Cross-Country Differences in Convergence in CESEE

Transcription

1 By Alan Stark - "Mini Stack" Interchange of Interstate 1, Loop, and State Route 51 at Night (), CC BY-SA., Bas B. Bakker and Krzysztof Krogulski June 17 Cross-Country Differences in Convergence FREE NETWORK ICY PAPER SERIES

2 Introduction 1 Since 1989, there have been large differences in the convergence of the income levels of the former communist countries with those in the US. Most Central European countries have seen a sharp rise in relative incomes, but many countries in former Yugoslavia and the CIS have not indeed, some countries, including Moldova and Serbia, are now poorer than they were in 1989 (Figure 1). The difference between Ukraine and Poland is particularly stark. In 1989, both had similar income levels, but Poland is now more than three times as rich (Figure ). Figure 1. Transition outcomes Change in income per capita gap to USA, (percentage points, PPP-adjusted) below above Note: for Bosnia and Herzegovina 1 Bas B. Bakker is the Senior Resident Representative and Krzysztof Krogulski an economist in the IMF s Regional Office for Central and Eastern Europe in Warsaw. The views expressed in this paper are those of the author(s) and do not necessarily represent the views of the IMF, its Executive Board, or IMF management.

3 Figure. GDP level in Poland and Ukraine GDP per capita (thousands of 1 US dollars, PPP-adjusted) 3 5 Poland 15 1 Ukraine As a result, cross-country income differences remain large. In 1989, the Czech Republic, Russia, Slovenia and Croatia had the highest income per capita in 1989, about times as high as in Albania and Moldova, the poorest in the group. Twenty-six years later, the differences are even larger. GDP per capita in Slovenia is 6 times as high as in Moldova (Figure 3). Figure 3. Cross-country income differences GDP per capita (percent of US GDP per capita, PPP-adjusted) below above 5 3

4 What explains these differences? These differences in convergence do not seem to reflect data problems. True, GDP statistics in 1989 were not very good. It is hard to measure value added when prices are not quite right. Moreover, GDP at that time was probably not a good indicator of consumer welfare. Much of what was produced was not wanted by consumers (e.g. military expenditures) and/or of low quality. Nevertheless, these issues apply to all post-communist countries in the regions it is not clear that some countries suffered from data problems more than others. Indeed, more direct measures of economic activity also suggest large initial output falls and large crosscountry differences. Between 199 and 1995 electricity consumption per capita fell by almost percent in Ukraine and Moldova. By then electricity consumption in Poland had nearly recovered to its 199 level (Figure ). Figure. Alternative measure of decline in economic activity Electricity consumption per capita (index, 199= 1) RUS Instead, several factors seem to have a played a role: The speed of transition to a market economy War and conflicts Boom-busts EU Membership Whether transition has been completed Source: IFA Statistics and IMF staff calculations. Countries that reformed early had a shorter and shallower post-transition recession. The lower the EBRD transition index in 1995 (i.e., the less the economy was reformed), the sharper the output decline between the beginning of transition and 1995 (Figure 5).

5 Figure 5. Market reforms and post-transition recession Cumulated change of GDP and advancement of reforms in the beginning of transition Cumulated change of GDP in the beginning of transition, t;t+ 5 (percent) t= 1991 for former USSR countries t= 199 for Bosnia and Herzegovina t= 1989 for other countries BIH SRB & MNE BLR SVN MKD ROU ALB RUS HRV LTU CZE EST Averag e of six EBRD transition indicators in 1995 Why was this? In late 1989, a fierce debate broke out over what came to be called gradualism versus shock therapy. Many gradualists argued that the structural flaws of the economy would frustrate attempts at liberalization, and therefore that reforms should be implemented in a gradual, sequenced way. But for others including key figures such as Leszek Balcerowicz in Poland understanding the nature of the problem meant the opposite: reform was a seamless web that could only succeed if all the changes happened together, because liberal prices, improved governance, and a stable economic and financial environment were needed to reinforce one another; little could be achieved with a partial reform. The evidence from the past 5 years has vindicated the seamless web theory of transition. There is no doubt that some reforms took much longer than anticipated, including privatization, both of banks and companies. But it seems clear that the countries that made sweeping changes, and that kept at reform and stabilization have done well. Countries that followed a more gradual path suffered from the decline of the old industries, and did not get the boost from the growth of new firms. Moreover, in some countries bouts of macroeconomic instability repeatedly undermined reforms and sapped political momentum. Weaker growth in the early transition years was not compensated by faster growth later. Countries where output declines were deeper in the early 199s did not see more rapid growth in subsequent years (Figure 6). This is not to say that the rapid and seamless approach was without problems, notably large losses of output and high unemployment in the short run. Thus, reform will always have to worry about the social safety net and, under some circumstances, may benefit from external assistance, which is where the IMF and others can come in. 5

