Average incomes in the EBRD region have risen markedly since the late 1990s. Back then, the average level of income in the region was only a quarter of that seen in the G7 advanced economies when measured at purchasing power parity (PPP).1 By 2011, it had reached 38 per cent of that level. Per capita incomes in the region today are around 50 per cent higher, on average, than they were in 1989. These average figures point to a strong performance in terms of income convergence, despite the fact that convergence has virtually stalled since 2011.2 However, they also mask large differences – differences both between economies and, more importantly, between individuals within economies.

Has everyone benefited from post-transition income convergence? Who has benefited the most and who has benefited the least? Whose growth experience are we actually referring to when we quote average figures for income growth and convergence? The answers to these questions vary from country to country. They also determine, to a significant extent, whether or not the broad gains of transition and globalisation are economically and politically sustainable.

In order to provide answers to these questions, this chapter looks at the income growth patterns experienced by different segments of the population (namely, the poor, the middle classes and the better-off), without necessarily passing judgement on the root causes of any trends observed. Income inequality is partly a reflection of differences in people’s efforts and abilities, but it also reflects differences in people’s opportunities to apply their skills or finance their ideas (for instance, if good jobs are reserved for those with connections). These issues are explored in greater detail in subsequent chapters, which look at equality of opportunity and financial inclusion.

When it comes to shifts in income inequality and differing experiences of growth since the late 1980s, the trends observed in countries that have experienced price liberalisation and a transition recession (referred to as “post-communist countries” in the interests of simplicity) often differ fundamentally from those witnessed in southern and eastern Mediterranean (SEMED) countries, as well as Cyprus, Greece and Turkey. This distinction is maintained throughout the analysis in this chapter.

We can see that people’s individual experiences of growth and convergence differ vastly depending on their position on the income ladder. In post-communist countries, only people in the top 27 per cent of the income distribution have experienced average or above-average income growth. In contrast, 23 per cent of people are actually worse off today than they were in 1989, while a further 33 per cent have experienced income growth below the G7 average, implying that only 44 per cent of people in those countries have personally experienced income convergence in the long run.3

Broadly speaking, the region has made considerable progress in terms of reducing poverty, but the extent to which wealth is concentrated among the very rich appears to be particularly high by global standards, partly reflecting the legacy of privatisation programmes implemented during the transition process. Poverty, income inequality and the concentration of wealth among the very rich are virtually uncorrelated across countries. These are separate phenomena that require separate policy responses, which will be discussed in the course of the report.

The chapter is structured as follows. A brief review of global trends in inequality is followed by an examination of income growth patterns across the EBRD region, with a focus on differences between the experiences of the poor, those in the middle of the income distribution and the better-off. The next section examines trends in poverty, income inequality and the concentration of wealth. This is followed by a discussion of broad economic policies that can help to tackle poverty, rising inequality and excessive concentration of wealth among the very rich.

Trends in inequality: a global perspective

The transition process has coincided with a period of technology-enabled globalisation. Over this period, inequality between countries has generally declined as income levels in emerging markets have risen towards those seen in advanced economies, while inequality within countries has increased. As a result of these two conflicting trends, the Gini coefficient measuring income inequality at the level of the world as a whole has been broadly stable over the last 30 years and may have begun to gradually decline.4

A number of factors have contributed to the shifts seen in the global distribution of income. Income convergence between countries has been supported by a long period of relatively high commodity prices which has benefited commodity-exporting developing countries, by improvements in macroeconomic policies in emerging markets and by technology-enabled globalisation of production based on global value chains. At the same time, the automation of routine jobs and new technologies that have increased productivity differentials between people with higher and lower skill levels have resulted in increased inequality within countries.5

At a global level, the main winners as a result of these changes have been the very rich, as well as those in the middle of the global income distribution (that is to say, the middle classes and the better-off in emerging markets and developing countries). This can be seen in Chart 1.1, which shows changes in people’s real incomes based on their position in the global distribution of income (with points A and C corresponding to the main winners). In contrast, the middle classes in developed countries (those around the 80th percentile in the global income distribution – point B in the chart) have seen their pay rising only slowly – if at all – as many of their jobs have been automated or outsourced to emerging markets.6

These shifts have coincided with increases in the concentration of wealth, with the result that the richest 10 per cent of people in advanced economies are now estimated to account for more than half of all wealth in those countries.7 At the very top end – that is to say, the top 0.1 per cent or 0.01 per cent of the global income distribution – the stock of wealth is now significantly more concentrated among top earners than annual income.

Chart 1.1

SOURCE: Lakner and Milanović (2016).

Is growing inequality within countries a concern?

There is growing evidence that excessive inequality hurts long-term growth prospects. Specifically, the concentration of earnings in the top quintile (20 per cent) of the income distribution may hamper subsequent growth. Furthermore, high levels of wealth concentration that are driven by political connections (as opposed to innovation, for instance) are associated with weaker long-term growth performance.8

An unchecked increase in inequality may jeopardise people’s ability to invest in their human capital or develop new ideas. In other words, excessive inequality of outcomes may, over time, negatively affect equality of opportunity in society. High levels of income and wealth inequality may also lead to a loss of confidence in core economic and political institutions and a backlash against market reforms, shifting public consensus towards populist policies that target redistribution and may potentially be detrimental to productivity growth.9

In this chapter, we look at measures of inequality of economic outcomes (as opposed to inequality of opportunity). Outcome-based measures are more readily available and feature much more prominently in policy debates; they also enable us to look more closely at poverty and the concentration of wealth among the very rich – important aspects of the overall distribution of income and wealth that are typically not captured by the survey data used to determine inequality of opportunity.10

For countries in the EBRD region, analysis of income growth on the basis of people’s position in the income distribution requires a number of crude simplifying assumptions in order to fill in data gaps.11 Indeed, many countries in the region do not feature in the analysis of global trends presented in Chart 1.1 owing to such data limitations. This chapter represents an important – albeit imperfect – attempt to fill in those gaps.

Post-transition convergence: differing experiences

In the EBRD region, the globalisation trends of the last three decades have been compounded by the transition to market economies. At the start of the transition process, reported income inequality was low by international standards, although official measures may have understated the level of inequality in economies with shortages of goods. As various goods and services (from cars to summer holidays to basic food staples) were often distributed in the form of employment-related privileges while being unavailable in shops, some people will have had similar incomes, but vastly differing opportunities to spend their money based on their status within society. Unfortunately, it is virtually impossible to correct for such inequality of spending power before the start of the transition process. Cross-country income comparisons (in US dollars at PPP) are likewise complicated by issues relating to currency convertibility and shortages of goods.

Reported inequality rose sharply in the 1990s. During the early years of the transition process, newly created markets placed a premium on new skills – such as business acumen – as well as political connections. This resulted in wage decompression, with upward and downward adjustments in wages for large sections of the population.12 In addition, many women left jobs owing to the loss of universal childcare. In numerous countries, the privatisation of large companies – as opposed to the privatisation of small and medium-sized enterprises (SMEs) – also made a major contribution to the rise in inequality and, in particular, the concentration of wealth among the very rich.13 Moreover, in a number of countries in the region, the armed conflicts and civil wars that followed the collapse of the Soviet Union and Yugoslavia further exacerbated the rise in inequality.14

The transition process has also involved an unprecedented shift from an economic model favouring manufacturing and agriculture to a more service-oriented structure. On average, the share of services in gross domestic product (GDP) has jumped from less than 40 per cent in 1990 to almost 60 per cent today. Moreover, in some economies (such as Bulgaria), almost all of this increase occurred in the space of several years in the early 1990s (see Chart 1.2). Mirroring this, the average share of agriculture in GDP has halved, falling from around a quarter in the early 1990s to around 12 per cent today.

While other emerging market economies, such as China, have undergone similar shifts, the pace of the structural change seen in the early years of the transition process was unparalleled. Urban areas were much better placed to facilitate the development of the new service-oriented economy and benefit from it than rural areas, which exacerbated the rise in inequality and deepened the rural-urban divide in many transition economies.

