II. How to Get Back on The Fast Track?

International Monetary Fund. European Dept.
Published Date:
May 2016
  • ShareShare
Show Summary Details

Despite the strong cyclical rebound, growth in CESEE remains well below precrisis levels. If lower potential growth in CESEE turns out to be the “new normal,” this would imply a much slower pace of income convergence with advanced Europe. This chapter explores the reasons behind the postcrisis growth slowdown by looking at labor, capital and productivity trends across the region. It also aims to identify the key gaps between CESEE and advanced Europe – with regard to capital deepening and productivity – as well as the specific institutional and structural features of CESEE economies that might explain these gaps. While there is no magic formula for fast convergence, the hope is that this chapter will provide some insights for ongoing policy discussions in the region on how to get back on a fast convergence track.

A. Historical Perspective

1. High-Speed Convergence

History shows that fast-track convergence is possible, but it still takes several decades. There are a few examples of successful and rapid catch-up with advanced economies from the starting per capita income levels that are comparable with the current levels of CESEE countries. The relatively small number of economies that have successfully converged to their advanced peers in the past century include Italy (1960-80), Spain (1980-09), Japan (1966-97), Korea (1988-10), and Taiwan (1968-08).1

Most of the rapid-convergence episodes share some common characteristics (Spence, 2008, Buera and Shin, 2013). First, the pace of growth typically accelerates following large-scale reforms. Second, a sizable fraction of growth is due to sustained growth of total factor productivity (TFP). Third, the investment-to-output ratio tends to increase during initial stages of growth acceleration and declines in the later stages. And finally, financial deepening occurs gradually along the transition path. Were the precrisis transition dynamics in CESEE economies any different?

2. CESEE Convergence before the Crisis

Following the dramatic economic transformation and initial period of instability, most Eastern European countries set on their convergence paths by the late 1990s. During the initial stages of transition, per capita income declined in most CESEE countries, with the notable exception of Poland. Countries that had higher starting income levels or had gone through greater political turbulence (wars, changing borders) incurred higher transition costs. In most CESEE countries, troughs in per capita GDP were reached by the mid-or late 1990s.

Compared to earlier fast-track convergence cases, CESEE economies had less favorable demographics and slower capital accumulation, which were offset by stronger TFP growth. Using the United States as a benchmark frontier economy in the post-World War II period, one can trace the evolution of relative per capita income, as well as other economic variables, during the growth acceleration episodes. Figure 2.1 shows the growth paths of CESEE countries (starting from their precrisis troughs, indicated in parentheses, and ending in 2008) juxtaposed against the growth paths of the so-called “miracle economies” of the past century to match starting income levels. A comparison between these two groups yields several observations:

  • The speed of convergence in CESEE before the global financial crisis – measured by the rate of growth of per capita GDP at PPP relative to the U.S. economy– was in many cases similar to that observed in the earlier fast-track convergence episodes.

  • However, the transition dynamics of CESEE differed from that of the miracle economies in several respects: on the upside, TFP growth has been stronger and the pace of financial deepening has been much faster than during the earlier fast convergence episodes, but on the downside, CESEE economies had slow-growing or shrinking labor force and slower pace of capital deepening (measured by the growth rate of the capital-labor ratio), as well as generally lower domestic saving and investment rates (Table 2.1).

Figure 2.1.Convergence

(Per capita income as a share of per capita U.S. GDP at PPP)

Source: Penn World Tables, Version 8.1.

Table 2.1.Convergence in CESEE before the Global Financial Crisis in Comparison to Previous Fast-Track Convergence Episodes
Average Annual Growth Rates

during the Initial 20 years of Transition 1
Average Ratios, as a Share of GDP

during the Initial 20 years of Transition 1

(per capita) 2
Labor 3Capital/Labor

Ratio 4
TFP 5Investment 6Domestic

Savings 7

Investment 8
Credit 9
Starting income level less than 0.2 of US GDP per capita
Macedonia (98)
Starting income level above 0.2 but less than 0.3 of US GDP per capita
Bulgaria (01)
Starting income level above 0.3 but less than 0.4 of US GDP per capita
Spain (60)
Korea (88)
Hungary (92)
Slovak Rep.(99)−0.01
Starting income level above 0.4 of US GDP per capita
Italy (60)
Japan (66)
Spain (85)−
Taiwan (82)
Korea (92)
Slovenia (92)
Czech Rep. (00)
Sources: Penn World Tables, Version 8.1; IMF, World Economic Outlook and International Financial Statistics databases.Note: Shaded blue areas are values that are below the benchmark (non-CESEE) country averages.

The number in parentheses refers to the first year of the latest growth spell for each country.

Growth in real GDP at purchasing power parity per capita.

Growth of the labor force.

Growth rate of capital stock per capita.

Growth of total factor productivity.

Investment rate, calculated as the share of investment in output.

Domestic saving rates, calculated as the share of output that is not consumed.

The share of public investment in GDP. Due to data availability, public investment/GDP ratio data for Taiwan Province of China starts from 1973, for Japan from 1980, for Poland from 1995, for Lithuania from 1999, for Latvia from 2000. for Bosnia and Herzegovina from 1998, for Serbia from 1997, for Hungary from 2000, and for FYR Macedonia and Belarus from 2005.

The growth rate of credit-to-GDP ratios. Credit-to-GDP ratio data for Italy start from 1963, for Spain from 1972, for Albania from 1994, for Lithuania from 1995, and for Bosnia and Herzegovina and Serbia from 1997.

Sources: Penn World Tables, Version 8.1; IMF, World Economic Outlook and International Financial Statistics databases.Note: Shaded blue areas are values that are below the benchmark (non-CESEE) country averages.

The number in parentheses refers to the first year of the latest growth spell for each country.

Growth in real GDP at purchasing power parity per capita.

Growth of the labor force.

Growth rate of capital stock per capita.

Growth of total factor productivity.

Investment rate, calculated as the share of investment in output.

Domestic saving rates, calculated as the share of output that is not consumed.

The share of public investment in GDP. Due to data availability, public investment/GDP ratio data for Taiwan Province of China starts from 1973, for Japan from 1980, for Poland from 1995, for Lithuania from 1999, for Latvia from 2000. for Bosnia and Herzegovina from 1998, for Serbia from 1997, for Hungary from 2000, and for FYR Macedonia and Belarus from 2005.

The growth rate of credit-to-GDP ratios. Credit-to-GDP ratio data for Italy start from 1963, for Spain from 1972, for Albania from 1994, for Lithuania from 1995, and for Bosnia and Herzegovina and Serbia from 1997.

