the decline. Since the year 2000, these countries not only experienced a clear recovery but in fact they saw a surge of GDP growth with rates typically 6%–10% and more. As Table 1 shows, the average for this CIS moderate reformers group was well above that for Central Europe and even the Baltics.⁷ Notably, the CIS limited reformers saw a decline in growth rates after 2000, having apparently – as discussed later – performed much better than the other CIS moderate reformers in the 1990s.

Table 1 - Annual growth of GDP since 1998.

Full table

Many observers attribute this surge in growth to the luck of higher energy prices, but this explanation only goes so far, however, and in particular cannot be the whole story for the six countries in the group that import energy; only Azerbaijan, Kazakhstan, Russia and Turkmenistan are major exporters. There are several other possible explanations explored below: spill-over of Russian and Kazakh demand to their neighbours; sufficient macro-stability that triggered growth unrelated to oil; the post-1998 devaluation of the ruble and other currencies of the region; sufficient liberalisation and institutional development to provide a hard-budget environment and hence incentives for new production and investment; and finally the inevitability of hitting bottom in the transitional recession and bouncing back with high growth rates.

The sharp increase in energy prices was certainly a boon to the exporters and is surely part of the explanation for the very high growth rates.⁸ Kazakhstan and Azerbaijan have seen several years of growth of about 10% per annum or more. But, as Ahrend and Owen and Robinson (2003) show, the growth was as much attributable to an increase of production levels as to the price rise, especially for the Central Asian exporters. Furthermore, it is necessary to explain two points: why the improved terms of trade for Russia did not translate into abuse of energy revenues as is so often the case with resource booms in developing countries, and why it did not cause a resurgence of high inflation but was accompanied by an improved fiscal position and control of inflationary pressures. This is discussed later in conjunction with the arguments of reform progress, but first consider the notion of the spill-over effects.

There is no doubt that the increased domestic demand in Russia and Central Asian energy exporters spilled over to stimulate exports from their neighbours. As an example, Ukraine's food –processing industry experienced a sharp revival of exports to Russia after a decade of decline. But the spill-over argument explains only part of the growth in the other countries. First, it is certainly not enough to explain why they had growth rates as high or even higher than the energy exporters. Surely the terms of trade loss should have kept their rates lower. Furthermore, the spill-over effect was declining over time, as the diversification of trade away from intra-CIS trade continued and for all but Belarus the share of exports to Russia had declined from well over 50% in the 1990s to about a third or less by 2002 (see Elborgh-Woytek, 2003). In other words, the boom in exports was just as strong in other directions, to the European Union (EU) and Asia, which suggests that other factors such as macro-stability, devaluation and a sufficient advance in reforms were more important.

The first non-oil explanation is the achievement of macro stability, and particularly control of inflation. As demonstrated in Vinhas de Souza and Havrylyshyn (2006) for Russia, Ukraine and, to a lesser extent, Belarus, an increasingly sensible fiscal and monetary policy began to be implemented from the mid-1990s with a subsequent decline of inflation rates. This was even more true of the other CIS moderate reformer countries. By 1999, inflation in CIS countries, especially the CIS moderate reformers group, while still high, had indeed fallen to about the same level that Central Europe had at the time of their recovery in 1992–1994.

The second important non-oil explanation often adduced is devaluation, although it should properly be considered as part of any effective stabilisation. Owen and Robinson (2003) demonstrate that even for Russia oil was not the whole story. At least as important was the beneficial side effect of the 1998 financial crisis of a real exchange rate devaluation, initially about 50%. Their analysis demonstrates that there was substantial growth in non-exporting sectors of the economy, suggesting that at least some import substitution was taking place. Most of the other CIS moderate reformer currencies eventually followed the Ruble devaluation; hence, they also benefited from this effect on growth of export and import-substituting domestic production. They also argue that increased income eventually led to increased consumer demand, which in turn led to a resurgence of investment in domestic industry, further adding to the growth impetus.

Berengaut et al. (2003) analyse various possible factors behind the growth surge in Ukraine, including oil spill-over, devaluation and stabilisation, and the simple possibility that Ukraine had hit such a low point in the decline that the rebound when it came was bound to be strong. It is useful to recall the very high growth rates of 5%–10% in the mid-1990s, when war and internal conflicts subsided in countries such as Albania, Armenia, Georgia and Tajikistan. Berengaut et al. (2003) also stress the distinct hardening of the budget constraint under the more reform-minded Prime Minister Yuschenko and his Deputy for energy Yulia Tymoshenko, especially as it relates to implicit energy rents and subsidies. Owen and Robinson (2003) describe a similar hardening in Russia under President Putin, with regional budgets subordinated to the federal one, tax collections greatly increased and oil revenues prudently used to pay off substantial portions of the external debt, which declined from over 60% of GDP in 1999 to about 17% in 2005, and to build up huge foreign reserves.

