EU income inequality decline: Views from an income shares perspective

In a recent blog post, I showed that the EU-wide Gini coefficient of income inequality hit an almost three decade low in 2016. The large decline in 2016 was driven by both income convergence across countries and reductions in within-country income inequalities. Following a few commentators’ requests for a more granular analysis, I will provide more scrutiny on this situation by looking at income share developments.

Figure 1 shows that the poorest 20% of the European population received 5% of total income in 2007 – a share which has gradually increased to 6% by 2016. This indicates that income inequality is substantial as the proportion of total income is well below the weight of the population segment. However, its developments over the past decade show that the bottom earning population is catching up. The second poorest 20% group - those who belong to the poorest 20% to 40% segment in the income distribution spectrum - have also benefitted from an increase in their income share form 11.7% in 2007 to 12.3% in 2016.

In contrast, the richest 5% of EU citizens - those who held 16.3% of total household income in the EU in 2007 (and even 16.7% in 2008) - saw a slight erosion to their income share to 15.7% in 2016. The second richest 5% segment also faced some, albeit minor, decline in their income share from 2007-2016. Thereby, on average across the EU, the low earners are catching up to European middle class while the high earners are slightly falling back. The decline in the EU-wide Gini coefficient is therefore driven by developments both at top and the bottom of the income distribution spectrum.

I should also add a few remarks on the data and the methodology used to analyse the income distribution situation. Country-specific data is aggregated into the EU’s distribution of income using purchasing power standards which considers price differences across countries (e.g. prices are much higher in Luxembourg than in Bulgaria, so fewer goods can be purchased with one euro in Luxembourg than in Bulgaria).

The income data was collected from the EU-SILC survey. For data on income, underreporting can always be a big hindrance to accurate information. Indeed, it is a reality that, for example, some cleaning ladies, construction workers, lawyers, doctors, etc. do not always declare their total income. Moreover, there are cases of high earners using tax havens to hide and increase their income and wealth. Such underreporting distorts information on the level of income distribution. Unfortunately I cannot accurately assess the importance of this distortion, yet I highlight that it influences my results on the changes in income inequality only if underreporting changes too.

The EU-distribution of income depends both on within-county inequalities and between-country inequalities. Changes to between-country inequalities result from a rise (e.g. Poland) or fall (e.g. Greece) in economic health and performance. For example, the average income in Poland is well below the EU’s average. If the average Polish income increases, then the EU-wide income inequality declines as the income gap between people living in poorer and richer countries decreases.

In order to see within-country inequality developments, the charts in the annex show the same income share data for each EU country. Among the current 27 EU countries for which data is available from 2007 to 2016, the income share of the poorest 20% declined in 17 countries, remained practically unchanged in five, and increased in another five. Thus, in most EU countries the relative income position of the lowest earners has deteriorated overall in 2007-16, though developments in 2015-16 are more favourable for the poor. The share of the top 5% declined in 2007-16 in 14 countries, remained practically unchanged in two, and increased in 11.  As the numbers highlight, the country-specific results for top earners are more mixed.

The impact of convergence/divergence of countries, which influences between-country inequalities, is illustrated in Figure 2. It shows the 2007 and 2016 country-composition of the EU-wide income distribution by income deciles. Income deciles are derived by ranking all people according to their income and subsequently dividing them into ten equal groups. The lowest decile includes the lowest earners, while the 10^th decile is made up out of the highest earners.

In 2007, almost one-third of the poorest 10% of EU citizens were from Romania and almost one-quarter from Poland. Altogether, in 2007, the ten central and eastern European countries accounted for 75% of the poorest 10% of EU population. The remarkable convergence of these countries is visible by comparing the two panels: their share in the lowest EU income decile declined from 75% in 2007 to 54% in 2016. This is still a large share and central Europeans hardly enter the lucky club of the richest 10% of EU citizens, but the direction of change is encouraging.

The progress of central Europeans away from the poorest deciles implies, by definition, that citizens of other countries have entered the bottom deciles. In simplified terms, if some people move up in a ranking, others must move down. The benign way such changes in relative income positions occurs when only upward convergence happens as the absolute income of people in other countries does not fall. In such a case, citizens of other countries slip into the poorest decile not because their living standards have fallen but because the living standards of central Europeans has increased. We can observe such slipping back in most northern and western EU countries.

But southern Europeans have taken an economic hit which has seriously pushed them back on the income distribution ladder. For example, Italy accounted for 4.8% of the poorest EU decile and 8.4% of the second poorest EU decile in 2007; however, by 2016, Italy’s share in these two poorest deciles had increased to 10.8% and 13.5% respectively. In terms of number of people, 6.4 million Italians belonged to the two poorest EU deciles in 2007 with their number increasing to 12.1 million by 2016. Yet the burden has not fallen equally on every shoulder in Italy or, for that matter, in Spain. Italy’s share in the richest EU decile was 11.8% in 2007 (5.9 million people) and has only marginally fallen to 10.8% in 2016 (5.5 million people). Spanish developments are even more striking. Indeed, while there was a large increase of Spaniards belonging to the two lowest EU income deciles, the share of Spain in the richest EU income decile has in fact increased from 2007 to 2016.

In sum, the wealthy Italian and Spanish lost very little (perhaps even increased their income) while lower earning Italians and Spaniards have lost a great deal.

The setback in Greece was more uniform because both the poor and the rich have fallen back. The number of Greeks in the two lowest EU income deciles increased significantly from 2.2 million to 5.7 million from 2007-16. All the while, the number of Greeks belonging to the richest 10% of EU population fell from 660 to 110 thousand.

Therefore, the overall decline in EU-wide income inequality is a welcome development, which partly reflects the progress of the poorest people in Central and Eastern Europe. That being said, the poorer southern Europeans faced a major setback, while richer Italians and Spaniards hardly suffered.

Annex: Country-specific income share developments