Faster TFP growth and financial deepening in CESEE during 1990-2008 could, in part, be attributed to an unusually favorable global environment, characterized by a combination of high commodity prices, low interest rates, rapid expansion of global supply chains and buoyant trade. During this period, not only CESEE, but many other economies in Southern Europe, Latin America and Asia experienced an acceleration of growth (IMF, 2013). Amid easy funding conditions and given their initial low capital-labor ratios, CESEE economies would have been expected to have much higher rates of capital accumulation than what they actually had. The latter may be in part due to fast financial deepening that enabled households to boost consumption at the expense of saving in anticipation of higher incomes.

3. The Postcrisis Growth Slowdown

The global financial crisis followed by the euro area crisis led to steeper growth declines in CESEE than elsewhere. Real GDP growth in CESEE dropped to an average of 1.9 percent over 2011-15 from an average of 6.1 percent over 2002-08, showing a much sharper decline than in other emerging or advanced economies (Figure 2.2). Part of this decline was cyclical, and in fact, much of it has already reversed in the last two years outside the CIS (as discussed in Chapter I). But a large part of the growth slowdown is thought to be structural. The average potential growth in CESEE is estimated to have dropped to 2 percent, which is about half of the precrisis potential growth rate (Figure 2.3). This stands in sharp contrast, for example, to the full recovery of Korea’s potential growth within three years following the Asian crisis. That said, potential growth deceleration since the global financial crisis has been a global phenomenon.

Figure 2.2.GDP Growth and Convergence, 2002–15

Source: IMF, World Economic Outlook database.

Note: CESEE = Central, Eastern, and Southeastern Europe; CIS =Commonwealth of Independent States; LatAM=Latin America; EMs=Emerging Markets. The per capita income growth paths of CESEE countries are juxtaposed on Korea’s income growth path.

The decline in potential growth across CESEE appears to have been mainly driven by slower TFP growth but also by weaker investment. Indeed, our growth accounting analysis points to lower TFP growth and slower capital accumulation as the main reasons behind the potential growth slowdown after the crisis (Figure 2.3).2

Figure 2.3.Potential Growth, 2002-2015

(Average year-over-year growth rate, percent)

Sources: Penn World Tables, Version 8.1, Haver, and IMF staff.

Note: Potential growth is shown as a range of estimates based on three methods: a multivariate filter without financial friction, a multivariate filter with financial frictions, and the production function approach (Podpiera, Stepanyan, and Raei, forthcoming). Potential output decomposition into TFP, capital, and labor contributions is based on the production function approach. The TFP contributions include human capital. Human capital accounts for very small part of TFP change, on average close to 0.04 percentage points. Data availability limits precrisis period as follows: 2005-08 for Moldova, 2006-08 for Ukraine, 2007-08 for Russia and Macedonia.

At current growth rates, convergence is effectively off the fast track. In order to get back on the fast track, CESEE countries would need to lift potential growth closer to precrisis levels. This proposition is not new. What is new is that CESEE countries are now facing considerable headwinds from unfavorable demographics, sluggish global growth (or even secular stagnation) and likely tighter global financial conditions going forward. Against this backdrop, achieving similar growth rates as before the crisis may prove to be challenging. The rest of this chapter will explore possible future engines of growth across CESEE countries, by looking at different factors of production and how efficiently they are used.

B. Growth Drivers: Labor

While demographics in the region are generally unfavorable for economic growth prospects, CESEE countries have some scope to mitigate these negative effects by increasing participation rates, reducing structural unemployment, and raising life expectancy. Increasing the share of the workforce with tertiary education and implementing more active labor market policies could reduce skill mismatches, and thereby, contribute to faster income convergence.

1. How Does CESEE Compare to Advanced Europe?

CESEE countries are facing some of the worst declines in working-age population in Europe, reflecting both unfavorable demographics and emigration. This trend is expected to continue or worsen in some cases (Figure 2.4, and Box 2.1). CESEE countries also tend to have lower life expectancy than the EU-15, which is strongly correlated with the quality of healthcare.

Figure 2.4.Working Age Population Growth and Life Expectancy

Sources: United Nations, Populations Prospects; Organization for Economic Cooperation and Development; Eurostat; IMF, World Economic Outlook database; World Health Organization; and IMF staff calculations.

Note: CEE = Central and Eastern Europe; CESEE = Central, Eastern, and Southeastern Europe; CIS = Commonwealth of Independent States; SEE = Southeastern Europe; SEE-XEU = Southeastern European countries outside of the EU; WHO =World Health Organization.

Labor force participation rates in CESEE are comparable to those in advanced Europe, but there are pockets of underutilized labor and structural unemployment is high in some cases (Figure 2.5). Based on full-time equivalent participation rates, only a few CESEE countries (Bosnia and Herzegovina, Moldova and Turkey) are below the EU-15 average. But there are more countries (notably in SEE), where participation rates among women and seniors are much lower than elsewhere in Europe. Structural unemployment tends to be high in SEE as well. Stubbornly high unemployment rates in SEE economies are linked to the rigidity of labor market institutions and persistent outward migration (Kovtun et al, 2014; IMF, 2015c). Labor market practices in SEE countries have traditionally afforded workers high degree of protection and union coverage is also high compared to the rest of CESEE. While structural unemployment acts as a push factor for emigration (Box 2.1), it may also be exacerbated by heavy dependence on remittances, which are the flipside of emigration. Sizable remittances allow their recipients to extend periods of job search and push up reservation wages, thus reducing domestic workers’ willingness to accept lower-paid jobs. Also, migration itself may reduce pressures to reform labor markets.

Figure 2.5.Labor Force Participation and Unemployment Rate


Sources: United Nations Populations Prospects; Organization for Economic Cooperation and Development; Eurostat; IMF, World Economic Outlook; and IMF staff calculations.

Note: Data for ALB, BLR, SRB, and UKR are not adjusted for part-time employment, due to lack of data. CEE = Central and Eastern Europe; CESEE = Central, Eastern, and Southeastern Europe; CIS = Commonwealth of Independent States; SEE = Southeastern Europe; SEE-XEU = Southeastern European countries outside of the EU.

Most CESEE economies appear to score well on the quality of human capital on aggregate, but there are some gaps (Figure 2.6). Based on the standard aggregate human capital indices, most CESEE countries compare well with advanced Europe. However, a closer look at the distribution of the labor force by education level reveals that some CESEE economies lag in the share of the workforce with tertiary education (SEE and Turkey). As a result, shortages of high-skilled labor – the difference between the share of high-skilled labor in total employment and the same share in total population – are worse in some parts of the region than in the EU-15. In the Baltics and SEE, these shortages have been exacerbated by large and persistent outflows of younger and relatively more educated people since the 1990s (Box 2.1). Furthermore, skill mismatches – whereby a sizable portion of available skills are not employed in relevant occupations – tend to be worse in CESEE than in advanced Europe, in part due to deficiencies in labor market policies or institutions.