Chart 1.2

SOURCE: World Development Indicators and authors’ calculations.

NOTE: Data for the EBRD region represent an unweighted average of figures for the following 19 countries: Albania, Armenia, Azerbaijan, Belarus, Cyprus, Egypt, FYR Macedonia, Georgia, Jordan, the Kyrgyz Republic, Moldova, Mongolia, Morocco, Romania, Tajikistan, Tunisia, Turkey, Ukraine and Uzbekistan.

Should inequality have increased?

Some of the increase in inequality seen during the early years of the transition process was inevitable, and perhaps even desirable. Income inequality is partly a reflection of differences in effort on the part of individuals (both in education and at work). To the extent that inequality strengthens incentives for people to excel, compete and invest in education and ideas, it can be necessary for growth.15 The transition process was expected to establish closer links between individuals’ efforts and rewards and thus improve economic efficiency, but the speed of the increase seen in inequality and the resulting distribution of income and wealth are cause for concern.

The transition experience at different points in the income distribution

The combination of the very low levels of inequality at the start of the transition process and the deep transition recession of the early 1990s mean that people’s experiences of income growth have differed widely depending on their position on the income ladder. Chart 1.3 plots cumulative growth in real income per capita since 1990 for each decile of Russia’s population today – the average for the poorest 10 per cent of the population, then the average for the second-poorest decile and so on, all the way up to the richest 10 per cent of the population. The calculation is based on real GDP growth data (adjusted for changes in the size of the population and shifts in the ratio of disposable income to GDP), as well as changes in the income shares of each decile of the population based on the World Bank’s Povcal database. (The earliest data on income shares are typically from 1988-89, while the most recent are from 2012-13.)

The chart reveals that, in Russia, average per capita income growth – the key statistic in a typical analysis of growth and convergence – corresponds to the individual circumstances of someone in the 77th percentile of the income distribution (that is to say, the point where the curve crosses the horizontal line denoting average growth). In other words, only 23 per cent of Russians have actually experienced average or above-average cumulative income growth over the last quarter of a century. Meanwhile, the richest decile have experienced income growth of more than six times the median growth rate. In sharp contrast, 13 per cent of the population (the people to the left of the point where the curve crosses the horizontal axis) have lower real incomes today than they did in 1989. Similarly, in many other countries, large parts of the electorate may feel that they have not benefited from the growth seen in the last two and a half decades. This is understandable, as in many countries average growth rates – the ones typically reported in the press and used by policy-makers – simply do not apply to significant sections of the population.

This analysis is based on a number of simplifications and does not account for people’s mobility between income strata. Indeed, these growth rates are obtained by comparing the incomes of today’s poor with those of the poor of the past (and likewise for the rich). In fact, people under the age of 45 today would typically have had no income of their own before the start of the transition process, while those at the bottom of the income distribution 20 years ago could theoretically be at the top today and vice versa. The results would be unlikely to change significantly even if such movements could be fully accounted for (see Box 1.1, which is based on panel survey data for Russia).16 Nonetheless, the results should be seen as attempting to compare the incomes of particular segments of the distribution, rather than seeking to track the fortunes of individual people, similar to the concept of shared prosperity used by the World Bank (which looks at the income growth of the bottom 40 per cent).

Equivalent data for 26 post-communist countries in the region (that is to say, countries that experienced price liberalisation and a transition recession) reveal a broadly similar picture. Chart 1.4 provides a representative curve for those countries, showing unweighted averages of the growth rates for each individual decile across 26 countries. Thus, each decile may potentially contain people with substantially different levels of income (as the poorest decile in Slovenia may be substantially better off than the poorest decile in the Kyrgyz Republic, for example). A different calculation is used later on to construct an income distribution for all individuals in the region.

This analysis confirms that most of the changes in the income distribution took place in the early years of the transition process. During that period, the income curve sank and acquired a pronounced slope (see Chart 1.4). A steeper slope generally corresponds to a stronger rise in inequality, as it means that the poor experience much weaker growth than the well-off. Since the late 1990s, income inequality in the region has been broadly stable. Correspondingly, the curve representing income growth by decile has shifted upwards, while becoming only marginally steeper, as convergence has benefited all deciles of the income distribution, albeit to varying degrees.17 Note that this synthetic analysis is shown for illustrative purposes only, as the use of unweighted averages hides many individual experiences. (For instance, the incomes of the bottom decile increased on average, but in many individual countries they did not.)

A similar pattern can be observed when looking at an index of real income for various deciles over time (see Chart 1.5). Inequality shot up during the early months and years of the transition process, as the incomes of the top deciles fell less sharply (and rose in some countries), while those of the majority of the population fell dramatically – mainly owing to wage decompression.

Growth and convergence during the 2000s benefited almost everyone, but the median citizen experienced overall growth of around 45 per cent – below the reported mean. Nevertheless, that median growth is still higher than the median rate estimated for the G7 economies. In those advanced economies, too, median income growth since 1989 has lagged behind the mean, with the two estimated at around 17 and 39 per cent respectively.18

Chart 1.3

SOURCE: National authorities, World Bank Povcal database, International Monetary Fund (IMF), UN and authors’ calculations.

Chart 1.4

SOURCE: National authorities, World Bank Povcal database, IMF, UN and authors’ calculations.

NOTE: Data represent unweighted averages for each decile across the following 26 countries: Armenia, Azerbaijan, Belarus, Bosnia and Herzegovina, Bulgaria, Croatia, Estonia, FYR Macedonia, Georgia, Hungary, Kazakhstan, the Kyrgyz Republic, Latvia, Lithuania, Moldova, Montenegro, Poland, Romania, Russia, Serbia, the Slovak Republic, Slovenia, Tajikistan, Turkmenistan, Ukraine and Uzbekistan.

Chart 1.5

SOURCE:  National authorities, World Bank Povcal database, IMF, UN and authors’ calculations.

NOTE: Data represent unweighted averages across the 26 post-communist countries listed in the note below Chart 1.4. The median is calculated as the mean of the fifth and sixth deciles.

Has the region’s transition experience been exceptional?

The experiences of other emerging markets have been similar in some respects, but very different in others. In China and much of emerging Asia, inequality has risen sharply, resulting in large differentials between the growth experienced by the rich and median rates (see Chart 1.6). However, as these countries have enjoyed consistently strong growth and have not experienced a major recession of the kind seen in the EBRD region in the early years of the transition process, even the poorest sections of those populations have enjoyed very strong growth.

Growth in Latin America has been weaker, but appears to have benefited the lower and middle segments of the income distribution to a greater extent (see Chart 1.7, which provides data for Brazil).19 This reflects the very high levels of inequality seen in the late 1980s, which have since improved, partly owing to increases in taxation, greater redistribution through government spending and increases in minimum wages. In advanced economies, most of the gains have been accrued by the top decile, while income growth in the middle and bottom segments of the distribution has been modest, as reflected in diverging mean and median growth rates.

Some individual countries within the EBRD region have also had different experiences. In Turkey, people in the middle section of the income distribution have experienced stronger growth than the poorest and richest in the country, while in Egypt, Jordan and Tunisia, people in the lower-middle segment of the income distribution have seen the strongest rises in incomes.

Chart 1.6

SOURCE: National authorities, World Bank Povcal database, IMF, UN and authors’ calculations.

Chart 1.7

SOURCE: National authorities, World Bank Povcal database, IMF, UN and authors’ calculations.

The convergence pattern for the region as a whole

Let us now consider the EBRD region in its entirety. This means looking at people with a particular level of income (expressed in US dollar terms at PPP) independently of their country of residence. Incomes within individual countries vary widely. For instance, with the exception of Tajikistan, Uzbekistan and the Kyrgyz Republic, every country contains a decile of people with an average income corresponding to the median income for the region as a whole, but in Moldova this is the top decile, whereas in Lithuania it is the second poorest decile (see Chart 1.8).