In summary, oil alone is far from adequate as an explanation for the growth surge after 2000; progress on reforms played an important underlying role. However a puzzle remains: earlier econometrics showed that the more advanced the stabilisation, liberalisation and institutions, the higher the growth. In fact, in 2000, Central Europe and the Baltics were far more advanced on all three measures than the CIS moderate reformers, and yet had lower growth rates. This interpretation can be seen using the typical specification in such studies:

where GR (I, t) is GDP growth in country i, year t; INFL is the corresponding annual inflation rate; REF is a vector of reforms comprising elements such as liberalisation and institutions and IC is a vector of initial conditions. The consensus at least agreed that 'b' was negative and 'c' was positive. Taking this literally and inserting the 1999–2000 values of inflation, which was higher in CIS moderate reformers than Central Europe and the Baltics, and reforms that were lower, the equation predicts that CIS moderate reformer growth should be lower. In fact, they were much higher as can be seen in Table 1, hence the puzzle.⁹

At the same time, the popular explanation of oil has been clearly demonstrated to be incomplete, and hence it cannot resolve the puzzle by itself. Going beyond oil, other explanations summarised above are broadly consistent with the spirit of the 1990s focusing on macro and micro reforms as growth determinants. Control of inflation and the devaluations after 1998 aimed at correcting large external imbalances are both very conventional stabilisation measures. Budget hardening comprises both a macro stabilisation policy through fiscal effects and a micro signal to stimulate more efficiency. The 'sufficient' stabilisation argument noted above hints at a formal modification of the model providing reconciliation of the puzzle: to explain the timing of recovery or restart of growth, as opposed to long-term sustained growth, it may be enough to find the minimum threshold in stability, liberalisation and institutional development that induces economic agents to begin expanding production and investments. The next section elaborates.

A THRESHOLD EXPLANATION OF GROWTH RECOVERY

Given the large inefficiencies of the central plan period, it was sensible to argue that the degree of decline and the first recovery of output would depend on achieving as quickly as possible the policy changes inherent in the process of transition to a market economy. Hence, lower inflation, greater progress on measures of liberalisation, privatisation and establishment of market-enhancing institutions would provide stronger incentives for the reallocation of resources that transition was meant to bring about, and consequently stronger growth performance. In retrospect, this may have been too static an interpretation of the relation between level of market progress attained and growth. Consider the EBRD measure of market reforms, the transition progress indicator (TPI), which, in the Central Europe and the Baltic countries by 1999–2000, had values close to 4.0, just short of the maximum 4.3 of the theoretical fully functioning market economy. By this time, inflation had reached single-digit levels and continued to decline over the next few years. That is, they were close to completing the transition, they had nearly achieved a low inflation stability, and hence for continued growth – as opposed to its restart earlier – further improvements in reforms or stabilisation are likely to contribute less to growth prospects than more traditional long-term determinants including factor inputs (investment, new skills development). It is then a considerable stretch of the 1990s' logic to say that as Central Europe and the Baltics had attained lower inflation and higher levels of the TPI than the CIS moderate reformer countries, one should expect that their growth rates would be higher.¹⁰

Let me propose a different approach that is consistent with the original rationale but overcomes the issue of how to compare country policies as all approach the top of the asymptote. Suppose that for each of the major transition policy elements there exists 'a critical minimum of systemic reforms that helps to revive growth.,'¹¹ formalised simply as:

where GR and REF are as in equation 1, and MRT is the notional minimum threshold of reforms needed to stimulate growth.

In the case of stabilisation as measured by inflation control, this is not a new idea at all, and has been discussed for developing countries in detail; most analysts have concluded that growth in an adjustment programme can begin with inflation levels in the lower double digits, 25%–50% Bruno and Easterly (1995). Since there is no benchmark for the other two dimensions of reform, and arguably, the threshold inflation found relevant for developing countries may not apply to transition, I propose here to use as a benchmark the levels reached for these three measures by Central Europe and the Baltic countries just before their recovery. It is comforting that all of them first began to recover within a short time of each other, in the years 1993–1994.

Three hypotheses follow from this idea:

• H1: for each of stabilisation,liberalisation institutions, there is a minimum threshold to be attained before growth recovery.
• H2: the benchmark for the minimum threshold may be approximated by the actual levels attained by the Central Europe and the Baltic countries just before their recovery.
• H3: CIS moderate reformer countries experienced their growth recovery when they had reached the benchmark levels of the threshold.

Only H3 is directly testable, and that will be done below. A test for H2 may only be possible vis-à-vis prior experience of developing countries; for stabilisation, it is shown below that the Central Europe and the Baltics threshold is very similar to that found in the earlier literature for other countries. Testing for the other two measures is in principle possible, but would entail considerable data effort and is not attempted in this paper. As to H1, for transition countries once it is shown that H3 is upheld by the evidence, that is, the same threshold applied for both the advanced and lagging reformers, this is to some extent prima facie evidence for the existence of such a threshold.

Thus, a new question is raised about the CIS moderate reformers' surge of growth: by 1999, had they reached the same level of transition reforms that one saw in Central Europe and the Baltics at the time of their first recovery? Table 2 shows for the Central Europe and the Baltics the values before recovery for annual rate of inflation, liberalisation as measured by the EBRD first-phase reforms (TPI-LIB) and institutional progress as proxied by the EBRD's second-phase reforms (TPI-INST).¹² The same values are shown for the CIS moderate reformers in 1999–2000. For comparison, the table also shows values for the overall EBRD index (TPI-All), and all four variables for Central Europe and the Baltics in 1999–2000. The CIS limited reformer countries are not considered here for two reasons: they did not experience the same surge, and in fact may have suffered a slight decline from earlier years; their apparently superior performance even in the 1990s comprises an altogether different puzzle deserving a separate study beyond the scope of this paper. SEE is also excluded as its unstable political situation resulted in growth performance that varied considerably over time and across countries.