Figure 2.6.Human Capital, Education and Skills

Sources: Penn World Tables, Version 8.1; World Bank, World Development Indicators; Eurostat; SEO Economic Research (2012); and IMF staff calculations.

Note: CEE = Central and Eastern Europe; CESEE = Central, Eastern, and Southeastern Europe; CIS = Commonwealth of Independent States; SEE = Southeastern Europe; SEE-XEU = Southeastern European countries outside of the EU.

Box 2.1.Economic Impact of Emigration on CESEE Countries

Emigration can have profound effects on economic outcomes in sending economies. Economic migration driven by individuals’ choices is part of economic development and has likely led to positive outcomes for CESEE emigrants themselves and for the EU as a whole. However, emigration—through its externalities—may have also slowed growth and convergence in CESEE countries. Such externalities could arise from sizable outflows of skilled labor, which can create skill shortages, and may impair productivity growth and convergence.

Figure 2.1.1Determinants of Bilateral Emigration of Skilled Workers

(Standardized coefficients, Sending countries: CESEE; Receiving countries: OECD)

Source: IMF staff estimates.

Note: Estimates are based on a panel data gravity model of cumulative outward emigration growth over 1990 to 2010.

Figure 2.1.2High Remittance Receiving Countries

(Percent of GDP)

Source: IMF staff estimates.

Note: High remittance countries: ALB, BIH, KOS, MDA, MNE.

Figure 2.1.3Per Capita Income in Purchasing Power Standard, 2014

(Percentage points; additional reduction in per capita GDP gap with the EU28)

Sources: Eurostat and IMF staff calculations.

Note: Coefficients derived from regressions of value added per worker on emigration ratios and control variables are used to estimate the contribution of emigration to cumulative per capita output changes during 1995-2012.

Over the past 25 years, close to 20 million people have emigrated from CESEE, accounting for 6 ¼ percent of the region’s working-age population. Two-thirds of CESEE countries are affected by net emigration, where it has often exacerbated adverse demographic trends and dampened working-age population growth by about 0.5-1.0 percentage points per year since 1990, and contributed to shortages of high skilled labor, especially in the Baltics. Lower average income vis-à-vis more advanced economies, but also poor institutional quality, and weak economic conditions at home are important factors behind emigration, particularly of skilled labor.

Remittances have been both a blessing and a curse. Analysis suggests that higher remittances are associated with lower labor supply incentives, and large remittance inflows may contribute to real exchange rate appreciation, and adversely affect the tradable sector. But they also appear to have supported consumption, and private investment, and facilitated financial deepening in high-remittance receiving countries.

Overall, emigration appears to have slowed growth and income convergence in CESEE. Analysis suggests that in 2012, cumulative real GDP growth could have been around 7 percentage points higher on average in CESEE in the absence of migration during 1995-2012. As a result, on average, CESEE members of the EU could have narrowed their per capita income gap with the EU average by an additional 2 percentage points. Emigration may have also created pressures on social security systems and hindered growth through increased growth-unfriendly labor taxes. Raising the labor market participation rates and better leveraging remittances to promote investment could help offset the negative impact of emigration. Improving institutions and economic policies would also encourage potential migrants to stay, promote return migration, and attract new immigrants.

1/ This Box was prepared by Faezeh Raei based on Atoyan et al (forthcoming).

C. Growth Drivers: Capital

Capital gaps relative to advanced Europe are still large, while investment rates are not sufficient for a rapid catch-up. Investment is partly held back by crisis legacies and subdued long-term growth prospects. Many CESEE countries would need higher domestic saving rates to sustain high enough investment rates to achieve successful convergence without hitting external sustainability limits.

1. How does CESEE Compare to Advanced Europe?

After more than 20 years of transition, there is still significant scope for capital deepening in CESEE (Figure 2.7). With the exception of Slovenia and the Czech Republic, capital stock per capita in a typical CESEE economy is about one third of that in advanced Europe. Most significant gaps are in infrastructure (buildings and civil engineering) and machinery equipment.

Figure 2.7.Capital Stocks

Source: Penn World Tables, Version 8.1.

Notes: Capital gap is the difference between the capital stock of country X and the U.S. capital stock. Asterisks denote countries that report only the basic structure of capital stock.

2. Was Capital Accumulation before the Crisis too Fast or too Slow?

Given large capital gaps, precrisis capital accumulation in CESEE was relatively slow. The relative scarcity of capital and the higher rate of return on investment would have justified investment rates higher than those observed in CESEE before the crisis. While precrisis investment rates in CESEE exceeded those in advanced Europe, the difference was fairly small. Average precrisis investment rates were relatively higher than in advanced Europe in CEE, SEE, and the Baltics, but lagged behind advanced economies in the CIS and Turkey (Figure 2.8).

Figure 2.8.Investment Rates in CESEE

(Percent of GDP)

Sources: Haver Analytics; Eurostat; and Penn World Tables, Version 8.1.

Note: CEE = Central and Eastern Europe; CESEE = Central, Eastern, and Southeastern Europe; CIS = Commonwealth of Independent States; SEE = Southeastern Europe; SEE-XEU = Southeastern European countries outside of the EU.

After the crisis, investment rates fell across all of Europe. Investment rates declined in the Baltics, CEE, and SEE, but also in the EU-15. The declines could have been even steeper in the Baltics, CEE and SEE had it not been for the sizable boost from EU Structural and Cohesion Funds (Figure 2.8). In contrast, across in most CIS economies and Turkey, post-crisis investment rates have improved significantly and moved ahead of their CESEE peers.

How do we assess the adequacy of the speed of capital accumulation across CESEE? This is not straightforward. In what follows, we use two benchmarks: a model-based steady state investment rate (“golden rule”) and an investment rate consistent with stylized transition dynamics derived from the historical experience of other European countries that have achieved convergence to present-day euro area income levels (henceforth, a “historical benchmark”). The benchmark values can be calculated for each country and each point in time, given the TFP and population growth rates, as well as the country’s capital-labor ratio (Figure 2.9).

Figure 2.9.Precrisis Investment Rates and Benchmarks

(2002–08 averages, percent of GDP)

Sources: Haver Analytics; Penn World Tables, Version 8.1; Eurostat; and IMF staff calculations.

Note: The “golden rule” is estimated with the social rate of time preference set equal to the estimated euro area average (5 percent) plus/minus 1 percentage point. CEE = Central and Eastern Europe; CESEE = Central, Eastern, and Southeastern Europe; CIS = Commonwealth of Independent States; SEE = Southeastern Europe; SEE-XEU = Southeastern European countries outside of the EU.