By combining all of the income deciles for the various countries, we can construct an income distribution for the region as a whole (see Box 1.2). The second richest decile in the region-wide income distribution for the present day comprises people from 22 countries, including: the richest decile in Armenia, Jordan and Montenegro; the second-richest decile in Kazakhstan and Serbia; several upper-middle income deciles in Poland, Romania and Russia; and several middle-income deciles in the Slovak Republic and Slovenia. The analysis presented in Chart 1.9 aggregates the income growth rates of specific deciles in individual countries on the basis of their position in the region-wide income distribution in 1989. (As before, calculations cannot account for the income mobility of individual citizens within countries.) Thus, the shifts depicted in the chart reflect the redistribution of income within individual economies, as well as differences in growth rates between economies (similar to the analysis for the global population presented in Chart 1.1). The pattern that emerges for the EBRD region as a whole is broadly in line with the global pattern, albeit with some differences.

In general, those in the middle of the 1989 income distribution have experienced weaker growth than the very richest and poorest. The well-off in higher-income transition countries have done particularly well. In fact, the income distribution for the EBRD region as a whole has acquired an unusual feature since the mid-1990s – a second mini-mode at high levels of income (in excess of US$ 50,000 at PPP in 2011 prices; see Box 1.2). Increases in inequality within countries and differences in rates of convergence across countries (with poorer countries generally growing faster) mean that the upper and middle classes in lower-income countries have also been among the main beneficiaries of transition and globalisation, while poor residents of lower-income countries have done less well.

The initial transition shock and the subsequent recession affected the entire population of the region to more or less the same extent, while benefiting (in relative terms) the few people who had the technical skills and entrepreneurial vision demanded by a market-based economic system. Low-income countries (which had a less developed industrial base) were also less affected by the transition recession.

The divergence seen in individual rates of income growth has become more pronounced since the 2008 crisis (see Chart 1.9). In recent years, people in the middle of the income distribution have seen particularly sharp slow-downs in income growth (as shown by fact that the income growth curve is now less flat than it was in the 2000s), while the incomes of poorer sections of the population have grown more strongly (partly because lower-income economies have been less affected by the 2008 crisis and partly because a number of countries have adopted post-crisis fiscal stimulus targeting social spending such as pensions).

In Turkey and the SEMED countries, which did not suffer a transition recession following price liberalisation, growth experiences at different points in the income distribution have been more homogeneous, and the middle of the income distribution has, if anything, experienced slightly stronger economic growth (see Chart 1.10). This is one of the factors that have lifted the lower-middle part of the curve for the region as a whole. If we look specifically at the post-communist countries (see Chart 1.11), the pattern is similar to that observed for the region as a whole, apart from the fact that the section of the distribution that has experienced the weakest growth has shifted to the left, with the lower-middle segment being worst affected.

Chart 1.8

SOURCE: National authorities, World Bank Povcal database, IMF, UN and authors’ calculations.

NOTE: Based on estimated income for 2016. Dots show the average income for each decile.

Chart 1.9

SOURCE: World Bank Povcal database, IMF, UN and authors’ calculations.

NOTE: See Milanović (2016) for a discussion regarding methodology. The income distribution on the x-axis is based on 1989 incomes in US dollars at PPP.

Chart 1.10

SOURCE: World Bank Povcal database, IMF, UN and authors’ calculations.

NOTE: See Milanović (2016) for a discussion regarding methodology. The income distribution on the x-axis is based on 1989 incomes in US dollars at PPP.

Chart 1.11

SOURCE: World Bank Povcal database, IMF, UN and authors’ calculations.

NOTE: See Milanović (2016) for a discussion regarding methodology. The income distribution on the x-axis is based on 1989 incomes in US dollars at PPP.

Whose growth?

Individual experiences vary widely within countries, yet they are typically summarised in a single figure: the average rate of income growth. But whose income growth does this average figure correspond to? The answer depends on the specific circumstances in the country in question.

Overall, only 27 per cent of the total population of the post-communist countries have experienced average or above-average income growth (see Chart 1.12). The remaining 73 per cent of the population have experienced income growth that is below the average for these countries. (The percentage experiencing at least average income growth rises to around 40 per cent if Turkey and the SEMED countries are included.) Only in Azerbaijan, Egypt, Jordan, the Slovak Republic, Tunisia and Turkey has median income growth equalled or surpassed the mean.

In most countries, the upper deciles of the income distribution have experienced the strongest growth, in line with the region-wide pattern presented in Chart 1.4. In several cases, however, some of the lower deciles have also experienced above-average growth, similar to the pattern observed for the SEMED region and Turkey in Chart 1.10 (see the data for Ukraine, Kazakhstan, the Kyrgyz Republic, Azerbaijan, Egypt, Jordan, Turkey and Tunisia in Chart 1.12).

Chart 1.12

SOURCE: World Bank Povcal database, IMF, UN and authors’ calculations.

NOTE: Data for each percentile are based on linear extrapolation of averages for each decile.

Has everyone benefited from transition and globalisation?

On the basis of this calculation, 15 per cent of the region’s population are worse off today than they were in 1989 (see Chart 1.13). The corresponding figure for post-communist economies is 23 per cent. A further 21 per cent of the region’s population have experienced long-term income growth that is below the average observed in the G7 economies – in other words, they have not seen their incomes converge with those of advanced economies. In post-communist economies, this figure is even higher, at 33 per cent.

The remaining 64 per cent of the region’s population (44 per cent if Turkey and the SEMED countries are excluded) have personally experienced long-term income convergence. In most countries, those who have experienced income convergence tend to be higher up the income ladder. The most notable exception here is Azerbaijan, where only those with lower incomes are estimated to have experienced growth above the G7 average (with the result that the green segment of the bar is located at the bottom). This probably reflects the redistribution of oil revenues across Azerbaijan’s economy.

Only in around half of all the countries has the majority of the population experienced per capita income growth above the G7 average. In nine countries (Turkey, the SEMED countries, Armenia, Belarus, Poland and Turkmenistan), income growth has exceeded the G7 average in all deciles, resulting in a near-universal (though far from complete) convergence experience.

Chart 1.13

SOURCE: World Bank Povcal database, IMF, UN and authors’ calculations.

NOTE: Data for each percentile are based on linear extrapolation of averages for each decile.

Inequality in the transition region from an international perspective

Today, most transition countries display levels of income inequality similar to those observed in many advanced economies, with significant variation from country to country. For instance, Georgia, Russia and Turkey have higher Gini coefficients, on a par with that of the United States, while measures of inequality are lower in Hungary, the Slovak Republic and Slovenia, on a par with advanced European economies. Inequality in the region tends, on average, to be lower than in other emerging markets, although the gap has narrowed since the start of the transition process (see Chart 1.14).

Chart 1.14
  • Post-communist countries
  • Other EBRD countries
  • Rest of world

SOURCE: World Bank Povcal database, Solt (2014) and authors’ calculations.

NOTE: Based on the latest available data. Calculations underlying Gini coefficients are not fully comparable across countries. Income per capita is measured at PPP. The dotted lines are the lines of best fit for the rest of the world and the sample of post-communist countries.

People’s perceptions of inequality

Most people in the region believe that inequality has increased in their country of residence (see Chart 1.15), despite official data indicating that there has been no clear trend in recent years – and, if anything, a slight decline in inequality since the mid-1990s. The lack of correspondence between perceptions and official data can clearly be seen in the results of the third round of the Life in Transition Survey (LiTS III). That survey, which was conducted by the World Bank and the EBRD in late 2015 and the first half of 2016, spanned more than 51,000 households in EBRD countries, as well as the Czech Republic, Germany and Italy. In all countries bar Tajikistan, the percentage of people who believe that inequality has risen over the last four years exceeds the percentage who believe it has fallen, despite official data indicating that the number of countries where inequality has risen is broadly similar to the number of countries where it has fallen. Furthermore, those perceived changes in inequality are only weakly correlated with official estimates of changes to Gini coefficients.