Table 2 - Threshold values for growth recovery: inflation, liberalisation and institutions.

Full table

Table 2 tells a very clear story. If one uses the actual reform index values attained by Central Europe and the Baltics just before they began to recover as a benchmark for the sufficient minimum threshold, then the CIS moderate reformer countries' timing of growth recovery since 2000 is entirely consistent with the broad view that stabilisation, liberalisation and institutions are critical determinants of growth. It is worth repeating that this does not imply that growth will continue and be sustained once the threshold is reached; indeed, it is more likely that it will not be sustained unless there continues to be progress on lowering inflation, reinforcing liberalisation and continuing to improve institutions. This is already suggested by the trend seen in Table 1: in Central Europe, but not Baltics, growth stalled slightly in the late 1990s, but then picked up again as they continued to progress on all three policy dimensions (Table 1). Consider each of the three measures.

Inflation prior to Central Europe and the Baltics before recovery was about 35% per annum, which, although still quite high, is very much in the middle of the range found in earlier writings for developing countries. The CIS moderate reformer group had, by the end of the 1990s, attained somewhat lower levels, about 25%. Arguably, the restart for some was as early as 1997 or the first half of 1998 but was interrupted by the Russian financial crisis. The growth rates in 1997 already had begun to turn very slightly positive in several countries, although in many they were still negative. At that time, inflation averaged 22% but shot up again in 1998–1999.

Consider next the overall TPI values. The average threshold level attained by Central Europe and the Baltics countries in the year preceding first positive growth was 2.55;¹³ by 1999, CIS moderate reformer countries had reached a similar level of 2.7. In many CIS moderate reformer countries, the 2.55 value was actually reached by about 1997 when the first signs of the turnaround began to be seen.

The debates of the early 1990s and the econometric findings of the late 1990s both emphasised the distinct but complementary effects of liberalisation compared to institutional development; it is therefore useful to look at these measures separately. The evidence remains strongly consistent with the minimum threshold hypothesis. The liberalisation index for CIS moderate reformers by 1999 is about 3.7, even higher than the Central Europe and the Baltics benchmark of 3.3. In other words, while the former were still lagging behind, they nevertheless were moving forward, some steadily, some in fits and starts; none had fully caught up to Central Europe and the Baltic countries, but all had reached a level of liberalisation at least as high as Central Europe and the Baltics experienced in the mid-1990s. A similar story applies to institutional development. The Central Europe and Baltics threshold was 1.9, while the CIS moderate reformers reached a level of 2.0 by the end of the decade just before the surge of growth.

In short, for the three main dimensions of transformation, stabilisation, liberalisation and institutional development, the CIS moderate reform countries reached levels of progress on the eve of their growth surge very similar to the levels reached by Central Europe and the Baltic countries just before their recovery. This result supports H3: for laggards in transition, growth recovery occurs when they reach the same minimum threshold of progress that the early reformers did prior to recovery. There is no real puzzle: the oil boom, broadly well managed as it was in these countries with the possible exception of Turkmenistan, doubtless added a few points to the growth rates, but macro and micro policies plus institutional development would have led to a recovery by themselves.

THE ACTUAL SEQUENCING OF INSTITUTIONS AND LIBERALISATION

The importance of institutional development and its relative sequencing vis-à-vis other policy reforms has been and will continue to be a hotly debated matter. This paper does not go into this debate in detail, but argues that a very important historical fact provides some valuable insights into the debate. Concretely, the evidence of Table 2 shows two things: institutional development lagged behind liberalisation in all transition countries, and the levels reached prior to a re-start of growth were surprisingly low in all countries. These 'facts', for the measures are disputable, are discussed in this section. First, however, a brief review of the main issues in the new institutions-growth literature is provided, with a few qualitative comments on the relevant experience of transition countries.

Key issues relating to institutions and growth¹⁴

That institutions matter is not in dispute in the literature, nor is there disagreement about the need for early stabilisation, but three big questions remain unanswered:

• Is there a minimum critical mass needed to stimulate growth?
• What are the most important market-enhancing institutions needed?
• Should they be developed before, during or after the main steps of stabilisation and liberalisation?

Consider the first issue: critical mass of institutions. Theoretically, there has been little effort to address the matter this way perhaps because the nature of institutions makes their many elements essentially non-additive, any unequal weighting is hard to justify and the high correlation among them makes it difficult to isolate the critical ones in econometric analysis. But as long as one accepts the available data as reasonably meaningful, it is very easy to look at actual historical levels and ask, as was done in the previous section for an aggregate index: was the level reached before recovery similar across countries?