How should one interpret the deviations of actual investment rates from the benchmark rates? If a country’s investment rate falls short of both benchmarks, this is likely a sign of investment undershooting. On the other hand, if it is above the golden rule rate, it does not necessarily imply overshooting (as the golden rule can be viewed as a lower bound – see Box 2.2). However, if the investment rate exceeds both benchmarks, this could be an indication that the investment rate is unsustainable.

With that interpretation in mind, the key finding is that investment rates tended to be on the low side, with some exceptions:

  • The precrisis investment rates in Turkey and most of the CIS were too low, with the exception of Belarus. Investment rates in Russia and Turkey were well below both benchmarks, while investment rates in Moldova and Ukraine were sub-optimal relative to the golden rule.

  • Across the Baltics, CEE, and SEE countries, there is more variation. The precrisis investment rates in Lithuania, the Slovak Republic, Bulgaria, and Poland also fall short of both benchmarks. In contrast, precrisis investment rates were well above both benchmarks in Croatia, Estonia, and Slovenia, suggesting possible overshooting.

Box 2.2.Golden Rule and Historical Benchmark Investment Rates

The golden rule can be interpreted as a lower bound to which an investment rate in a country would eventually converge as it approaches its own steady state level underpinned by its deep structural characteristics and exogenous parameters.

The neo-classical growth model modified to allow for exogenous growth of labor-augmenting productivity (Cass-Koopmans model) predicts that—for given parameters of the aggregate production function, social rate of time preference, depreciation, exogenous growth rates of the labor force and labor-augmenting productivity, and initial conditions with positive values—an economy converges to a steady-state equilibrium, in which income, consumption, and capital all grow at a fixed rate equal to the sum of the growth rates of labor force and labor-augmenting productivity.

Under typical calibration of the parameters, the model implies that the investment rate would fall monotonically as the economy converges to its steady state. As such, the closed-economy, “golden rule” saving/investment can be interpreted as a lower bound for the investment rate along the CESEE countries path of convergence to euro area income levels. The interpretation of the “golden rule” as a lower bound of the optimal investment rate also holds in the case of a similar open economy, for which the world interest rate is lower or equal to the value in the steady state of the closed economy. The main advantage of the “golden rule” approach is that it provides a benchmark that is invariant with respect to country’s initial conditions (i.e., it is not sample-dependent as is the case with all regression-based approaches). The main disadvantage is that it requires knowledge of the unobservable social rate of time preference (Miranda, 1995).

The historical benchmark provides a yardstick investment rate (for a given K/L ratio and technology) that is consistent with capital accumulation path of selected advanced European economies during 1951-2011 that has proven to be sustainable. The main advantage of this approach is that it does not require any assumptions about the social rate of time preference and the position of the country on the saddle-path. The main disadvantage is that it assumes similarity in economic structures of CESEE countries and their advanced peers. See Annexes V and VI for details.

3. Has Investment Growth Slowed Since the Crisis?

With the crisis, investment slumped in most CESEE economies. The global financial crisis triggered a “sudden stop” in capital flows. At the same time, incomes fell and risk premia increased, souring the outlook and making legacy debt burdens unsustainable for many borrowers (see Spring 2015 REI). The resulting push for deleveraging led to higher savings and further declines in investment, broad-based in some cases (for example, in Slovenia), but concentrated in construction in other cases (Hungary, Lithuania, and Estonia) (Figure 2.10). As a result, the net saving-investment balances improved in most countries.

Figure 2.10.The Postcrisis Investment Slump in CESEE

The postcrisis decline in investment rates likely reflects both cyclical and structural factors, though the two are difficult to pry apart:

  • In those countries that had precrisis investment booms, investment rates have become unsustainable after the onset of the crisis, as borrowing costs rose and near-term growth prospects soured. Indeed, investment rates tended to decline more in countries with larger precrisis investment gaps – defined as the actual investment rate minus the benchmark rate (Figure 2.10). As discussed above, the largest positive investment gaps were observed in Croatia, Estonia, and Slovenia. Precrisis investment overshooting was often associated with excessive credit growth that led to a build-up of debt. After the crisis, debt overhang and attendant high nonperforming loans became a drag on investment (see Spring 2015 REI).

  • The postcrisis investment slump may, in part, also reflect a perceived structural shift in trend growth (Figure 2.10). The estimated golden rule and historical benchmark investment rates for CESEE countries shifted lower after the crisis. The decline in golden rule benchmarks, for example, was mainly due to lower TFP growth and, to a lesser extent, worsening demographics. 3 Thus, some of the postcrisis downward correction in investment rates may be mirroring the declines in the benchmark investment rates for CESEE countries that reflect changes in potential growth drivers after the crisis.

In most economies, investment rates are now below their estimated historical benchmark rates, though not the golden rule. While it is not clear whether postcrisis TFP growth slump will persist (this will be discussed below), if it does persist, then the steady-state investment rates for CESEE – as implied by the golden rule – may be permanently lower. However, most CESEE countries would need to have much higher investment rates – according to historical benchmark – in order to get back on the fast convergence path. This would require substantial efforts to boost TFP and, in a number of economies, higher saving rates to fund the additional investment without running into the external financing constraints. We turn to these issues in the next sections.

4. What Explains Investment Undershooting in Some CESEE Countries?

So, investment rates tended to be on the low side before or also after the global financial crisis in a number of CESEE economies, as discussed in the previous sections. Was this because of low domestic saving rates, limited external borrowing space or other reasons?

Indeed, in several CESEE economies, saving rates have been and remain low in comparison with actual investment rates, optimal investment benchmark rates and earlier fast-track convergence episodes:

  • Comparing saving rates across CESEE countries: Saving rates are particularly low in SEE non-EU, Ukraine, Moldova and Turkey (Figure 2.11). In FYR Macedonia, Moldova, and Bosnia and Herzegovina, low saving rates are partly offset by high net remittance inflows.

  • Comparing saving rates with actual investment rates (Figure 2.11): Gross domestic saving rates were below investment rates in many countries before the crisis (the Baltics, SEE, and Moldova). After the crisis, saving rates increased and matched or exceeded the investment rates in SEE EU and the Baltics, while saving-investment gaps persisted in the SEE non-EU, Moldova, Belarus, and Turkey.

  • Comparing saving rates with optimal investment benchmark rates (Figure 2.12): Domestic saving gaps – the difference between actual gross domestic saving rates and estimated benchmark investment rates – are large and persistent in SEE non-EU, Moldova, and Turkey.