6 Figure 6. Permanent output loses in early transition Convergence per capita to USA in and (percentage points) 5 Note: for Bosnia and Herzeg ovina Change in income per capita gap to USA, (pp) RUS SRB & MNE LTU EST BLR ROU CZE SVN MKD ALB BIH HRV Change in income per capita gap to USA, (pp) Wars and conflicts also played an important role. It is striking that the five countries with the lowest growth all had a war or serious conflict between 199 and 15 (Figure 7). Figure 7. Wars and conflicts impact on long-term growth Average GDP growth, , and its level in 1989 Red point indicates whether a country experienced a war in BIH Averag e annual GDP per capita g rowth, (percent) ALB BLR EST ROU MKD CZE LTU SVN RUS HRV SRB & MNE - Note: for Bosnia and Herzeg ovina GDP per capita in 1989, thousands of 1 US dollars, PPP-adjusted Avoiding boom-busts helped boost longer-term growth. Steady growth rates seem to be more conducive to higher longer-term growth than booms followed by busts. Between and 8, Romania had a capital inflows-fueled boom and grew much faster than Poland, but thereafter it suffered a deep bust, and between and 15, Poland has grown faster (Figure 8). 6

7 Figure 8. The hare and the tortoise 1 GDP growth, -1 GDP per capita, index = Poland Poland 1 Romania 13 - Romania EU accession was a powerful catalyst for reforms and upgrading of institutional frameworks. CESEE countries that joined the EU were required to bring their regulations and institutions up to Western European standards. There is a striking difference in the level of the EBRD transition indicators between the EU countries and non-eu countries (Figure 9). Thus, prospects of EU Membership have led to more reforms and, as a consequence, to stronger growth (Figure 1). Figure 9. EU accession as reform catalyst Average of six EBRD indicators in 1. non-eu countries EU countries 3.5 Note:7 for Czech Republic BLR BIH SRB RUS ALB MKD SVN ROU HRV CZE LTU EST Source: EBRD and IMF staff calculations. 7

8 Figure 1. Market reforms and changes in income levels GDP per capita and average of EBRD transition indicators Beg in of the arrow indicates data for 1995, the end of the arrow - for 1 3 GDP per capita in 1995 and 1, thousands of 1 US dollars, PPP-adjusted 1 SEE non-eu SEE EU CIS Baltics CE Average of the EBRD transition indicators in 1995 and 1 Source: EBRD, Total Economy Database and IMF staff calculations. Countries that upgraded their institutions to EU standards saw a decline in cross-country income differences. Countries that joined the EU in the s show a clear pattern of convergence. The difference between Bulgaria and Slovenia has narrowed by 15 percent of Slovenia s GDP since the former begun EU accession negotiations in (Figure 11, right panel). Similarly, a group of candidate and potential candidate countries, including Croatia (which joined the EU only in 13), have converged as well (Figure 11, left panel). Figure 11. Convergence with regions CESEE-EU countries (as of 7) 6 EU candidate countries and Croatia 6 Averag e annual GDP per capita g rowth, -15 (percent) ROU LTU EST CZE SVN y = x R² = GDP per capita in, thousands of 1 US dollars, PPP-adjusted Averag e annual GDP per capita g rowth, -15 (percent) ALB BIH SRB & MNE MKD y = -.36x R² =.79 HRV GDP per capita in, thousands of 1 US dollars, PPP-adjusted Note: The EU has recognized Bosnia and Herzegovina as potential EU candidate countries. 8

9 By contrast, there was no convergence among the European CIS-countries. Russia, the richest of CIS, countries grew by only.6 percent annually since 1989, while output per capita declined in Moldova and Ukraine. Only Belarus achieved growth rates comparable to the non-cis countries, but its largely unreformed economy may have approached the limits of the current extensive growth model (Figure 1). Figure 1. Convergence in European CIS region Average GDP per capita growth, , and its level in BLR Averag e annual GDP per capita g rowth, (percent) RUS GDP per capita in 1989, thousands of 1 US dollars, PPP-adjusted Source:Total Economy Database and IMF staff calculations. Countries that have a more completed transition are richer. There is a strong correlation between progress in market reforms and a country s income level (Figure 13). Similarly, richer countries have a more vibrant private sector (Figure 1). Correlation does of course not mean causality, but is it telling that there is no highly reformed poor country. Figure 13. Market reforms and income level GDP per capita as percent of US and average of EBRD transition indicators, 1 GDP per capita as percent of US, BIH RUS SRB & MNE SVN ROU MKD ALB CZE EST LTU HRV Averag e of six EBRD transition indicators, 1 Note:EBRD data for CZE from 7 and for SRB & MNE only for Serbia Source: EBRD, Total Economy Database and IMF staff calculations. 9