The second question, which are the most important institutions for growth, requires much more detailed analysis beyond the scope of this paper. There remains considerable disagreement both on relative importance and sequencing, although a comprehensive World Bank (2002) study suggests there are some that should come early, some that can be developed simultaneously with introduction of liberalising reforms and others that can be allowed to evolve over a much longer period. The first category includes elements such as the state's ability to enforce basic law and order, new elements such as a market-oriented laws and government agencies: a Central Bank, a Finance Ministry that enforces budget discipline, a separate Treasury for transparent and uncorrupted implementation of budgets, regulatory agencies for enforcing codes of commercial behaviour, an anti-monopoly regulator. Most transition countries established such institutions formally, but the effectiveness of implementation varied. A second category relates to privatisation, freedom of private-sector activity, a legal basis for secure property rights, plus ensured competitive environment. Without property rights, entrepreneurs will be very cautious about engaging in new ventures, or expanding output and employment. In some transition countries, these were established early, with a reintroduction of pre-communist commercial codes in Poland, or borrowing Western European codes in Hungary and the Baltics. In Belarus, Russia, Ukraine and other CIS countries, new laws allowing free enterprise coexisted for some years with Soviet-period laws deeming it illegal, an inconsistency that probably discouraged early development of new small enterprises – although it was not much of an impediment to the large insiders and future oligarchs.

The third issue, sequencing relative to other policy reforms, has been a very intensely debated one in the transition discussion of the early 1990s and continues to be a favourite topic for those who criticise the Washington Consensus failures.¹⁵ The critics most often point to the rush involved in liberalisation and privatisation and attribute the poor results to the lack of effort in building up institutions in advance. This has often been a heated debate with both sides using the same evidence to draw opposite conclusions. For example, the pioneering work of North on institutions is adduced by gradualist-institutionalist proponents as historical evidence of the need for good institutions; big-bang proponents point to North's caution that in market economies institutions evolved over very long periods of time, and so delaying liberalisation is impractical and risky. Here, the issue is indirectly addressed by simply asking what the observable path and sequencing for institutions-liberalisation was in the transition period.

The path and sequencing of institutional development since 1989

Many quantitative measures of institutional development are available providing very narrow detail on different types of institutions and frequently averaging these into an overall index. They are largely measures of perceptions by local business and/or academic experts, domestic and foreign, of the effectiveness of institutions; hence, they can be criticised for the inaccuracies of subjectivity. However, this very subjectivity may give a better measure of the strength of these institutions than harder measures of the existence or non-existence of formal laws and regulations.¹⁶ In developing and transition countries, looking only at the laws on paper often tells little about implementation and effectiveness. It was a widely held view of kremlinologists that in Soviet states, informal rules were far more important than the paper laws; for example, private economic activity was largely banned, but as Handelman (1994) elucidates, the existence of underground economic activity and a Soviet-mafia speaks volumes about what the real law was. Until recently, economists have paid limited attention to the informal institutions that political scientists have long studied, but in the transition debates it has come to be recognised that the key issue is effectiveness of implementation.

Recognising this, indices of institutional development intentionally rely on subjective perceptions. Early compilations of such synthetic indicators of market institutions were performed by Freedom House, the Heritage Foundation and Transparency International with its well-known Corruption Index. More recently, a comprehensive compilation of various sources and many new indicators from now-annual World Bank surveys has been carried out and analysed in Kauffman et al. (2004). The accuracy of such measures can be questioned; however, any analysis of institutional development wishing to have some quantitative purchase on the issue cannot but rely on these data sources.

The major criticism that measures of institutions may not capture the quality or effectiveness of implementation is addressed in great detail in the above work of the World Bank. Their approach attempts to reflect the related problem of informal institutions in which vested interests quickly develop to subvert the original policy intentions.¹⁷ Furthermore, subjective perceptions also address, at least in principle, the issue of enforcement. Indicators that use subjective assessment by experts or panels of experts, or opinion surveys of practitioners, 'how easy is it to do business' and 'how much is bribery a problem,' suffer from the inaccuracy of subjective measures, but paradoxically may get at the important issue of implementation and effectiveness better than hard quantifiable measures.

Weder (2001) used such data to assess the degree of institutional development in the region, and Table 3 summarises the results for 1997–1998. For a comparison of rank order among transition countries, Table 3 also shows 2003 values for a World Bank measure; as the scaling is not the same, comparison over time is not possible in Table 3.

Table 3 - Quality of institutions in transition economies 1997–1998 and 2003.

Full table

Three broad conclusions follow. First, in the late 1990s, transition economies were still far from the level of institutional quality prevailing in the advanced industrial countries, although Central Europe and the Baltics reached levels similar to the leading emerging market countries. Second, the rank ordering of institutional development by country group is the same as that for progress towards the market as measured by the EBRD – this is even more clear in Table 4. Third, the rank ordering has remained the same 5 years later.

Table 4 - EBRD Transition Progress Index, (TPI) Selected Years Liberalisation (LIB) versus Institutional Development (INST).

Full table

The best available facts on growth and institutions in transition economies show that despite the very low levels of institutional development in some (Table 3), all countries have seen significant economic recovery by now (Table 1). This suggests the following hypothesis on the important debate between big-bang and gradualist schools of thought: the start of a growth recovery is possible even if the process of developing market-enhancing institutions lags behind that of stabilisation and liberalisation.