  • Comparing CESEE saving rates with investment rates in previous fast-convergence episodes: While average domestic saving rates in previous fast-convergence episodes were around 28 percent of GDP, the average was just 12 percent for CESEE and slightly under 20 percent in CEE (Figure 2.11). In most CESEE countries, investment and saving rates remain well below 25 percent. While 25 percent is not a magic number, the received wisdom in the growth literature is that countries should maintain such investment rate for a sufficiently long period in order to achieve successful convergence with advanced economies (Spence, 2008).

Figure 2.11.Saving-Investment Balance

(Percent of GDP)

Sources: Penn World Tables, Version 8.1; and IMF staff calculations.

Note: The domestic saving rate is calculated as GDP less consumption, in percent of GDP. The national saving rate equals the sum of the balance of the current and capital accounts of the balance of payments in percent of GDP plus the investment rate. Remittances in Southeastern European countries outside of the EU and foreign direct investment profits in CEE account for the bulk of the difference between national and domestic saving rates. The saving-investment balance is calculated using national saving rates. The redline benchmark denotes a level of saving/investment rate that is consistent with fast-track convergence. CEE = Central and Eastern Europe; CIS = Commonwealth of Independent States; SEE = Southeastern Europe, SEE-XEU = Southeastern European countries outside of the EU.

Figure 2.12.Domestic Saving Gap (Domestic Saving Less Optimal Investment)

(Percent of GDP)

Sources: Penn World Tables, Version 8.1; and IMF staff calculations.

Note: CEE = Central and Eastern Europe; CESEE = Central, Eastern, and Southeastern Europe; CIS = Commonwealth of Independent States; SEE = Southeastern Europe; SEE-XEU = Southeastern European countries outside of the EU.

External borrowing space may have been constrained by sustainability concerns in some cases. Before the crisis, quite a few CESEE countries exceeded sustainable borrowing (the Baltics, Bulgaria, Romania, the Slovak Republic, Serbia, FYR Macedonia, Montenegro, Bosnia and Herzegovina, and Turkey). Postcrisis adjustments have dragged down consumption and investment and created external borrowing space for many CESEE countries, except Albania, Bosnia and Herzegovina, Belarus, and Turkey (Figure 2.13).

Figure 2.13.External Borrowing Space

(Percent of GDP)

Source: IMF staff calculations.

Note: Negative values indicate available borrowing space; positive values indicate breaching borrowing limits consistent with sustainable external balance. The figure shows the gap between debt-stabilizing and actual current account primary balances. SEE-XEU = Southeastern European countries outside of the EU.

Finally, it is also possible that productivity gains may have been too slow, thereby constraining return on capital and keeping both investment and saving rates low. The next section will take a closer look at the composition of domestic savings across CESEE countries.

5. A Closer Look at Private Sector Saving Rates in CESEE Countries

A comparison of savings rates across EU countries reveals that household saving rates tend to be lower in CESEE EU countries than in advanced Europe, while the opposite is the case for corporate saving rates (Figure 2.14). These differences can be largely explained by differences in the structural characteristics of these economies. Specifically, households’ saving rates tend to be positively correlated with GDP levels, labor share of value added, and labor force participation, but negatively related to public debt levels, remittances, tax rates and age dependency (see Annex IV). Corporate saving rates appear to be lower in countries with higher corporate debt, corporate taxes and government expenditures. Faster wage growth relative to productivity growth may have a dampening effect on corporate savings as well (Figure 2.14). Since the crisis, saving rates of CESEE firms and households have generally improved, with economy-wide saving rates now comparable across CESEE and advanced EU countries.

Figure 2.14.CESEE: Saving Rates and Wage-Productivity Growth

Sources: Penn World Tables, Version 8.1; Eurostat; European System of National and Regional Accounts, 1995 Annual Sector Accounts; and IMF staff calculations.

Note: Interest payments include financial intermediation services indirectly measured (FISIM).

CESEE countries have some scope to boost corporate and household savings:

  • Corporate savings account for the bulk of total domestic savings in CESEE (Figure 2.14). The fact that a large share of earnings are distributed rather than reinvested suggests that the return on investment may not be high enough. Thus, some additional savings could come from encouraging the reinvestment of FDI profits (CEE and Croatia) and reducing dividend payouts (Baltics) through tax incentives and productivity-enhancing reforms. In other cases (Bulgaria), improving financial affordability could help lift corporate savings as well. After the crisis, corporate saving rates were boosted by reductions in distributed profits, a shift from direct to indirect taxes, and realigning wages to productivity (Baltics and CIS). Indeed, the experience of Korea shows that a sustained positive productivity-wage growth differential would support higher corporate savings and investment.

  • Household saving rates are, in general, lower in CESEE than in other EU countries (Figure 2.14). This reflects a higher share of autonomous consumption in household incomes and smaller labor share in national incomes in CESEE. Broad social safety nets and rapid financial deepening before the crisis may have provided additional disincentives to save. After the crisis, household saving rates improved due to lower consumption and net taxes/transfers. Further improvements may be needed, especially in Romania and Bulgaria, where domestic household saving rates continue to be negative. Given the demographic profiles, household savings can also be stimulated by greater emphasis on Pillar II and III pension schemes (Box 2.3 discusses some of these measures in the context of Turkey).

Box 2.3.Raising Domestic Savings in Turkey

Figure 2.3.1Turkey: Saving Rates

(Percent of GDP)

Sources: CBRT; IMF, WEO; and IMF staff calculations.

Turkey has a large external imbalance, mostly due to a structurally low private saving rate. While the private sector saving rate averaged 18 percent over 1998–2003, it dropped to 9 percent in 2013 and has stayed below 13 percent since 2010. The decrease in the saving rate was particularly pronounced in the years since 2003. Meanwhile the public saving rate stands at around 3 percent, while the investment rate increased from around 17 percent in 2002 to 20 percent in 2014. Thus, domestic savings, private and public, no longer covered investment, opening up a large gap between savings and investments and hence a current account deficit, which averaged over 6½ percent of GDP between 2010 and 2015.

The decline in the private saving rate was mainly a consequence of economic stabilization and financial deepening. The fast drop in the saving rate in the years directly following the 2001 crisis suggests that the quick implementation of a thorough macroeconomic stabilization program may have played a large role. The more gradual, but still rapid, decline thereafter is consistent with rapid financial deepening, primarily through bank credit becoming available to a large proportion of households. Urbanization likely also played a role: it is generally thought to lower the need for precautionary savings, as better and more public services are available in urban centers, and income volatility is lower for city dwellers.

Figure 2.3.2.Third Pillar Pensions

Source: Pension Monitoring Center.