10 Figure 1. Market reforms and private sector share in the economy GDP per capita in 15 and private sector share in GDP in 1 GDP per capita as percent of US, SRB BIH RUS MNE SVN HRV ROU MKD LTU ALB EST Source: EBRD, Total Economy Database and IMF staff calculations. Convergence post-9 crisis Post-9, catch-up has slowed down. Pre-crisis convergence was rapid and widespread. In some countries, the GDP per capita gap to the US narrowed by more than 1 percentage points in 3-8. Since 1 only two thirds of countries in the region have continued to catch-up with the US, while Ukraine and Slovenia saw a widening of income differences (Figure 15). Moreover, if we include the 9 crisis, which was deeper than in Western Europe, convergence has been even less. Figure 15. Convergence pace pre- and post-crisis Change in income per capita gap to the US (percentag e points, PPP-adjusted) Private sector share in GDP in 1 (percent) below - pp above 1 pp. Source: WEO database and IMF staff calculations. Source: WEO database and IMF staff calculations. 1

11 More recently, there have also been large differences across regions: while the CIS was in recession, the non-cis countries doing much better. The CIS countries suffered from falling commodity prices, and from the impact of sanctions on Russia. By contrast, the non-cis countries saw a gradual acceleration of GDP growth, on the back of a pick-up of domestic demand in the euro area. Labor markets in many EU New Member States (NMS) are tightening rapidly, and unemployment is quickly approaching pre-crisis lows, though GDP growth rates are well below those in the pre-crisis years. How can we boost convergence going forward? 3 GDP per capita is the product of GDP per worker (labor productivity) and the share of the population that works (the employment rate): GDP POP = GDP E E POP Low GDP per capita can thus be the result of both low labor productivity and a low employment rate. In CESEE, both factors play a role: In most CESEE countries, the employment rate is below that in Western Europe (Figure 18). Low employment rates are a particular problem in SEE and some CIS countries. The labor productivity gap with Western Europe is still large, even though it has declined in the past twenty years. Figure 16. Big differences in growth among regions 3-year rolling average PPP weighted real GDP growth (percent) CESEE EU CESEE non-eu - CIS Source: WEO database and IMF staff calculations. 3 The IMF addressed this question in depth in the spring 16 issue of CESEE Regional Economic Issues. 11

12 Figure 17. Labor markets in EU new member states Unemployment rates (percent) LTU 6 6 EST CZE Source: Eurostat. Source: Eurostat. Figure 18. Labor utilization and productivity Beg in of the arrow indicates data for 1995, the end of the arrow - for 15 9 Employment rate (percent) BLR ALB ROU EST LTU RUS MKD SRB & MNE CZE HRV SVN DEU NLD 3 BIH Labor productivity (thousands of 1 US dollars, PPP-adjusted) Source: Total Economy Database, UN population statistics and IMF staff calculations. To raise labor productivity more investment is needed. The capital stock per worker in a typical CESEE economy is only about a third of that in advanced Europe. Domestic saving rates are too low in most the region; policies should therefore focus on institutional reforms that reduce inefficiencies and increase returns on private investment and savings. Boosting total factor productivity (TFP) is important as well. CESEE countries have to address structural and institutional obstacles that prevent efficient use of available technologies, or lead to inefficient allocation of resources. The recent IMF CESEE report suggests the largest efficiency gains are likely to 1

13 come from improving the quality of institutions (protection of property rights, legal systems, and healthcare); increasing the affordability of financial services (especially for small but productive firms); and improving government efficiency. Conclusion Since the fall of communism, there have been large differences in convergence of income levels with the US among CESEE countries. Much of these differences reflect differences in policies. Countries that reformed more and earlier saw faster growth than countries that reformed less or later. Macro-stability also helped, and countries that avoided boom-busts tended to grow faster. Continued convergence will require higher investment, higher TFP and higher employment rates. The capital stock per worker is still below that in Western Europe. Higher investment rates will require higher saving rates, lest large current account deficits emerge anew. Addressing structural and institutional obstacles would also help convergence, as it will support higher labor force participation and allow for a more efficient allocation of resources. 13

14 Bas B. Bakker Krzysztof Krogulski International Monetary Fund International Monetary Fund Bas B. Bakker is the IMF s Senior Regional Resident Representative for Central and Eastern Europe in Warsaw. He joined the IMF in 1993 and has held in four IMF departments, working on a range of countries, and policy and research issues. He has worked extensively on central and Eastern Europe, including as head of the Emerging Europe Regional Division and as mission chief for Bulgaria, and is the co-author of the book How Emerging Europe Came Through the 8/9 Crisis: An Account by the Staff of the IMF's European Department. He is a national from the Netherlands and obtained his PhD in economics from the University of Groningen. He is married with three kids. Krzysztof Krogulski is an Economist in the International Monetary Fund s Regional Office for Central and Eastern Europe in Warsaw. He is responsible for surveillance and analysis of cross-regional developments and supports outreach of the IMF in the region. He graduated from the Warsaw University. freepolicybriefs.com FREE NETWORK The Forum for Research on Eastern Europe n Emerging Economies is a network of academic r s on economic issues in Eastern Europe and the r r Soviet Union at BEROC ins, BICEPS i, CEFIR s, CenEA in, KEI Ki and SITE h l. The Network Policy ri Series and Policy Paper Series provides research-based analyses of i policy issues relevant to Eastern Europe and emerging markets.