The earlier discussion on sequencing among types of institutions noted that there may not be clear examples of institutions that are postponable until after the basic reform steps are carried out. But a more useful view of sequencing may be that transition should begin with a basic and simplified legal-regulatory framework that is comprehensive but not deep, and follow with refinements of these regulations over time. As North (2006) re-emphasised, it is always a case-by-case judgement call what the basic minimum may be, and how soon further refinements take place. In retrospect, it is clear that most of these economies proceeded along such a pragmatic path, albeit at different speeds. The Central Europe group moved the fastest; the countries of the CIS, with some exceptions such as Armenia's land laws, moved much more slowly, intermittently introducing new laws and abolishing Soviet ones.

To get some view of the actual sequencing over time between institutions and liberalisation reforms, I use the simpler EBRD indicators, which go back to the early 1990s and have the advantage of cardinal comparability, explained below, and same-source consistency for both indices. The 10 dimensions can be grouped into two parts: those indicators that measure market liberalisation, EBRD's initial phase reforms and those that involve institutional changes, their second-phase reforms. Table 4 shows values for liberalisation (LIB) and Institutions (INST) for 3 years, 1994, 2000 and 2005. The first includes price liberalisation, foreign trade liberalisation and small-scale privatisation,and the second, governance and enterprise restructuring, large-scale privatisation, competition policy, banking and financial sector reform, which together can be thought of as institutional reforms.¹⁸ In its own analysis, the EBRD (2003) more or less equates second-phase reforms with institutional development; for our purposes, it is comforting that the correlation coefficient between the 1990 values of INST and those of Weder in Table 3 is very high (0.94192) for 25 countries.

Two facts seem clear from the data of Table 4, and each of them leads to some tentative interpretations. First, throughout the region, no matter how quickly or slowly liberalisation proceeded, institutions lagged behind. By 2005, even the most advanced reformers of the Central Europe and the Baltics group, which had reached LIB values of 4.3, that is, essentially completed the process of achieving full liberalisation equivalent to existing market economies, still had INST values about 3.2–3.3, well below maximum levels. The other countries were even farther behind, with the rank ordering of institutions exactly the same as the rank ordering for liberalisation. It has already been noted that the observed actual sequencing of growth and institutional development implies that growth can re-start with institutional development lagging behind liberalisation. The sequencing between liberalisation and institutions shown in Table 4 confirms such an interpretation. This relative sequencing also suggests that where there was the will and capacity to move fast on liberalisation, there was an equal will and ability to move forward on institutional development, and vice versa.¹⁹

The second important fact is that not a single case exists of a country following the theoretical path of an institutionalist-gradualist strategy, with institutions preceding or at least moving in parallel with market liberalisation.²⁰ This is true not only for the gradual reformers of the CIS moderate reform group but also and importantly for the three countries in the CIS limited reformers group. In all three cases, the political leadership often stated that they were pursing a more gradual path to reforms than the shock therapy that caused so much pain in their neighbours, and many outside observers point to some of them like Belarus as examples of avoiding the sharp decline in growth.Whether in fact they managed to avoid the transition recession and the social costs of reform or only delay it is beyond the scope of the paper, but one thing is clear: none of them used the opportunity of gradual liberalisation to move ahead with institutional development. Indeed, they moved even more slowly on institutions than the CIS moderate reform countries, not to speak of Central Europe and the Baltics.²¹ The values in Table 4 clearly show that Belarus lagged well behind Russian and Ukraine not only on liberalisation, but even more so on institutional development.

In all regions institutional reforms lagged behind liberalisation and in general with less of a surge than seen for liberalisation in the early 1990s. This pattern is consistent with the view that institutional reforms cannot be implemented as quickly as many components of liberalisation, but it may also be consistent with the criticism aimed at the Washington Consensus, that they were not given enough importance early on. This debate cannot be easily resolved, as there is no obvious benchmark to define the apt phrase of Vaclav Klaus 'reform as fast as possible'. The evidence is therefore easily used by both gradualists and big-bang proponents in support of their arguments. However, a great deal can still be inferred by comparing the time patterns across country groups.

By 1994, the Central European and Baltic countries had reached the very high LIB level of 3.7, broadly comparable to many mixed economies with some state ownership and price regulation of a few basic goods. By 1999, this had increased substantially to nearly 4.0 for most, and by 2005 had reached the maximum 4.3 rating of the EBRD index. It is particularly notable that the Baltics, starting later than the others, already caught up to them by 1994 and kept pace in the final liberalisation drive.

Southeast Europe lagged somewhat behind, although not nearly as much as the CIS group, which, in 1994, was far below the level of SEE.²² Indeed, at the mid-1990s point, the gap between the CIS moderate reformers and CIS limited reformers is not visible. This quickly changes however, and by 1999 the CIS moderate reformers progress is evident, as it surges ahead in liberalising measures to achieve about the same position that Central Europe and the Baltic countries had 5 years earlier. I argued in the previous section that it is not just a coincidence that the CIS moderate reformers output recovery began after reaching about the same level of reforms as had been reached by Central Europe and the Baltic countries when their recovery began. What is striking is that by 2005, although the CIS moderate reformer countries had not yet caught up to the liberalisation levels of Central Europe, they were very close to the 4.0 mark, reflecting reasonably well-functioning market mechanisms, if not yet institutions. The nearly stagnant process of liberalisation for CIS limited reformer countries and the even greater lag for institutions is particularly evident in Table 4.