The Turkish authorities have emphasized raising the private saving rate as an important policy goal to reduce the economy’s external vulnerability. To this end, they have considered various policy options. They have introduced a subsidized third pillar pension scheme, and, more recently, a savings subsidy for dowry accounts. They have also piloted an auto-enrollment funded pension scheme, and are committed to scaling up this pilot. Lastly, proposals to reform the severance pay scheme by making it a funded and transferable benefit have been put forward. The authorities have also used macro-prudential tools to limit credit growth, prompted by systemic prudential risks in the banking system that may be caused by very high credit growth.

Going forward, full and swift implementation of the pension and severance pay reform plans is key. Given the urgency of reducing vulnerabilities and the time lag with which new policies will affect the saving rate, efforts should begin as soon as possible. In addition, Turkey’s relatively young population and declining fertility rate imply the country is enjoying a demographic dividend. This provides a window of time to increase savings in anticipation of almost inevitable population aging in the future. Macro-prudential policies limiting credit growth should also remain part of the policy mix. IMF (2016b) provides more details and background on these policy options.

1/ Prepared by Alexander Tieman, drawing on the IMF (2016b), Turkey: Selected Issues Papers.2/ These years included major economic crises, when the saving rate fluctuating between 12.4 and 28.5 percent.

D. Growth Drivers: Productivity

Significant productivity gaps relative to advanced Europe can be largely explained by structural and institutional obstacles that limit efficient use of available technologies or efficient allocation of resources in CESEE. In the absence of favorable external tailwinds that helped to boost productivity growth before the crisis, productivity-enhancing reforms become a must for CESEE countries.

1. How Does CESEE Compare with Advanced Europe?

The TFP levels in CESEE are notably lower than in advanced Europe and the postcrisis TFP growth slowdown was much sharper (Figure 2.15). Because TFP is typically estimated as a residual after accounting for the contributions of other factors of production, it could reflect either supply side (technological) or demand side drivers.4 Because there is no good model of TFP, it is often referred to as a “measure of ignorance”. For example, the much more dramatic declines in TFP growth in CESEE compared to advanced Europe (Figure 2.16) may suggest that either CESEE were much more sensitive to some common global TFP growth drivers or that their precrisis potential growth estimates were overstated.5 In what follows, we’ll attempt to shed some light on TFP levels and growth across the CESEE region.

Figure 2.15.Total Factor Productivity Levels

(Average 2002-14, EU-15 =100)

Sources: Penn World Tables, Version 8.1 and IMF staff calculations.

Note: CESEE = Central, Eastern, and Southeastern Europe.

Figure 2.16.Difference Between Average TFP Growth before and after Crisis

(Percentage points)

Focusing on the supply side, aggregate productivity reflects the level of technological progress, technical efficiency, and allocative efficiency. Technical efficiency is the efficiency with which firms use available technology, while allocative efficiency is the extent to which firms with higher productivity have more resources. Next sections will explore each of these aspects in turn.

Technical Efficiency

CESEE countries appear to be less efficient users of available technologies than advanced Europe. In the stochastic frontier analysis (Annex VII), the relative technical inefficiency of a country is measured by its distance from the frontier, with the latter representing the maximum amount of output that can be obtained from given inputs. Figure 2.17 shows the estimated technological frontier and the position of CESEE and advanced European economies relative to the frontier as of 2014. Technical efficiency is estimated relative to the frontier (=100). For example, if a country’s score is 60, this means that it uses available technology 40 percent less efficiently than the frontier economy.

Figure 2.17.Technological Frontier, 2014

Lower technical efficiency may be due to structural or institutional obstacles that prevent the diffusion and efficient use of available technologies. Our empirical analysis of the sample of advanced economies and CESEE identifies several determinants of the distance to the technological frontier:

  • Structure of the economy: A relatively low share of the service sector and, in some cases, a still-sizable share of agriculture places countries further away from the frontier.

  • Quality of institutions: Greater judicial independence, impartial courts, and better protection of property rights reduce the incidence of corruption, and tend to be associated with higher efficiency; and so does higher life expectancy, which partly reflects the quality of healthcare.

  • Restrictiveness of regulation: Lighter general business and FDI-specific regulations tend to increase efficiency as well.

In contrast, the levels of research and development (R&D) spending, infrastructure gaps and labor market flexibility do not appear to be statistically significant determinants of technical efficiency in our analysis. That said, these findings should be seen as tentative, given the limitations inherent in gauging technical efficiency with macroeconomic data, and sample specific issues6.

Structural reforms – most notably upgrading legal systems – could bring significant efficiency gains. Figure 2.18 presents estimates of potential efficiency gains from improving structural and institutional characteristics of CESEE countries to EU-15 average level based on the stochastic frontier analysis. In the case of Croatia, for example, efficiency gains from all structural reforms shown in Figure 2.18 would allow it to close the gap with the frontier economy. But, in the case of Estonia, potential gains are limited since in many of these areas Estonia is already very close to the EU-15 average levels.

Figure 2.18.Potential Efficiency Gains From Structural Reforms


Sources: Penn World Tables, Version 8.1; and IMF staff calculations.

Note: Potential efficiency gains for Albania and Serbia are tentative estimates, since due to data limitations these countries are not included in the regression analysis. CEE = Central and Eastern Europe; CESEE = Central, Eastern, and Southeastern Europe; CIS = Commonwealth of Independent States; SEE = Southeastern Europe; SEE-XEU = Southeastern European countries outside the EU.

Allocative Efficiency

Aggregate productivity is influenced by the structure of the economy. Countries with a larger share of employment in low-productivity sectors, such as agriculture, tend to have lower aggregate labor productivity. As one would expect, countries where a larger share of workers moved away from agriculture had higher productivity growth over 2000-14 (Figure 2.19).

Figure 2.19.Structure of the Economy and Productivity

Sources: Eurostat; and IMF staff calculations.

Note: Aggregate labor productivity is the ratio of real value added in the total economy as a percentage of the number of employees.

More flexible labor regulations may facilitate better allocation of resources. More flexible labor markets tend to be positively associated with aggregate productivity growth (Figure 2.20).

Figure 2.20.Labor Regulation and Productivity

Sources: Eurostat; World Economic Forum; and IMF saff calculations.

Note: The Labor Market Flexibility Index captures hiring and firing regulations. Higher values represents more economic freedom.

The link between productivity and allocative efficiency within sectors can be analyzed using firm-level data. Aggregate productivity, defined as average firm-level productivity weighted by the number of employees, can be decomposed into two elements (Olley and Pakes, 1996): (1) underlying productivity, which is captured by the unweighted average of firms’ productivity and (2) productivity stemming from more efficient resource allocation across firms, whereby firms with higher productivity have more resources. The second component—allocative efficiency—is measured by the covariance between firm’s productivity and its share of employment in a given industry (Box 2.4).

