How has income inequality evolved among the citizens of the 28 current EU members?
Lacking correct official statistics, I have estimated the EU-wide Gini coefficient using a method which is very precise, as I explained in a recent paper. I also estimate the Gini coefficient for world citizens as well as for other major regions of the world.
My most striking result shows that income inequality in the 28 countries which are now members of the European Union has been on a unique course in recent decades. Inequality is much lower among EU citizens than in other parts of the world, and actually fell in 1994-2008.
As the figure above indicates, there was a sharp increase in income inequality among the citizens of the EU28 in 1989-93, reaching the level observed in the United States at that time.
A major reason for this increase was the massive output declines in the central and eastern European states during their transition from socialist to market-based economies. As their income declined by about a quarter or more in 1989-93, citizens in these countries became even poorer relative to citizens in western European member states, and EU28-wide income inequality therefore increased.
Nevertheless, the EU-wide Gini coefficient of income inequality in 1993 was 36, well below the 50+ values observed in Africa, Asia and Latin America.
After 1994, inequality declined steadily until 2008, since when it has remained relatively stable. In contrast, income inequality in the United States increased almost continuously from the late 1970s until 2013.
How could EU-wide income inequality decline when income inequality actually increased by more than 1 Gini point in 15 of the 28 countries in 1994-2008? One reason is that there were 7 countries in which it declined by more than 1 Gini point, and there were 6 other countries in which income inequality hardly changed.
But, more importantly, I find that that the decline in EU-wide income inequality in 1994-2008 was almost entirely due to the convergence of average incomes. People in poorer regions of the EU increased their income relative to richer regions.
Unfortunately, this process has stopped with the global and European financial and economic crisis: while central European member states continue to close their gaps with the richest member states of the EU, some southern European member states, like Italy and Greece, are falling behind. Thus overall convergence within the EU was minimal in the past five years.
How should we interpret my new findings?
It is important to remember that high levels of income inequality have various negative consequences (including poor health outcomes, weak social mobility and potential swings towards political populism), as I analysed in a recent report on inclusive growth with Guntram Wolff.
With this in mind, the positive message of my calculations is that income inequality was falling among the citizens of the current 28 EU member states in 1994-2008. Moreover, levels of income inequality in Europe are well below levels observed in the other main regions of the world.
However, the bad news is that the decline stopped in 2008, and the current EU Gini coefficient of 33 is still high in comparison to Norway (23), Japan (30) or Canada (31). Above all there are large differences within the EU. While in Belgium, the Czech Republic, Slovenia and the Scandinavian countries the national Gini is about 25, it is around 35 in Bulgaria, Greece, Spain and the Baltic countries.
Recent work on income inequality has found that many of the negative impacts are related to local levels of inequality (see, for example, my recent work about the impact of income inequality on votes for Brexit and Trump). Therefore, efforts to address income inequalities should be stepped up in a number of countries as well as at the EU level.
Finally, I note that Eurostat publishes a misleading Gini coefficient of income inequality for the EU28, which is simply the population-weighted average of country-specific Gini coefficients. However, the average of the Gini coefficients of individual countries does not correspond to the Gini coefficient of the combined population of those countries, partly because of the differences in average income in different countries, and partly because of differences in within-country income distributions (see more details on this issue in my paper).I therefore recommend that Eurostat stop publishing these misleading Gini coefficients for the EU and instead calculate the EU-wide indicators of income distribution either by combining household level data from all countries, or by using my estimates.