This disconnect between people’s perceptions of inequality and official estimates may reflect the large concentration of wealth at the top end of the income distribution, which tends to be poorly reflected in household surveys and national statistics. It may also be influenced by the legacy of the transition experience or symptomatic of people’s generally poor track record when it comes to perceptions of income distribution. Regardless of the factors underlying those misperceptions regarding inequality, perceptions matter. In fact, perceptions of inequality tend to matter more than officially reported figures when it comes to social conflict and backlashes against reforms, according to recent studies.20

Chart 1.15
  • Post-communist countries
  • Other EBRD countries
  • Comparator countries

SOURCE: World Bank Povcal database, Solt (2014), LiTS III and authors’ calculations.

NOTE: Changes in Gini coefficients are over four years, based on available data for the period closest to 2011-15.


While trends in income inequality have been mixed, poverty rates in the EBRD region have declined rapidly since the late 1990s as countries have benefited from higher rates of growth and convergence. Poverty headcounts have declined, in terms of both national definitions and the World Bank’s universal definition (namely, people living on less than US$ 3.10 per person per day in 2011 prices at PPP). For instance, on the basis of its national threshold, Russia’s poverty rate has declined from 29 per cent in the aftermath of the 1998 crisis to 11 per cent in 2014.21 Tajikistan, the country in the EBRD region with the lowest income per capita, has also seen rapid improvements, with its poverty rate falling from 86 per cent in 1999 to 23 per cent in 2009 (on the basis of the US$ 3.10 threshold). In comparison, 78 per cent of the population of Bangladesh – an economy with comparable levels of income per capita – are classified as being in poverty under that definition. Today, based on internationally comparable rates of absolute poverty, the region compares favourably with emerging markets and developing countries elsewhere in the world that have similar levels of income per capita (see Chart 1.16).

Chart 1.16
  • Post-communist countries
  • Other EBRD countries
  • Rest of world

SOURCE: World Development Indicators, IMF and authors’ calculations.

NOTE: Based on the latest available data. The poverty threshold is set at US$ 3.10 per person per day at PPP in 2011 prices. The dotted lines are the lines of best fit for the rest of the world and the sample of post-communist countries.

Concentration of wealth among the very rich

The global trend of strong and growing concentration of wealth appears to be even more pronounced in the EBRD region. In particular, the transition process appears, in a number of countries, to have contributed to strong concentration of wealth among the very rich. The latest list of the world’s billionaires published by Forbes magazine includes more than 1,800 individuals from 67 countries, with a combined wealth of around US$ 6.5 trillion. Eleven countries where the EBRD works are featured in that list: Cyprus, Egypt, Georgia, Greece, Kazakhstan, Morocco, Poland, Romania, Russia, Turkey and Ukraine.

With a combined wealth of almost US$ 400 billion, billionaires in the EBRD region account for 7 per cent of all wealth held by the world’s billionaires – slightly more than the region’s 6 per cent share of global GDP. In the case of emerging Asia, Latin America and other emerging and developing countries, the opposite is true (see Chart 1.17).

The EBRD region’s share in global billionaire wealth has increased sharply as a percentage of its share in global GDP, rising from around 50 per cent in 2002 (the earliest year with broad data coverage) to around 115 per cent in 2015. For a number of years, Russia came second worldwide in terms of the total combined wealth of the listed individuals (after the United States), but it has dropped to fourth place in recent years as commodity prices have declined. Cyprus, Georgia and Ukraine also feature among the countries with the highest billionaire wealth-to-GDP ratios in the world.

In terms of the sources of their wealth, billionaires in the EBRD region owe much more to commodities and much less to innovation and competitive manufacturing than those in other regions. In advanced economies, 17 per cent of billionaire wealth is derived from innovation (in sectors such as software and IT hardware), while a further 36 per cent of their money is derived from various manufacturing industries (such as clothing, food and beverages). A similar picture can be seen in emerging Asia. In the EBRD region, however, innovation and manufacturing account for only 1 and 7 per cent of billionaire wealth respectively, as billionaires derive a disproportionate amount of wealth from commodity-related sectors (such as oil, gas and basic metals; see Chart 1.18).22 Moreover, in a number of transition countries, large-scale privatisation made an important contribution to the initial accumulation of wealth by a select group of individuals.23

This distinction may be important, as concentration of wealth caused by rent-seeking behaviour and lobbying is found to be significantly more detrimental to economic growth than concentration caused by innovation in the manufacturing and service sectors.24

The picture that emerges from the Forbes list is undoubtedly incomplete. In particular, as the list only includes people whose wealth exceeds US$ 1 billion, it may tell us less about the concentration of wealth in small countries. Nonetheless, in the absence of tax data for the EBRD region, it provides a useful snapshot of the concentration of wealth and its sources.

Chart 1.17

SOURCE:  Forbes, IMF and authors’ calculations.

NOTE: This chart reports the ratio of (i) the combined wealth of each region’s billionaires as a share of global billionaire wealth to (ii) the relevant region’s share of global GDP. Advanced economies are classified using the IMF’s definition.

Chart 1.18

SOURCE: Forbes and authors’ calculations.

NOTE: Advanced countries are classified using the IMF’s definition. Data relate to 2016.

Distinct policy challenges

Strikingly, measures of the concentration of wealth among the very rich are virtually uncorrelated with overall measures of inequality or poverty in a large sample of countries (see Panel A of Chart 1.19). Nor is there a strong relationship, on average, between inequality and poverty (see Panel B of Chart 1.19). In other words, countries where most people have relatively low incomes may have a number of very rich people (as in the case of India or Georgia); relatively equal societies may have a poverty problem (as in the case of Ethiopia or Tajikistan); and societies with moderate levels of inequality overall may see significant wealth being accumulated by a handful of individuals (as in the case of Sweden or Ukraine).

Chart 1.19
  • EBRD countries
  • Rest of world

SOURCE: World Bank, Forbes, IMF and authors’ calculations.

NOTE: The poverty threshold is set at US$ 3.10 per person per day at PPP in 2011 prices.

Determinants of poverty, inequality and the concentration of wealth

This section looks at determinants of poverty, income inequality and the concentration of wealth for a large sample of developed and developing countries. The regression specification used to examine determinants of the concentration of billionaires’ wealth is a two-stage Heckman selection model. The first stage explains the probability of a country having at least one billionaire (which around half of all economies do) as a function of the size of the economy, its income level and other variables. The second stage explains the levels of non-zero billionaire wealth-to-GDP ratios using a similar set of economic factors (the first two columns in Table 1.1). This two-stage procedure takes account of the fact that billionaires are only found in a subset of countries with certain characteristics – for instance, their population size or income per capita.25 The equation for poverty (fourth column) is estimated using a Tobit model to account for the large number of zero observations in advanced economies. The findings presented should be viewed as indicative and not necessarily causal, as inequality, for instance, may itself influence income per capita or the extent to which a country is open to trade.

The results indicate that levels of income inequality tend to be lower in countries with higher-quality economic institutions (as reflected in average scores for Worldwide Governance Indicators measuring the rule of law, control of corruption, regulatory quality and government effectiveness). A 1-standard-deviation improvement in the quality of economic institutions – the difference between the levels in Albania and Poland – is associated with a reduction of more than 3 points in the Gini coefficient. In contrast, democratic institutions (as measured by the Polity II index) appear to play an important role in limiting excessive accumulation of wealth among the very rich, without having a significant effect on income inequality or poverty.26 Recent armed conflicts increase the probability of having billionaires in a country and the country’s wealth-to-GDP ratio by as much as 10 percentage points.

Higher levels of government spending (excluding military expenditure) are associated with lower levels of inequality, as they increase redistribution relative to pure market outcomes. The size of government does not appear to influence the accumulation of wealth or poverty, which points to the importance of adopting targeted measures (rather than simply increasing the volume of government spending) when it comes to reducing poverty. Poverty generally declines as incomes rise and birth rates fall.