Box 2.4.Assessing Allocative Efficiency Using Firm-level Data

The allocative efficiency score measures the relative productivity gain that a country enjoys owing to its actual allocation of employment across firms relative to a case where employment allocation is random. For instance, in Sweden, the country’s aggregate productivity is 40 percent higher than what it would have been if resources (employment) were allocated randomly (see Figure 2.21 in the main text). Firm-level productivity is measured as the log of the ratio of a firm’s turnover to the number of employees. Allocative efficiency is measured at industry level (according to NACE 2 Rev. first two digits), and then aggregated to the country level as the weighted average, weighted by each industry’s labor share.

Allocative efficiency is higher when larger firms are more productive. Between two countries with a similar level of productivity among large firms (say, Poland and the Slovak Republic), a country with less productive small firms (Slovak Republic) exhibits a higher allocative efficiency score than the country with more productive small firms (Poland). This means that in Poland there is more room for further productivity gains if more resources are allocated to more productive firms.

The allocative efficiency score appears to be very sensitive to the productivity distribution by firm-size in a sample used in the analysis. For example, it measures differently depending on whether micro firms (less than 20 employees) are included or excluded from the sample, as micro firms represents 80 percent of sample firms while accounting for only 20 percent of employment and turnover. This is particularly the case in countries where the productivity of micro firms is relatively high (Poland and Russia), which implies that their inclusion brings down these countries’ allocative efficiency scores significantly (see Figure 2.21 in the main text). It should be noted that further analysis may be needed here, as the unusually high productivity of micro firms may be due to mis-reporting which is linked to specific threshold-based regulations. In most CESEE countries, however, allocative efficiency appears lower when micro firms are excluded—particularly in the case of Bulgaria and Latvia, where productivity of micro firms is significantly lower. See Annex VIII for more details.

Some CESEE countries lag behind in allocative efficiency. Figure 2.21 shows that some CESEE countries, such as Serbia and Slovenia, have much room to improve their productivity through more efficient allocation of resources.

Figure 2.21.Allocative Efficiency Scores, 2013

Sources: ORBIS; and IMF staff calculations.

Note: Allocative efficiency score is a correlation between firm’s labor share and its productivity. CEE = Central and Eastern Europe; CESEE = Central, Eastern, and Southeastern Europe; CIS = Commonwealth of Independent States; SEE = Southeastern Europe; SEE-XEU = Southeastern European countries outside the EU.

Structural reforms that help improve allocative efficiency would also help narrow the productivity gaps. We find several structural indicators to be significant determinants of allocative efficiency: (1) quality of institutions (government efficiency); (2) labor regulation (flexible wages); and (3) financial development (affordability of financial services). Notably, these are exactly the areas where one would expect improvements to yield the largest benefits for relatively small but productive firms. The analysis suggests that more efficient resource allocation through improvement in government efficiency and affordability of financial services up to the level of Sweden could bring significant potential productivity gains for CESEE economies (Figure 2.22).

Figure 2.22.Potential Productivity Gains from Improved Allocative Efficiency

(Relative labor productivity, Sweden=100)

Sources: ORBIS; and IMF staff calculations.

Note: Labor productivity is calculated as average firm-level log productivity, weighted by the number of employees. The potential gains through improved allocative efficiency are calculated based on the estimates from regression analysis and the gap between structural indicators of each country and Sweden.

In sum, CESEE countries could realize sizable gains from improving both technical and allocative efficiency. In the case of technical efficiency, the biggest improvement could be achieved by upgrading institutions (legal systems), while in the case of allocative efficiency, the largest benefits would stem from greater affordability of financial services, especially for SMEs. On both fronts, more country-specific research is needed for more precise diagnostic.

2. What Explains the TFP Growth Slowdown in CESEE After the Crisis?

The recent TFP slowdown is broad-based, suggesting an important role of common factors. Eichengreen, Park, and Shin (2015) use historical episodes of sharp and sustained decelerations in TFP growth to identify common and country-specific factors behind the TFP slumps. They argue that neither the secular stagnation hypothesis nor the middle-income trap hypothesis can explain the recent slowdown in TFP growth. Secular stagnation would have affected mostly advanced economies, while the middle-income trap would have only affected middle-income countries. Therefore, the authors argue that there are global factors in addition to country-specific ones that are behind the recent deceleration in TFP growth. They find that oil price shocks and increases in risk are among the significant global factors, while human capital, investment rates, and poor political systems are identified as country specific factors behind the TFP slowdown.

What factors – global or country-specific – have played a larger role in the postcrisis TFP growth slowdown in CESEE? We consider several hypotheses:

Hypothesis 1: The TFP growth slowdown in CESEE has been largely driven by lower potential growth of trading partners. Basically, we conjecture that the observed decline in TFP growth is simply a result of lower potential output and aggregate demand of the CESEE economies’ main trading partners – the euro area countries. Our analysis suggests that strong TFP growth in CESEE before the crisis was indeed largely driven by common or external rather than country-specific factors (Figure 2.24). More specifically, we find that trading partners’ potential growth rates, as well as policy uncertainty, expansion of global supply chains, and global trade are closely correlated with the common component of TFP growth in CESEE (Figure 2.24).7

Figure 2.23.Change in Technical Efficiency, 2012-14

Figure 2.24.Common and Idiosyncratic Components of Total Factor Productivity Growth, 2001–14

Sources: Penn World Tables, Version 8.1; World Bank, World Development Indicators; World Economic Forum; Baker, Bloom and Davis (2015); and IMF staff calculations. Note: The common component of TFP growth used in the scatter plots is the other common component obtained from the regression analysis (see Annex IX). TFP = total factor productivity.

Hypothesis 2: The TFP growth slowdown in CESEE reflects slower technical progress in frontier economies, in other words, it reflects a TFP growth slowdown in advanced economies. We find some empirical support for this hypothesis as well. Technological frontier is estimated to have been expanding by 1 percent per year before the crisis, on average, but has remained unchanged since 2007. Meanwhile most of the advanced countries that were close to the technological frontier before the crisis appear to have moved away from it since the crisis.

Hypothesis 3: Less efficient economies suffered larger declines in TFP growth than those that were closer to the technological frontier before the crisis. Indeed, one might expect that countries that have significant structural or institutional obstacles to efficient use of technology or resource allocation could have experienced a larger TFP growth slowdown due to their limited ability to adjust. However, we find that countries with low initial levels of technical efficiency have experienced larger subsequent improvements in efficiency (Figure 2.23), likely reflecting improvements in their structural and institutional characteristics. However, the pace of these improvements has slowed after the crisis.

In sum, it appears that the strong productivity growth enjoyed by CESEE before the crisis was largely associated with favorable external or common factors. This also means that CESEE countries will have to do more on their own rather than rely on external tailwinds if the current less supportive global environment becomes a “new normal”.