Finally, what do the values of Table 4 imply for the debates about the relative importance and sequencing of stabilisation, liberalisation and institutions? First, note that the degree of institutional reform reached before growth started was not that high in Central Europe and the Baltics. Indeed, even by 2003, the level of development of institutions still had a long way to go in the advanced countries, and yet no major damage to the performance of the economies appears to have occurred. Johnson and Subramanian (2005) propose that, generally, good economic policy alone can give growth a start, but sustained growth requires improved institutions; the case of Central Europe appears to fit this hypothesis particularly well. In the first recovery phase, 1993–1998, GDP grew after the surge of liberalisation but then slowed as institutional reform lagged. As it accelerated, doubtless abetted by the EU Accession process, stronger growth returned after 1999. Second, there is not a single instance of slow liberalisers moving ahead more rapidly or even at the same pace with institutional reforms as recommended by the institutionalist-gradualist. On the contrary, those that liberalised fastest also moved fastest on institutional reforms, albeit with a lag.

The EBRD makes the point that partial liberalisation creates 'winners...who block further progress in reforms' (EBRD, Transition Report 2000, p. 30). Perhaps the most important inference is that the slow reformers were not slow reformers for the public reasons often given by politicians urging gradual liberalisation: 'the economy is not ready for market operations', and 'we must avoid the huge pain of shock therapy'. If that had been the sincere motivation for gradual reform, one should have seen after 15 years some instances of institutional reform moving faster or at least at the same pace as liberalisation. To repeat, this pattern is not seen in any country; where liberalisation has been very slow (CIS limited reformers), institutional development has been even slower.

All this leads to a tentative conclusion as to why, in practice, gradualism could not work as intended theoretically. Gradualist arguments were motivated by the legitimate concern that liberalising too fast ahead of institutional developments would be less effective, and would create more dislocation and pain. In theory, and using mathematical models, this can be and has been shown to be a more rigourous argument than the one for a big-bang. In practice, however, gradualism has been less effective than rapid reforms at preventing the flourishing of rent-seeking activities and subsequent evolution of oligarch regimes, because the rationale of gradualism was abused by political leaders who had no commitment either to liberalisation or to institutional development.

CONCLUSIONS

This paper was motivated by the apparent puzzle that the nine moderately reforming CIS countries experienced a surge of growth from 2000 despite reforms still lagging considerably behind Central Europe and the Baltics. This seemed inconsistent with earlier studies that generally pointed to reforms in three areas – stabilisation, liberalisation and institutions – as the most important determinants of growth. The conventional econometric specification implied that the greater the progress in these three, the higher the growth rate – thus even after 2000, Central Europe and the Baltics should have had higher, not lower, growth. But the common alternative explanation of an oil price boom, though it had a clearly positive effect, at most explains the very high growth rates attained in energy exporters but not the other CIS moderate reform countries.

To reconcile this puzzle, this paper argues that the conventional econometric specification with growth rate as the dependent variable and level of liberalisation and institutional reform as independent variables was not mathematically meaningful by the year 2000 because these indices increase asymptotically and in the advanced reformers had reached or were approaching the maximum. The question was reformulated to ask: is there a minimum threshold level of these reforms sufficient to re-start growth? The paper tries to explain only the timing of growth recovery, and not its relative levels or sustainability. Using the actual experience of Central Europe and the Baltic countries in the period 1992–1994 as the benchmark, a minimum recovery threshold was calculated for the three reform dimensions. This paper then showed that the CIS moderate reform countries generally began their recoveries when they reached approximately the same threshold levels for each dimension.

Because the apparent level of institutional development before recovery in all countries appeared to be much lower than that for liberalisation, this paper went on to investigate more closely the path and sequencing of these two aspects of transformation. Two main conclusions are noted. First, in all countries, institutions lag far behind development, and second the catch-up of institutional development is faster in the rapid reformers with early and significant liberalisation than in the gradual reformers with delayed and/or slow liberalisation. An interesting policy story follows with indirect implications for the big-bang versus gradualism debates. In countries where leaders had a commitment to liberalisation, there was an equally strong commitment to institutional development. In countries where the political rhetoric propagated a gradual policy to achieve better future performance, the theoretically expected precedence of institutions ahead of liberalisation was not observed in a single case; this strongly suggests that the rhetoric was just that, rhetoric.

One implication for future research is that testing the sequencing hypothesis – institutions preceding liberalisation are superior – may be virtually impossible, as no empirical case study exists where this sequence was followed. Nevertheless, several directions of further research emerge from these findings. First, recognising that the three countries where reforms have been very limited appear to have better growth performance, research is still needed to determine if these rates overestimated as were rates in later years in the USSR, or if not how to explain the superior performance, given the clearly inferior achievements of these countries on all three policy dimensions? Second, the threshold analysis might be performed with more rigourous econometric analysis, and measuring more precisely the change of the transition index that, by definition, follows an asymptotic path. Third, given the surprisingly low level of the institutions threshold one should ask whether some particular subcomponents of such indices are more important and need to attain much higher levels for recovery to occur. Fourth, while it may be slightly early in transition countries to distinguish the sufficient conditions for recovery as done here from those necessary for sustainability, future work on this would be important. It is possible that some lessons could already be drawn from the Central Europe and the Baltics group where the mid-1990s recovery was followed by weakening in some, and then a new vigour after 2000. Finally, the record of developing countries that suffered a financial crisis and output slowdown or decline provides fertile ground for testing further the minimum threshold hypothesis.