Once these and various other factors have been taken into account, the EBRD region currently has significantly lower levels of inequality and poverty than other countries in the global sample. In the case of the concentration of wealth among the very rich, however, those differences are not statistically significant.27

Overall, this analysis suggests that the concentration of wealth, income inequality and poverty are separate – albeit related – phenomena that may require separate policy responses. Relative to other parts of the world, the EBRD region exhibits high levels of wealth concentration, moderate levels of income inequality and fairly low levels of absolute poverty. However, circumstances vary from country to country: some have particularly high levels of wealth concentration; others exhibit high levels of income inequality across the board; and in certain countries, poverty remains a pressing concern.

Table 1.1
Determinants of poverty, inequality and the concentration of wealth
Dependent variable Selection equation Billionaires’ wealth as a % of GDP Inequality (Gini coefficient) Poverty rate
Population (log) 1.01*** -4.03** -1.11* 1.26
(0.20) (1.68) (0.66) (1.08)
Income per capita (log) 1.16*** 1.80 0.01 -19.18***
(0.39) (3.93) (1.44) (2.21)
Quality of institutions 0.73* -0.53 -3.42** 3.69
(0.44) (4.26) (1.71) (3.21)
Democracy (Polity II index) -0.04 -1.07** 0.18 -0.08
(0.05) (0.46) (0.18) (0.30)
Conflict index 10.72* -1.54 0.01
(6.24) (3.05) (4.85)
Conflict since 1989? (yes/no) 1.22***
Non-military government spending (% of GDP) -0.01 -0.05 -0.18* -0.02
(0.02) (0.22) (0.10) (0.15)
Working age population (% of total) -0.08 0.58 0.25
(0.05) (0.49) (0.20)
Birth rate (children per female)  7.09***
Commodity rent (% of GDP) -0.02 -0.64** -0.04 -0.30
(0.03) (0.32) (0.10) (0.18)
Openness to trade (% of GDP) -0.01 -0.02 0.01
(0.01) (0.02) (0.05)
EBRD dummy 0.69 -4.07 -9.08*** -11.94***
(0.49) (5.16) (2.06) (3.41)
Constant -14.37*** 6.31 42.42*** 159.00***
(3.18) (40.67) (12.51) (25.45)
No. of observations 129 129 109 113
R2 0.36 0.27

SOURCE: World Bank, IMF, Forbes, Polity and authors’ calculations.

NOTE: Estimates for the ratio of billionaires’ wealth to GDP are calculated using a Heckman selection model. Poverty rates are estimated on the basis of a Tobit model. Standard errors are reported in parentheses, and *, ** and *** denote values that are statistically significant at the 10, 5 and 1 per cent levels respectively. The conflict index takes a value of 1 if a country is in conflict and declines to 0 after 30 years without a conflict.

Policy responses

Avoiding excessive concentration of wealth

Excessive concentration of wealth warrants the attention of policy-makers, as it may negatively affect equality of opportunity and cause a backlash against key economic and political institutions underlying market economies, which may in turn lead to weaker growth in the long run. A number of polices can be pursued to limit the concentration of wealth among the very rich.

Taxing wealth (a significant percentage of which tends to be held in immobile assets) may be an effective method of fiscal redistribution, as well as a means of raising additional revenue. Taxes on inheritance, in particular, tend to be less distortionary, in the sense that they affect people’s level of effort or employment decisions to a lesser extent. They may, in practice, be difficult to collect in the absence of sufficient international cooperation, as some wealth that is subject to inheritance tax may be moved to tax havens. The taxation of financial wealth faces similar challenges. Recurrent taxes on wealth, particularly immovable wealth, can also be a relatively non-distortionary instrument. In order to be effective, these require a comprehensive and regularly updated register of land and property values. Taxes on property transactions (such as stamp duty in the United Kingdom) are easier to implement, but may prevent economically efficient transactions from taking place.

Historically, taxes on wealth (including property taxes and taxes on inheritance) used to generate significant amounts of public income, but their role has gradually diminished. These days, they generate average annual revenues totalling only 2 per cent of GDP in member countries of the Organisation for Economic Co-operation and Development (OECD).28

In the EBRD region, property taxes and other taxes on wealth raise an average of only 0.8 per cent of GDP (see Chart 1.20). Levels of taxation in the EBRD region are generally somewhat lower than in OECD countries. Nonetheless, average wealth tax receipts in the EBRD region are about 2 percentage points below the OECD average when expressed as a percentage of total tax revenue.

Strong concentration of wealth also calls for further improvements in the overall quality of economic and political institutions that limit the ability of the elite to appropriate economic rents. This can be particularly important in countries where rents – from natural resources, tourism or agricultural commodities – account for a large proportion of total value added. In this regard, efforts to diversify economies away from excessive dependence on natural resources may also help to prevent excessive concentration of wealth in the hands of economic elites. Initiatives promoting greater accountability in commodity-related industries (such as the Extractive Industries Transparency Initiative, which aims to make company and project-level data relating to natural resource management widely available) can also play an important role in this regard. Furthermore, future privatisation initiatives need to learn lessons from the early years of the transition process and employ transparent and fully competitive procedures.

It is also essential to pursue higher standards of governance and transparency, as well as strengthening competition and ensuring consistent enforcement of competition laws. While there is no evidence that having a higher percentage of small businesses in the economy is associated with lower levels of inequality,29 a lack of SMEs may be symptomatic of a poor business environment that holds back entrepreneurs and results in the concentration of wealth.

Chart 1.20

SOURCE: OECD, IMF, national authorities and authors’ calculations.

NOTE: Based on data for 2015 or the latest available year.

Addressing inequality

Tackling broader inequality requires a combination of redistribution through taxation and public spending and measures to reduce inequality of opportunity in society.

The provision of high-quality education, health care and social services can play an important role in reducing inequality of outcomes, as well as tackling inequality of opportunity (as discussed in Chapter 3). If one attaches a notional monetary value to almost-free education and health care services (the result of “predistribution policies”30), the distribution of income becomes less unequal. In fact, public services account, on average, for more than two-thirds of the reduction seen in inequality relative to pure market outcomes (as reflected in income inequality before taxation). Civil society also has a role to play in improving equality of opportunity (see Box 1.3).

Direct fiscal measures aimed at income redistribution (such as progressive income taxes and cash transfers) account for around a third of the reduction in income inequality.31 In contrast, fiscal spending focused mainly on infrastructure, public administration and defence may exacerbate – rather than mitigate – income inequality.

Consumption-based taxes may also be regressive, as the tax paid by the poor may account for a larger percentage of their income. In addition, the recent influx of refugees arriving in countries such as Turkey, newer EU member states and SEMED countries highlights the difficulty of delivering equality and economic inclusion for migrants – and refugees in particular (see Boxes 1.4 and 1.5).

Chapters 3 and 4 of this report discuss policies aimed at facilitating access to financial services and improving equality of opportunity for all people, regardless of their gender, social background and ethnicity and any other characteristics that are beyond their control.

Reducing poverty

Subsidies and cash transfers are often used to improve the lives of the poor. Untargeted subsidies (such as energy subsidies) can be blunt and costly ways of reaching out to those with the lowest incomes. The fact that energy accounts for a much higher percentage of consumption among the poor relative to the rich is often used to justify energy subsidies. However, such subsidies are a highly inefficient and costly way of helping the poor. The majority of the subsidies, in volume terms, accrue to the rich, who use much more petrol and air conditioning. Moreover, low energy prices may further discourage more affluent consumers from saving energy. As a result, energy subsidies undermine governments’ finances and reduce authorities’ ability to finance other spending programmes aimed at helping the poor.

A switch to targeted subsidies may therefore be highly beneficial in terms of reducing poverty. At the same time, the provision of targeted subsidies and means-tested benefits requires a high degree of administrative capacity, which may often be lacking in less developed economies and countries with weaker economic institutions.

