You are here: Home
During the present crisis beginning in 2008, people worry about rising inequality, weaker social protection and the divergence of income levels between the core and the periphery of the European Union (EU). The financial crisis has been blamed on inequality (Rajan 2010; OECD 2015) as poor strata of the population (in the United States, but also in Europe’s periphery) borrowed funds to acquire housing or maintain consumption levels in spite of low and stagnating wages. On the side of lenders, high inequality contributed to an overhang of savings as the rich have a higher propensity to save, and investment in the real economy stagnates in the face of weak demand.
When governments increased their debt to bail out a financial sector where bankers and investors had enjoyed astronomical revenues and incomes, public discontent had increased (Occupy Wall Street). Prominent economists like Piketty (Piketty 2013) and Stiglitz (Stiglitz 2012) pointed out the rising levels of wealth, debt and the related income inequality and warned about their consequences. Even mainstream institutions such as the International Monetary Fund (IMF) or the Organisation for Economic Co-operation and Development (OECD) criticised the negative impact of rising inequality (Kumhof and Rancière 2010; Gupta 2014; Ostry et al. 2014; OECD 2015).
Concerns about declining cohesion within Europe grew as southern European countries faced shrinking economies and rising poverty and unemployment. The following chapter discusses the dimensions of inequality in the EU and analyses their relationship with the crisis.
The Dimensions of European Inequality
Inequality can be considered between different entities (such as countries, regions, households, individuals) with regard to different qualities (such as income, wealth, life expectancy) using different indicators and measures. Here, we focus on (disposable)1 income inequality within the EU. The analysis of inequality in a multi-country context implies certain problems, which have been discussed in depth on a global level by Milanovic (2016) and Bourguignon (2015). They differentiate between three types of international inequality: (1) between nations regardless of their population; (2) between nations weighted by population; (3) between people (households). The last measure takes into account the distribution of income within and between countries.
To compare incomes in an international context, one can use two measures:
(1) at exchange rates and (2) at purchasing power parity (PPP). The use of these two different measures makes a lot of sense when one compares income levels between countries with different currencies, as the value (e.g. converted into Euros) might change with the (real) exchange rate, which depends on variations in the nominal exchange rate and on inflation, which are different from country to country. Prices might change at different rates within countries between different regions, too. The analysis of international inequality and its results depend on the choice of indicator, too. First, there is an almost ethical question: Are absolute differences between incomes more relevant than relative ones? Are poor people content to see their income grow faster than that of the rich or do they want to reduce the absolute difference? In the context of convergence between countries, the first concept is called beta convergence; the second sigma-convergence, as the standard deviation, indicated in mathematics by the Greek letter sigma, measures absolute differences. To offer a (not unrealistic) numerical example: if at the beginning of the comparison the average GDP/capita of the richer country is 5 times the one of the poorer country and the rich country’s economy grows at an annual rate of 2% and the poor at 5%, it would take the poor country 55 years to catch up, and only after 24 years would the absolute difference between the two average incomes begin to shrink (it would still increase for the first 24 years in spite of the higher growth rate). Second, an indicator of international inequality should better be decomposable into intra-and inter-country inequality. This condition is fulfilled by the Theil index2 and the quintile ratio (S80/S20), but not by the Gini index. The Gini index varies between 0 in the case of perfect equality and 1 in the case that all income goes to one entity (e.g. household). If one compares only the degree of poverty rather than the distribution of income as such, the Foster-Greer-Thorbecke (FGT) index is decomposable as well.3 The indicator we use most often is the ratio between the average income of the richest and poorest quintiles (= 20%) of the respective population (the so-called quintile ratio S80/S20).
If one analyses income inequality in a multi-country setting like the EU, different dimensions are of interest.
(A) Disparities between the EU Member States measured in terms of average per capita income; in this case, the inequality within the countries is neglected;
(B) Disparities between regions of the EU; in this case, the inequality within the regions is neglected;
(C) Disparities between households within countries;
(D) Disparities between households within the EU as a whole taking into account both inequalities, (A) and (C).
The reduction of disparities between countries (dimension A) and between regions (dimension B) is usually called ‘convergence’ or ‘cohesion’.
The funds used by the EU to reduce regional inequalities are cohesion funds. There are regular reports by the EU on the development of regional disparities.4 Dimension (C) refers to the well-known inequality within countries. Let us briefly consider the three other dimensions (A, B and C) before focussing on European-wide inequality (D).