Notes

¹ This paper was originally prepared for presentation at the Seventh Annual Global Development Conference, St. Petersburg, Russia, 19 January 2006. I am grateful to Lucio Vinhas de Souza, Marek Rohozynski and Waldemar Skrobacki, the participants of a seminar presentation at the Munk Centre, October 2006., and two anonymous referees for suggestions. Nikola Milicic provided valuable research assistance.

² Two surveys of this literature provide a thorough comparative analysis of the approaches and findings: Campos and Coricelli (2002) and Havrylyshyn (2001).

³ Gros and Steinherr (2004) use several quantitative measures to show that after 10 years, the transition may have been almost over for some of the Central Europe and the Baltics countries, but was far from over in South-east Europe and the CIS; pp. 116–127.

⁴ Growth studies estimating TFP have not been common, but those carried out generally find that it is high, confirming the notion that efficiency gains are the most important part of early recovery (DeBroeck and Koen, 2000).

⁵ Williamson (2005) discusses whether anchors should be seen as part of the Washington Consensus or not.

⁶ While many surveys of privatisation effects show this, it is perhaps most thoroughly explored in the econometrics of Zinnes et al. (2001).

⁷ The groupings shown reflect the degree of progress in transition as measured by the EBRD, following the analysis in Havrylyshyn (2006), which shows that the EBRD measure is highly correlated with many other possible measures of transition progress. The correspondence with geography is apparent, but some marginal cases exist; thus, Slovenia and Croatia may be part of SEE, but their income and advanced transition progress are much closer to Central Europe.

⁸ See Ahrend, Chapter 5 in Vinhas de Souza and Havrylyshyn (2006).

⁹ However, growth in the Baltics is higher and hence consistent with the model. Part of the explanation there may be that they have avoided the deterioration of fiscal balances and possible crowding-out effects seen in Central Europe.

¹⁰ The dilemma is approached in a different way by Rzonca and Cizkowski (2003), who argue convincingly that since the TPI increases asymptotically to the 4.3 maximum, comparing absolute levels or even changes in absolutes for countries at different stages of transition is not correct.

¹¹ Mohacs-Nagy (2000, p. 30).

¹² The reason for using this proxy rather than much more detailed available indices is given in the penultimate section.

¹³ The variation for individual countries was not substantial.

¹⁴ This section is based largely on Johnson and Subramanian (2005), although of course the author is fully responsible for the interpretation.

¹⁵ A very balanced assessment, although it takes the gradualist and institutionalist position, is in Roland (2001).

¹⁶ North (2006), in a succinct summary of his well-known work on institutions, emphasises that one must understand three levels of institutions: the formal ones, their informal counterparts or complements and the enforcement mechanisms. The vast surveys from which data now come have usually been carefully designed to reflect all this in the summary of perceptions by affected agents and expert observers.

¹⁷ Alina-Pisano (2006) gives many examples of such 'institutional facades' in the post-communist period.

¹⁸ Their index of infrastructure reform is arguably not relevant to either and I have excluded it. It typically has values lower than liberalisation and about the same as institutional development.

¹⁹ One can think of this in terms of the recently popular notion of policy ownership or commitment: where the overall commitment to a liberal market and liberal democracy was strong, institutional development proceeded as fast as possible, although this often meant not as fast as liberalisation simply because the latter could be faster.

²⁰ The table's group averages are calculated from individual country values as in EBRD, but are not shown here to save space.

²¹ Admittedly, the INST measure here can be criticised for giving most weight to market institutions. But many studies exist showing that these three countries lag very far behind in measures of democracy and civil society.

²² No 1994 data available for conflicted Bosnia-Herzegovina and Serbia-Montenegro.