The effectiveness of social transfer programmes may, in certain cases, be further enhanced if they include the monitoring of outcomes and address certain behavioural traits that exacerbate inequality of opportunity. For instance, parents receiving assistance could be obliged to send their children to school (thus forgoing the additional income that working children can provide, in return for improvements in human capital and higher expected incomes in the future). Such programmes have been successful in Brazil, Colombia and Mexico.32 Similarly, the long-term unemployed (that is to say, those who have been out of work for more than 12 months) could be required to sign up for job search programmes or undergo retraining.


The EBRD region as a whole has achieved a remarkable degree of income convergence over the last 20 years. However, the benefits of this convergence have been distributed unequally, with fairly complex patterns in evidence.

In fact, people’s experiences of growth and convergence have differed vastly depending on their position on the income ladder. In post-communist countries, average income growth corresponds to the experience of someone in the top 27 per cent of the income distribution, while 23 per cent of those countries’ populations are, on average, worse off today than they were in 1989. Only 44 per cent of those populations have experienced income growth in excess of the average for the G7.

In addition to broader globalisation trends, income patterns over the last two-and-a-half decades also reflect experiences unique to the region – namely, wage decompression and the deep recession seen in the early years of the transition process, as well as a very rapid shift from manufacturing and agriculture-based economies to a more service-oriented model.

Before the start of the transition process, levels of inequality in the region were very low by international standards (at least as far as measurable inequality is concerned). Although they then increased dramatically in the early years of the transition process, they remain moderate by comparison with other parts of the world. Furthermore, significant progress has been made in terms of reducing poverty.

Despite this, people are overwhelmingly of the view that levels of inequality are high and rising. This may, in part, reflect the legacy of the transition experience. These perceptions may, to some extent, also be a result of the strong concentration of wealth among the very rich, with the region displaying high levels of concentration even relative to other emerging market economies. The chapters that follow look in more detail at the ways in which the transition process has affected people’s well-being and people’s perception of that process, as well as their attitudes towards open markets, democracy and reform.

Poverty, inequality and excessive concentration of wealth among the very rich represent distinct challenges requiring separate policy responses. The fact that wealth is strongly concentrated among the very rich across the region’s economies is a source of concern, as it may limit equality of opportunity and undermine confidence in key economic and political institutions, resulting in weaker long-term growth. This highlights the need for further improvements in the overall quality of institutions, higher standards of governance and transparency, the consistent enforcement of competition laws and efforts to diversify economies away from excessive dependence on natural resource rents. Taxation of wealth (as opposed to the taxation of income or consumption) could also be given a more prominent role as a source of government revenues.

The reduction of poverty requires targeted, well-designed social transfer programmes. These programmes may also need to address certain behavioural traits that exacerbate inequality of opportunity – for instance, by forcing parents receiving assistance to send their children to school.

Tackling broader inequality requires a combination of redistribution through taxation and public spending and measures to reduce inequality of opportunity in society. Policies aimed at boosting equality of opportunity for all people (regardless of their gender, social background and ethnicity and any other characteristics unrelated to their abilities and efforts) include better access to higher-quality education and measures supporting financial development. These are discussed in chapters 3 and 4 of the report.

Box 1.1. Income mobility and income growth

Estimates of the income growth experienced by people with differing levels of income assume that there is no mobility between income deciles – that is to say, those who were at the bottom of the income distribution two decades ago are assumed to remain among the poorest today. This assumption is driven purely by the availability of data. The income shares of the various deciles are derived from household surveys, and survey respondents differ from year to year.

How significant has income mobility been since the start of the transition process? And does it alter the main conclusions regarding the very large differences between the growth experiences of individuals within a given country? Data from the Russian Longitudinal Monitoring Survey (RLMS) conducted by the Higher School of Economics in Moscow may help to shed some light on this issue. That survey monitored a representative sample of more than 8,000 people from 1994 to 2014 and, despite some people dropping out of the survey, contains useable income data on more than 1,600 people covering the entire period. Thus, it is possible to calculate the probability of someone who was in a given decile in 1994 being in a particular decile in 2014 (for instance, the probability of someone remaining in the poorest decile or rising all the way to the top decile). These probabilities can then be applied to national survey data in order to construct “adjusted” income growth curves. In addition, each respondent’s income standing can be assessed within their age cohort, in order to take account of the fact that people’s incomes tend to grow, on average, as they move from their 20s to their 40s.

The survey reveals a substantial degree of income mobility, comparable with – and perhaps even higher than – estimates obtained for the United States.33 For instance, people who were in the bottom tercile of the income distribution in 1994 had a 20 per cent chance of being in the top tercile by 2014 (see Chart 1.1.1). Even so, people in all parts of the income distribution were still most likely to remain in the same tercile (see the large rectangles on the diagonal in the chart).

Estimates of income mobility can be used to carry out a “forward-looking” adjustment of the income growth curve for each decile of the population (see Chart 1.1.2). This analysis takes individuals from each income level in 1989 and calculates their income in 2016, resulting in a very egalitarian growth curve relative to the unadjusted “zero-mobility” curve (which is identical to the one in Chart 1.3).34 This is partly because the distribution of lifelong income will typically be more equal than a snapshot of any given year, as the rich have nowhere to go but down the income ladder and the poor can only move upwards. At the same time, the results for Russia also highlight the fact that incomes during and after the transition process may have had little in common with incomes under central planning. It is not clear to what extent today’s generation can expect a similar degree of income mobility in the future.

We can also adopt a “backward-looking” approach, taking the rich and poor of today and asking where they have come from in terms of their standing back in 1989. This results in differences between individual growth experiences which are even larger than those observed under a zero-mobility assumption. For instance, the top decile experience cumulative income growth of more than 400 per cent, while the bottom 25 per cent see no growth at all.

Ultimately, the question of whether income mobility affects conclusions about differing income growth experiences depends on how mobility is corrected for and whether we are interested in the future prospects of today’s workers or their past experiences. Besides, the circumstances in Russia, for which longitudinal survey data are available, may be different from those in other countries in the region. With this in mind, the “zero-mobility” assumption appears reasonable – albeit by no means an accurate reflection of the situation – when looking at disparities between individual income growth rates.

Chart 1.1.1

SOURCE: RLMS and authors’ calculations.

NOTE: Each bar shows the probability of being in the top, middle or bottom tercile of the income distribution in 2014 on the basis of an individual’s position in the income distribution in 1994 (taking into account that person’s age). Darker colours correspond to lower levels of mobility.

Chart 1.1.2

SOURCE: World Bank Povcal database, IMF, World Development Indicators, RLMS and authors’ calculations.

NOTE: The “forward-looking” curve traces the incomes of each decile of the 1989 distribution over time, using inter-decile mobility assumptions based on the RLMS data for the period 1994-2014. The “backward-looking” curve traces the incomes of each decile of the 2016 distribution using those same assumptions.

Box 1.2. Distribution of income for the EBRD region as a whole

What would an income distribution for the EBRD region as a whole look like? Chart 1.2.1 provides the answer, presenting region-wide distribution curves based on 2011 US dollars at PPP.

By 1996, in the wake of the transition process, the income distribution for the EBRD region had shifted strongly to the left, with the median income declining by 26 per cent. At the same time, a large number of high earners emerged at the right-hand end of the distribution. That second mini-mode represents around 2.5 per cent of the region’s population, whose incomes increased and surpassed the median income in the G7 economies.

That second mode then remained in place as the entire curve gradually shifted to the right of the 1989 distribution (see the 2016 curve). While the incomes of both the better-off and the poor have shifted to the right, as has the median income, the distribution has become more unequal, as convergence between the income levels of poorer and richer countries has not been strong enough to offset increases in inequality within individual countries.

Indeed, inequality within countries (as opposed to income differences between countries) now accounts for two-thirds of the region’s total income inequality, up from 57 per cent in 1989.35 In fact, the incomes of the top quintile are now 19 times those of the bottom quintile, up from 13 in 1996 and around 7 in 1989. This contrasts with the global distribution of income, which has, if anything, become slightly less unequal, as poor countries’ convergence with the income levels of advanced economies has more than offset rising inequality within individual economies.