E. Key Takeaways and Reform Priorities

CESEE may face much slower pace of convergence unless they step up reform efforts. CESEE countries made significant progress along the convergence path during 1990-2008, on the back of strong TFP growth and to a lesser extent, investment. After the crisis, TFP growth slowed significantly across most economies, raising concerns that CESEE may be entering a prolonged period of lower global growth, less trade and capital flows, and less scope for expansion of global supply chains. This also means that external financial conditions may be less supportive, implying that CESEE may have to rely more on domestic savings. Coupled with negative demographic trends, these challenges imply the need to put greater emphasis on productivity-enhancing reforms as well as on active labor market policies. For example, in the case of Korea, a strong acceleration of TFP growth after the Asian crisis was largely due to large-scale structural reforms, including corporate restructuring and upgrading corporate governance, which in addition to a more benign global environment at the time, helped it to stay on a fast convergence path.

1. What are the CESEE Economies’ Strengths and Weaknesses?

Labor: In contrast to many emerging markets that have enjoyed the so-called “demographic dividend”, CESEE countries have experienced some of the worst declines in the working age population due to unfavorable demographics and emigration. On the upside, the relatively high level of human capital is a plus, though sizable skill gaps and skill mismatches remain.

Capital: Capital gaps relative to advanced Europe are still large, while investment rates are not sufficient for rapid catch-up. Currently, investment is held back by crisis legacies and subdued long-term growth prospects. More generally, investment and saving rates have been fairly low across CESEE, when compared to the optimal benchmarks or to previous fast-track convergence episodes. This could be because of still low level of productivity coupled with the negative productivity-wage growth differential, which reduce returns on capital and incentives to save and invest. Furthermore, limited access to financial services may constrain the expansion opportunities for small, but productive firms. In some cases, space for external borrowing may be constrained due to already high external debt.

Productivity: Productivity gaps between CESEE and advanced Europe are significant due to both technical and allocative inefficiencies.

2. What should be the reform priorities?

CESEE countries rank below advanced Europe on a number of institutional and structural characteristics (Figure 2.25). This is well known. The colors in Figure 2.25 indicate the relative rankings for each of the characteristics in a sample of economies from CESEE and from Organization for Economic Cooperation and Development (OECD). The Baltics and some CEE countries tend to outperform other CESEE economies, but are still quite far from the frontier.

Figure 2.25.Institutional Quality: Relative Rankings of OECD and CESEE Countries

Note: CEE = Central and Eastern Europe; CESEE = Central, Eastern, and Southeastern Europe; CIS = Commonwealth of Independent States; SEE = Southeastern Europe; SEE-XEU = Southeastern European countries outside the EU. See Annex X for more details.

Which structural reforms would yield the largest growth benefits? The answer depends on the growth elasticity of different reforms and on the size of institutional gaps in a given country. Out of the host of indicators shown in Figure 2.25, Figure 2.26 includes only a subset of institutional and structural characteristics that turned out to be statistically significant and robust in our empirical analysis of productivity or the ones that directly affect the supply of production factors, such as labor and capital (infrastructure). The colors are based on the estimated potential gains (percent increase in GDP level) from improving these characteristics from their current levels to EU-15 levels. The main advantage of this approach is that potential gains from different reforms can be compared across the entire spectrum of policy choices. So, putting it all together suggests that the largest potential growth benefits can be achieved by (1) improving protection of property rights and upgrading legal systems; (2) increasing government efficiency; (3) improving affordability of financial services (especially for small productive firms); (4) upgrading infrastructure; (5) increasing quality of healthcare (life expectancy); and (6) facilitating structural transformation of economies by increasing the share of high-productivity sectors (services) (Figure 2.25). Given the caveats discussed above, these results should be seen as indicative, providing a starting point for more detailed country-specific diagnostic of impediments to growth.

Figure 2.26.Potential Growth Benefits from Specific Policy Actions

The thresholds are based on 25th (2.4 percentage points) and 75th (8.9 percentage points) percentiles of the distribution of growth impact of all policy measures across all CESEE countries. The growth impact of policy measures is calculated assuming that CESEE countries will improve their structural characteristics to the average level of EU-15. SEE-non EU countries are not included due to data limitations. The assessment of productivity-improving measures is based on our stochastic frontier and allocative efficiency analysis. Potential gains from improving labor force participation and upgrading infrastructure are based on the production function approach. Note: CEE = Central and Eastern Europe; CESEE = Central, Eastern, and Southeastern Europe; CIS = Commonwealth of Independent States; SEE = Southeastern Europe; SEE-XEU = Southeastern European countries outside the EU.

Active labor market policies could help offset some of the negative effects of demographic changes and emigration. Structural reforms and the improvement of institutions not only help convergence but also encourage potential migrants to stay as well as attract skilled workers from other countries. Further efforts are also needed to reduce skill mismatches, increase labor force participation (women and seniors), promote return migration, engage with the diaspora, better leverage remittances, and better utilize the EU Structural and Cohesion Funds.8

The productivity-enhancing reforms are crucial for re-accelerating convergence. CESEE countries have little prospects of converging to euro area living standards within a generation (25-year horizon), if TFP growth remains the same as in euro area, implying investment rates close to current levels (Figure 2.27). At the same time, a sustained 1 percentage point TFP growth differential against the euro area, which would raise average investment rates by 2½ percentage points a year, would lead to faster convergence for many CESEE countries. Some of the CIS and SEE-non EU countries, however, might need a much larger boost to TFP growth to achieve similar results. The analysis in this report suggests a possible menu of reforms that could help CESEE countries to achieve a necessary boost to TFP growth.

Figure 2.27.Illustrative Convergence Scenarios

(Percent of euro area GDP per capita at PPP)

Sources: Penn World Tables, Version 8.1; and IMF staff calculations.

Note: Income projections are derived using the estimated transition dynamics in Annex 2. In combination with the current total factor productivity (TFP) and K/L ratio, the transition dynamics recursively determine the future path of investment rates, the K/L ratio, and income per capita for all CESEE and the euro area. Population stays unchanged. The first projection assumes no difference in TFP growth between CESEE and the euro area—TFP grows in all countries by 0.5 percent annually. The second projection assumes a positive TFP growth differential of 1 percentage point a year in favor of CESEE countries against the euro area (TFP growth of 0.5 percent). It is consistent with GDP growth of 1.5 percent in the euro area and on average 3.5 percent in the CESEE, approximating Barro’s “iron law of convergence” (Barro 2015). Note that the more per capita income converges, the slower TFP growth is likely to become - this is not taken into account in the calculations presented in this figure.

    Other Resources Citing This Publication