References

1. Aslund, A 2001: The myth of output collapse after communism. Carnegie Endowment for International Peace Working Paper 18, Washington DC.
2. Alina-Pisano, J. 2006: Klychkov i Pustota: Post-Soviet bureaucrats and the production of institutional facades. Paper presented to Second Annual Danyliw Research Seminar, Chair of Ukrainian Studies, University of Ottawa, October 12–14, 2006.
3. Babetskii, I and Campos, NF. 2007: Does reform work? an econometric examination of the reform-growth puzzle. IZA Discussion paper no. 2638, Bonn, February 2007.
4. Bakanova, M, Vinhas de Souza, LV, Kolesnikova, I and Abramov, I. 2004: Transition and growth in Belarus. In: Ofer, G and Pomfret, R (eds). The Economic Prospects of the CIS: Sources of Long-term Growth. Edward Elgar Publishing: UK, pp. 57–75.
5. Barro, R and Sala-i-Martin, X. 1995: Economic growth. McGraw-Hill: New York.
6. Beck, T and Laeven, L. 2006: Institution building and growth in transition economies. Journal of Economic Growth 11: 157–186. | Article |
7. Berengaut, J, de Vrijer, JEJ, Elborgh-Woytek, K, Lewis, M and Lissovolik, B. 2003: An interim assessment of output developments in Ukraine, 2000–2001. IMF Working paper 02/97, The International Monetary Fund: Washington.
8. Bruno, M and Easterly, W. 1995: Could inflation be expansionary? Transition Newsletter, July–August 1995. World Bank: Washington DC.
9. Campos, N and Coricelli, F. 2002: Growth in transition: What we know, what we don't and what we should. Journal of Economic Literature XL: 793–836. | Article |
10. DeBroeck, M and Koen, V. 2000: The great contractions in Russia, the Baltics and the other countries of the former Soviet Union. IMF Working paper 00/32, International Monetary Fund: Washington.
11. De Melo, M, Denizer, C and Gelb, A. 1997: From plan to market: Patterns of transition. In: Bléjer, M and Skreb, M (eds). Macroeconomic Stabilization in Transition Economies, Chapter 1. Cambridge University Press: Cambridge, MA.
12. EBRD. various years: The Transition Report. The European Bank for Reconstruction and Development: London.
13. Elborgh-Woytek, K. 2003: Of openness and distance: Trade development in the commonwealth of independent states. International Monetary Fund: Washington.
14. Esanov, A, Merkl, C and Vinhas de Souza, L. 2005: Monetary policy rules for Russia. Journal of Comparative Economics 33(3): 484–499. | Article |
15. Gros, D and Steinherr, A. 2004: Economic transition in central and eastern Europe: Planting the seeds. Cambridge University Press: Cambridge, UK.
16. Handelman, S. 1994: Comrade criminal. Penguin Group: London.
17. Havrylyshyn, O. 2006: Divergent paths in post-communist transformation: Capitalism for all or capitalism for the few?. Palgrave MacMillan: Houndsmills, UK.
18. Havrylyshyn, O and van Rooden, R. 2001: Recovery and growth in transition: A decade of evidence. IMF Staff papers, Vol. 48, Special Issue: "Transition Economies: How Much Progress?".
19. Havrylyshyn, O and van Rooden, R. 2003: Institutions matter in transition, but so do policies. Comparative Economic Studies 45(1): 2–24. | Article |
20. Havrylyshyn, O, Wolf, TA, Berengaut, J, de Castello Branco, M, van Rooden, R and Mercer-Blackman, V. 1999: Growth experience in transition countries, 1990–1998. IMF Occasional paper no. 184, International Monetary Fund: Washington.
21. Johnson, S and Subramanian, A. 2005: Aid, governance, and the political economy of growth and institutions. Paper presented at Seminar on Foreign aid and Macroeconomic Management, Maputo, 14–15 March 2005.
22. Kauffman, D, Kraay, A and Mastruzzi, M. 2004: Governance matters III: Governance indicators for 1996–2002. World Bank paper. The World Bank: Washington.
23. Moers, L. 1999: How important are institutions for growth in transition countries? Tinbergen Institute Discussion paper 99–004/2 (Amsterdam).
24. Mohacs-Nagy, P. 2000: The Meltdown of the Russian State: The Deformation and Collapse of the State in Russia. Edward Elgar: Cheltenham, UK.
25. North, DC. 1993: Institutions and credible commitment. Journal of Institutional Change and Theoretical Economic 149(1): 11–23.
26. North, DC. 1995: Institutions, institutional change and economic performance. Cambridge University Press: New York.
27. North, DC. 2006: Understanding economic change and economic growth. Kozminski Academy and TIGER, Distinguished lecture series no. 7, Warsaw, Poland. ( www.tiger.edu.pl ).
28. Owen, D and Robinson, D. 2003: Russia rebounds. International Monetary Fund: Washington.
29. Polanec, S. 2004: Convergence at last? Eastern European Economics 42(July–August): 55–80.
30. Roland, G. 2001: Ten years after...transition and economics. The IMF Staff papers, Vol. 48, Special Issue "Transition Economies: How Much Progress?", pp. 29–52.
31. Rzonca, A and Cizkowicz, P. 2003: A comment on the relationship between policies and growth in transition countries. Economics of Transition 11: 743–748. | Article |
32. Vinhas de Souza, L. 2004: Financial liberalization and business cycles: The experience of the new EU member states in the Baltics and Central Eastern Europe, Deutsche Bundesbank. Working paper series no. 22/04, Frankfurt A.M., Germany.
33. Vinhas de Souza, L and Havrylyshyn, O (eds). 2006: Return to growth in cis countries and the macroeconomic framework. Springer Publishing: Berlin.
34. Weder, B. 2001: Institutional reform in transition economies. How far have they come? Working paper 01/114, International Monetary Fund: Washington.
35. Williamson, J. 2005: Differing Interpretations of the Washington Consensus. TIGER Distinguished Lecture Series No.17. (accessible at www.tiger.edu.pl).
36. World Bank. 2002: The World Development Report. The World Bank: Washington DC.
37. Zinnes, C, Eilat, Y and Sachs, J. 2001: The gains from privatization in transition economies: Is change of ownership enough? IMF Staff papers, Vol. 48 (Special Issue), pp. 146–170.