Divergence and Convergence
Greater wealth is one reason why poor countries joined the EU. For the EU itself, convergence is an official goal. Historically, for the first two poor countries that became Member States (Ireland in 1972 and Greece in 1981), progress was slow. Portugal and Spain (entry in 1986) experienced good catch-up growth for several years. For the post-communist countries of Central and Eastern Europe (CEE), catching up has been key. The biggest success story so far has been Ireland, which showed spectacular growth in the 1990s (i.e. 20 years after entry), thus becoming the second-richest country in the EU (measured at per capita GDP). Income disparities within the EU are huge. The poorest countries (Bulgaria, Romania and the Baltic states) have per capita incomes below €20,000, while this figure exceeds €70,000 for the richest country (Luxemburg). The differences become greater when one compares incomes at exchange rates, as PPPs reflect lower price levels in poorer countries (in particular rents and services).
Regarding different forms of convergence (beta and sigma, see above) one can see that there has been beta convergence since 1999 as most new Member States in CEE (top of Table 2.1) have grown much faster than the core EU countries. But there was no clear sigma convergence in the EU. Only after 2007, income disparities between countries have declined by approximately 5% or 10% if measured by the standard deviation of their average per capita income at exchange rates or at PPPs, respectively. If one calculates the S80/S20 ratio for the EU as a whole by adding up countries until their total aggregate population reaches a fifth of the EU (about 100 million), the ratio has declined between 2005 and 2014 from around 5.4 to 3.7 at exchange rates and from 2.6 to 2.0 at PPP.6 This ratio neglects income disparities within countrie Since the beginning of the crisis, recovery in Europe has been unequal. The resulting divergence does not appear in the general measures as most of the poorer Member States in CEE returned to their former growth path while the depressed, austerity-struck economies of Ireland, Greece, Spain, Portugal and Cyprus belonged to the European top or middle ‘class’, regarding income per capita. Has there been convergence after all? The answer depends on the measure or metric. If one measures relative disparities (e.g. by using S80/S20), countries did converge. If one considers absolute differences (e.g. by using standard deviation), countries did not converge.
Income (average per capita income) disparities between regions are higher than between the Member States (see above A) because regional income disparities within countries are high and tend to increase. Many economic activities are concentrated in growth centres, often the country’s capital. In Great Britain, for instance, the ratio of average income between London and Wales (the poorest region) is 1:5. In the EU as a whole, the richest region (on the NUTS-2 level8) is the City of London with a per capita income (at PPP) of more than €80,000 compared to €7200 in the Romanian border region Nord-Est.
For the EU as a whole, regional inequality (measured by the standard deviation) has increased (no sigma convergence). Nonetheless, there has been beta convergence as regions in poorer Member States have grown faster thanks to the faster growth of their national economies. If one calculates a European S80/S20 ratio by creating the poorest and richest European quintiles (of 100 million people) by adding up poorest and richest regions neglecting intra-region inequalities, the resulting values are 4 in 2000 and 2.8 in 2011. This decline again reflects the catching-up growth of poorer Member States. Within countries, regional inequality has increased. For the 22 Member States of the EU-28 that are divided into NUTS-2 regions (all except the smaller countries Luxemburg, Malta, Cyprus and the three Baltics), the standard deviation increased on average by 106% between 2000 and 2011, while within the new Member States, the rise was even stronger. Regional inequality in Romania increased by 300%.9
Recently, concerns about national inequality have increased. Even international institutions not known as progressive or concerned about social justice such as the OECD and the IMF have started to publish critical studies of inequality and its consequences for growth and stability. On average in the EU, inequality within countries has hardly increased. The average of the national S80/S20 ratios has remained at about 5 (see the bottom curve in Fig. 2.1 below). But that average hides substantial disparities. In Croatia, Denmark and France, the ratio increased between 2007 and 2013 by more than 15%, in Greece by 10%, while it declined by more than 10% in Romania, the United Kingdom and the Netherlands. National inequality is affected by redistributive policies such as progressive taxation, social protection and transfer payments to old, sick and unemployed people. The resulting distribution of disposable income is more equal than the primary distribution of market income. However, the effect of these policies varies widely among the Member States. The differences between the Gini coefficients for market and disposable income range from 0.14 (e.g. Finland, Slovenia, France) to 0.09 (Spain, Netherlands) and 0.08 (Estonia).