The 1989 and 2016 distributions intersect at around US$ 11,000 in 2011 prices at PPP. At higher levels of income (which are earned by around a third of the region’s population today, compared with around 15 per cent in 1989), the share of the population is larger in 2016 for every level of income.

The income distribution for the subregion comprising Turkey and the SEMED countries has evolved in a different manner. As that subregion did not experience the price liberalisation and transition recession of the early 1990s, its income distribution has shifted further and further to the right over time (see Chart 1.2.2). It has also become less skewed as people at the left-hand end of the distribution have experienced stronger income growth. In contrast, the income distribution for the post-communist countries is broadly similar to that of the region as a whole.

Chart 1.2.1

SOURCE: World Bank Povcal database, World Development Indicators, UN, national authorities and authors’ calculations.

Chart 1.2.2

SOURCE: World Bank Povcal database, World Development Indicators, UN, national authorities and authors’ calculations.

Box 1.3. Civil society and inclusion

Civil society – organisations and groups of individuals that occupy the space between the state and the private sector, promoting a wide range of interests and values through voluntary collective actions – are an essential component of inclusive political and economic systems. Factors that are conducive to civil society include a supportive legal framework and access to justice, access to diverse sources of information, and respect for civil rights and political freedoms. In the days of communism, civil society was largely silenced. Several countries have made extraordinary progress since the start of the transition process in terms of supporting civil society, while others still have a long way to go when it comes to putting the necessary framework in place. In some cases, though, there has also been some reversal of progress.

Many of the activities organised by this sector in developing and transition countries are about inclusion – promoting equal rights, equal opportunities and access to services. Indeed, many civil society organisations and groups are set up specifically in order to address inclusion-related issues, such as labour rights, gender equality, minority rights, rural development, the economic empowerment of young people and the needs of an ageing population. Civil society organisations help to voice the concerns of those segments of the population that are excluded from educational or social opportunities. Civil society stakeholders often promote work-based learning initiatives, social entrepreneurship, and inclusive and sustainable resourcing and management practices at municipal level. Since 2013, the EBRD has, through its Civil Society Capacity Building Framework, provided support for initiatives of this kind targeting economic opportunities for young people and rural communities.

Civil society also plays a crucial role in providing products and services to disadvantaged groups – empowering them, for example, with skills and knowledge. In particular, local civil society organisations have a direct link to local communities and can achieve a significant impact on the ground by serving as a bridge between citizens, local authorities and businesses. Such organisations are therefore well placed to reach out to economically and socially excluded groups, seeking to understand their needs and helping to find local solutions.

In Ukraine, for instance, civil society organisations promote rural youth employment opportunities in sustainable dairy farming. Since 2015, the EBRD has been working with the Danone Ecosystem Fund, Danone’s corporate social responsibility arm, and ICF Community Wellbeing, the Ukrainian chapter of civil society organisation Heifer International, with a view to upgrading the training offered by a demonstration farm in Dnipropetrovsk and facilitating youth employment opportunities in agribusiness through work-based learning. This initiative also supports the use of mobile units to train more than 450 small farmers across Ukraine, with a view to helping to raise quality standards in milk production and enable small farmers to work with large buyers of milk. The programme also involves twinning arrangements with farms in the Caucasus.

Given the inclusion work that is carried out by such organisations, the shrinking of civil society in some countries in the region is a cause for concern. There have been increased reports of civil society actors being intimidated and threatened by state and non-state actors with a view to delegitimising them and isolating them from their communities.36 These developments do not bode well for the future inclusiveness of political and economic systems.

Box 1.4. Economic inclusion of refugees

More than 60 million people around the world are currently displaced by conflict and instability, the highest level ever recorded. The current proliferation of security and environmental threats suggests that such displacement of large numbers of people is set to become the norm, and the EBRD region is directly affected by this trend.

After five years of brutal civil war, Syrians now represent the largest refugee population fleeing a single conflict in a generation (see Chart 1.4.1). In February 2016, the Office of the United Nations High Commissioner for Refugees (UNHCR) estimated that more than 4.7 million Syrians had fled to Syria’s immediate neighbours. Given the extent of Syria’s economic devastation,37 the majority of those refugees will not be able to return home for many years. With 90 per cent of refugees in Jordan, Lebanon and Turkey living outside camps, the economic and social integration of refugees into their host communities represents a major challenge for those host countries.

The influx of refugees has further increased inequality in the host countries. Nearly nine out of ten registered Syrian refugees living in Jordan and Turkey are either living in poverty or expected to be in the near future. Around half of those Syrian refugees are children or adolescents, and the majority are women (with the percentage of women particularly high among refugees in Turkey).

In Turkey, only around a quarter of Syrian children outside camps are in formal education, although this figure may increase with EU funding. Despite an estimated 195,000 Syrians aged between 18 and 25 living in Jordan and Lebanon alone, the numbers enrolled in tertiary education programmes are negligible. Low levels of education among this new generation of young Syrians may impede their social and economic inclusion, trapping them in poverty. With this in mind, governments and international donors are taking steps to promote the verification/mapping of skills and provide education and language training for refugees.

Thus far, the large influx of refugees has had only a limited impact on formal labour markets.38 However, some displacement has been noted with regard to low-skilled workers – mostly in the informal sector, where many refugees have ended up being employed. In some cases, they have displaced other migrant workers.39

That increase in informal employment has helped to reduce the prices of some manufactured goods and provided basic incomes for some refugees. In the longer term, however, reliance on informal labour is likely to weaken the competitiveness of the private sector by discouraging innovation and limiting productivity growth. Recognising this, governments are taking steps to regularise the employment of refugees through work permits.40

The key medium-term priority identified by the governments of the host countries is to ensure that refugees are able to make a living for themselves on a sustainable basis, while preserving the social cohesion of their host communities, which often suffer from pre-existing economic constraints (such as high levels of poverty, scarce resources and poor infrastructure). The key challenge for those host governments is to upgrade their strained municipal infrastructure in a timely and cost-efficient manner.

The private sector can play an important role in creating economic opportunities for communities hosting refugees (in the form, for instance, of work-based learning schemes and access to finance for SMEs established by refugees). Indeed, entrepreneurship has become the key factor contributing to the economic inclusion of Syrian refugees, particularly for those living in larger cities in Turkey and Jordan.

According to the latest figures from the Union of Chambers and Commodity Exchanges of Turkey, the number of Syrian-partnered firms established annually in Turkey increased from 30 in 2010 to 1,599 in 2015, with most of them concentrated in the catering, construction, retail, real estate, transport, textile and food industries.

Chart 1.4.1
  • Syrian refugees as a percentage of host country's population

SOURCE: UNHCR and Jordan Times.

NOTE: Data relate to February 2016 or more recent updates. Actual numbers of refugees (including people who have not been registered) may be substantially higher.

Box 1.5. The refugee crisis and economic migration from the Western Balkans

The Syrian refugee crisis has resulted in renewed attention being paid to migrant flows from the Western Balkans to western Europe. Migration from the Western Balkans has been a consistent phenomenon since at least the 1990s – and several decades earlier in the case of the former Yugoslavia. The migration pattern that has emerged in this most recent episode bears many of the traditional hallmarks of economic migration, but also features some surprising aspects.

In March 2015, the number of asylum applications in Germany – the main destination in western Europe for refugees and migrants alike – was triple what it had been just four months earlier (see Chart 1.5.1). The fact that so many Kosovans and Albanians decided to migrate at that precise moment in time appears to be due, in part, to the sharp decline seen in the cost of migration on the back of the wave of Syrian refugees.

