Convergence and inequality

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.

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.

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.

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.

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

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.

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.

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).

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.

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.

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.

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.

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.

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.

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.

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.

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.

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.

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