However, Kosovans’ consistently strong preference for migration whenever an opportunity presents itself seems to stem largely from conditions at home (the sluggish economy, the slow pace of institutional development and political change, and so on). While the unemployment rate in Kosovo is high at 35 per cent, employment status appears to play a limited role in the decision to emigrate, with many people leaving local private-sector jobs in order to move abroad. Moreover, that strong willingness to migrate spans all sections of the Kosovan population, running counter to the perception that most migrants are young, male and unemployed.41

A lack of adequate health care and reliable electricity and water supplies are frequently cited as key factors affecting people’s quality of life. Indeed, the LiTS II and III household surveys reveal a strong correlation between dissatisfaction with utilities and plans to move abroad in the next year. As political fatigue and disillusionment with a country’s progress grow, conditions that were once deemed acceptable may start to seem intolerable.

Developing growth-enhancing policies that reduce people’s desire to emigrate is a challenge. Fostering the development of the private sector and economic vitality can help to reduce unemployment and create higher-quality jobs. In addition, policy-makers should not overlook the provision of public services and infrastructure when seeking to encourage skilled individuals to stay. However, these represent medium-term objectives. In the short term, large numbers of people will probably continue to want to emigrate.

The management of economic migration requires proactive policy responses by the governments of both home and host countries. Circular migration policies – such as fixed-term work permits (potentially targeting specific skills), reductions in the cost of transferring remittances through formal financial institutions and tax regimes facilitating the reinvestment of migrants’ earnings in home countries – can help to mobilise short-term migrant and diaspora resources in support of economic development in home countries,42 as well as addressing specific labour market imbalances in host countries. Alternative ways of handling migration, such as the use of fences, deportations and strict penalties for employers, may reduce the number of migrants, but they can also have a negative impact on migrants’ skill profiles and incentivise smuggling.43

Chart 1.5.1

SOURCE: German Federal Office for Migration and Refugees.


P. Aghion, U. Akcigit, A. Bergeaud, R. Blundell and D. Hémous (2015)
“Innovation and Top Income Inequality”, NBER Working Paper No. 21247.

A. Alesina and D. Rodrik (1994)
“Distributive Politics and Economic Growth”, Quarterly Journal of Economics, Vol. 109,  pp. 465-490.

S. Anand and P. Segal (2008)
“What Do We Know About Global Income Inequality?”, Journal of Economic Literature,  Vol. 46, pp. 57-94.

S. Bagchi and J. Svejnar (2015)
“Does wealth inequality matter for growth? The effect of billionaire wealth, income distribution, and poverty”, Journal of Comparative Economics, Vol. 43, pp. 505-530.

R. Barro (2000)
“Inequality and growth in a panel of countries”, Journal of Economic Growth, Vol. 5, pp. 5-32.

F. Bastagli, J. Hagen-Zanker, L. Harman,  V. Barca, G. Sturge, T. Schmidt and  L. Pellerano (2016)
“Cash transfers: what does the evidence say? A rigorous review of impacts and the role of design and implementation features”, Overseas Development Institute report.

T. Beck, A. Demirgüç-Kunt and  R. Levine (2005)
“SMEs, Growth, and Poverty”, Journal of Economic Growth, Vol. 10, pp. 199-229.

A. Berg, J. Ostry and J. Zettelmeyer (2008)
“What Makes Growth Sustained?”, IMF Working Paper No. 08/59.

C. Bircan, T. Brück and M. Vothknecht (2016)
“Violent conflict and inequality”, Oxford Development Studies, forthcoming.

A. Chong and M. Gradstein (2007)
“Inequality and Institutions”, Review of Economics and Statistics, Vol. 89, pp. 454-465.

CIVICUS (2016)
State of Civil Society Report 2016.

B. Clements, R. de Mooij, S. Gupta  and M. Keen, eds. (2015)
Inequality and fiscal policy, International Monetary Fund.

T. Cowen (2013)
Average Is Over: Powering America beyond the Age of the Great Stagnation, Penguin, New York.

E. Dabla-Norris, K. Kochhar, N. Suphaphiphat, F. Ricka and E. Tsounta (2015)
“Causes and Consequences of Income Inequality: A Global Perspective”, IMF Staff Discussion Note No. 15/13.

D. Dollar, T. Kleineberg and A. Kraay (2015)
“Growth, Inequality, and Social Welfare: Cross-Country Evidence”, Economic Policy, Vol. 30, pp. 335-377.

EBRD (2013)
Transition Report 2013: Stuck in Transition?, London.

European Commission (2008)
“A Common Immigration Policy for Europe: Principles, Actions and Tools”, communication to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions.

A. Fakih and M. Ibrahim (2016)
“The impact of Syrian refugees on the labor market in neighboring countries: empirical evidence from Jordan”, Defence and Peace Economics, Vol. 27, pp. 64-86.

C. Freund and S. Oliver (2015)
Rich People Poor Countries: The Rise of Emerging-Market Tycoons and their Mega Firms, Peterson Institute for International Economics, Washington, D.C.

G. Friebel and S. Guriev (2006)
“Smuggling Humans: A Theory of Debt-Financed Migration”, Journal of the European Economic Association, Vol. 4, pp. 1085-1111.

V. Gimpelson and D. Treisman (2015)
“Misperceiving Inequality”, NBER Working Paper No. 21174.

J. Gobat and K. Kostial (2016)
“Syria’s conflict economy”, IMF Working Paper No. 16/123.

Group for Legal and Political Studies (2015)
“Potential migrant’s profile: Who are the Kosovars most willing to migrate?”, mimeo.

S. Guriev and A. Rachinsky (2006)
“The Evolution of Personal Wealth in the Former Soviet Union and Central and Eastern Europe”, UNU-WIDER Working Paper No. 2006/120.

J. Hacker (2011)
“The institutional foundations of middle-class democracy”, Policy Network, 6 May.

J. Heckman (1979)
“Sample Selection Bias as a Specification Error”, Econometrica, Vol. 47, pp. 153-161.

S. Jenkins (2015)
“World income inequality databases: an assessment of WIID and SWIID”, Journal of Economic Inequality, Vol. 13, pp. 629-671.

C. Jones and P. Klenow (2016)
“Beyond GDP? Welfare across Countries and Time”, American Economic Review, Vol. 106, pp. 2426-2457.

P. Krugman (2014)
“The Rich, the Right, and the Facts: Deconstructing the Income Distribution Debate”, The American Prospect.

C. Lakner and B. Milanović (2016)
“Global Income Distribution: From the Fall of the Berlin Wall to the Great Recession”, World Bank Economic Review, Vol. 30(2), pp. 203-232.

A. Lukiyanova and A. Oshchepkov (2012)
“Income Mobility in Russia (2000-2005)”, Economic Systems, Vol. 36, pp. 46-64.

B. Milanović (2016)
Global Inequality: A New Approach for the Age of Globalization, Harvard University Press, Cambridge, MA.

B. Milanović and L. Ersado (2010)
“Reform and inequality during the transition: An analysis using panel household survey data, 1990-2005”, UNU-WIDER Working Paper No. 2010/62.

B. Nolan, M. Roser and S. Thewissen (2016)
“Models, regimes, and the evolution of middle incomes in OECD countries”, INET Working Paper No. 2016-01.

T. Piketty (2013)
Capital in the Twenty-First Century, Harvard University Press, Cambridge, MA.

A. Shleifer and D. Treisman (2005)
“A Normal Country: Russia after Communism”, Journal of Economic Perspectives,  Vol. 19, pp. 151-174.

F. Solt (2014)
Standardized World Income Inequality Database, version 5.0. (

K. Sonin (2003)
“Why the Rich May Favor Poor Protection of Property Rights”, Journal of Comparative Economics, Vol. 31, pp. 715-731.

D. Treisman (2012)
“Inequality: the Russian experience”, Current History, Vol. 111, pp. 264-269.

E. Tsounta and A. Osueke (2014)
“What is Behind Latin America’s Declining Income Inequality?”, IMF Working Paper No. 14/124.

United Nations High Commissioner for Human Rights (2016)
“Practical recommendations for the creation and maintenance of a safe and enabling environment for civil society, based on good practices and lessons learned”.