Twenty years of European east-west household income convergence

While average household income is lower than GDP per capita, the poor have especially benefitted from income gains in central and eastern EU members

Publishing date
06 June 2024
Zsolt Darvas

The European Union’s eastern enlargement has been an enormous economic success over the past two decades. While the convergence of the eastern EU countries with their western neighbours in GDP per-capita terms has been studied widely, household disposable income has received less attention. GDP includes the incomes of workers, corporate and self-employed profits, and the balance between production and import taxes and production subsidies. While GDP is the most widely used indicator of economic performance, household income is more relevant for individuals, since it indicates the means they have to live on. Looking at household incomes allows an analysis of how different segments of society have benefitted from the past twenty years of income convergence.

Average income levels

The countries that joined the EU two decades ago embarked on a spectacular convergence process towards the income levels in rich Western European nations (Figure 1). Research suggests that Western European countries also benefitted from the expanded union 1 See, for example, European Commission (2009), Baas and Brücker (2010), Oberhofer and Winner (2017), Caliendo et al (2021). .

Figure 1 indicates that in terms of household income, the gap between the twelve EU newcomers (CEE12; see list in note to Figure 1) and ten richer western and northern EU countries (EU10; see note to Figure 1) was and still is wider than the GDP per capita gap. I report these indicators at purchasing power standards, which corrects for cross-country differences in price levels. I compare CEE12 to EU10 and to the EU average, because the EU10 countries are the top performers to which convergence would be desirable. The EU average also includes southern European countries that faced various crises and economic weaknesses over the past fifteen years, and the CEE12 countries themselves. 

In 2004, CEE12 GDP per capita was 43 percent of the EU10 level, but average household income was at only 34 percent. In 2023, the CEE12 reached 69 percent of EU10 GDP per capita and 62 percent of EU household income, which was an impressive pace in narrowing the income gap. Household income converged somewhat faster than GDP.

Household income distribution

How have the incomes of the poorest, the middle class and the richest developed over the past two decades?

To answer this question, I divide society into 20 parts (ventiles in the statistical jargon). The first ventile (V1 in Figures 2 and 3) includes the poorest 5 percent of the population. The second ventile (V2) includes the next poorest 5 percent, and so on. The twentieth ventile (V20) includes the top 5 percent earners (see the annex for some data issues and methodological notes).

Figure 2 shows that in 2004, the poorest 5 percent in the CEE12 possessed only 13 percent of the income of the poorest 5 percent in EU10 – well below the overall average of 34 percent reported in Figure 1. Yet the poorest CEE12 citizens benefitted from very fast income growth: by 2023, their income relative to the poorest 5 percent of EU10 increased to 54 percent 2 To be more precise, I compare the income of the poorest 5 percent in 2004 with the poorest 5 percent in 2023, but it may not include the same people. For example, the bottom 5 percent in the CEE12 in 2004 included 3.4 million Romanians and 101,000 Hungarians, while in 2023, it included 2.1 million Romanians and 675,000 Hungarians. . This is still below the 62 percent average reported in Figure 1, yet income growth was very fast.

In contrast, the top 5 percent in the CEE12 had a better relative position compared to the top 5 percent in EU10 in 2004 (41 percent), while the improvement from 2004-2023 was less impressive, with their relative share increasing to 59 percent, which is also below the 62 percent average reported in Figure 1. 

It is the upper middle class (from ventiles 10 to 19) in the CEE12 who achieved a relative income position above the 62 percent average in 2023.

Real income growth

While Figures 1 and 2 expressed CEE12 household income as a percent of EU10 household income, it is instructive to show the total growth of real income from 2004-2023 (Figure 3).

In real terms, the incomes of the poorest 5 percent of CEE12 citizens were lifted by a staggering 256 percent in total from 2004 to 2023. That is, if income was €100 in 2004, it increased, in real terms, to €356 by 2023. This corresponds to a 6.8 percent average annual real income increase, well above the EU average annual of growth of just 0.7 percent. Figure 3 shows lower income growth rates for richer CEE12 people. Real income growth for the top 5 percent was the lowest among the twenty ventiles, though still high at 60 percent in total from 2004-2023, or 2.5 percent per year on average. Such inequality-reducing income growth developments in CEE12 are in contrast to the experience of the EU10, where the highest earners benefitted from the fastest income growth, even if its rate at 0.5 percent per year (in real terms) falls short of the CEE12 growth rate.

Looking forward

Two decades ago, EU membership offered the CEE12 various opportunities to boost growth. For example, market institutions have been strengthened, improving the business environment and attracting investment from abroad, which has boosted the capital stock and productivity, and has improved corporate governance. Access to the single market has increased competition, fostering the efficiency of firms. 

Meanwhile, labour mobility resulted in an exodus from CEE12 to EU10 and the United Kingdom, which offered the opportunity to gain work experience in a more productive environment, upgrade skills and accumulate savings from higher incomes than at home. Such capital accumulation of CEE12 emigrants also supported their home countries in the form of remittances and investments in property and businesses back home. Access to high-quality education in other EU countries improved the human capital of CEE12 young people. All these effects supported the economies of CEE12 countries, even if many of those who left the CEE12 have not returned. Financial support from EU funds might have also supported growth.

Such beneficial effects, coupled with much lower wage levels in CEE12 than in EU10 and the geographical closeness of the region to EU10, were likely instrumental in the CEE12 economic successes over the past two decades.

The annual pace of catching up measured in terms of per-capita income remained broadly stable over the past two decades, despite the crises that some of these countries suffered during the global financial crisis, and the gradual increase in local labour costs, which reduces the attractiveness of these countries for labour-intensive investments. At an unchanged pace from the past two decades, both GDP per capita and household income would reach the EU10 level by about 2046. But is this realistic?

The key question is whether CEE12 economies can shift towards higher value-added activities as their relative wage levels increase. Some of the EU-enlargement-related productivity gains might be one-offs (ie it is easier to improve productivity from a low level, but more difficult from a higher level), while CEE12 economies already suffer from labour shortages, and the demographic outlook of these countries is even worse than in EU10. Innovation activities in the region are weaker than in EU10, as reflected in the European innovation scoreboard 3 See European Commission, ‘European innovation scoreboard’,…. , which, in 2023, ranked five CEE12 countries in the worst of the four innovation performance categories, six other CEE12 countries in the second worst category, and only Cyprus making it into the second-best category. Secondary school education quality scores are mixed across the region, with five CEE12 (Estonia, Poland, Czechia, Slovenia and Latvia) having better scores than the OECD average in at least two of the three subjects considered (mathematics, reading and science). Two countries (Lithuania and Hungary) are close to the average, while the remaining five countries are below average. Most CEE12 universities are poorly ranked 4 See . Weak innovation and education performance might adversely affect longer-term growth prospects.

Therefore, while there is a lot to celebrate, twenty years on from the EU’s eastern enlargement, major challenges lie ahead to maintain the momentum of economic convergence.


Baas, T. and H. Brücker (2010) ‘Macroeconomic impact of Eastern enlargement on Germany and UK: evidence from a CGE model’, Applied Economics Letters 17(2): 125–28, available at 

Bhalla, Surjit S. (2002) Imagine there is no country: poverty, inequality, and growth in the area of globalization, Institute for International Economics, available at 

Caliendo, L., L.D. Opromolla, F. Parro and A. Sforza (2021) ‘Goods and Factor Market Integration: A Quantitative Assessment of the EU Enlargement’, Journal of Political Economy 129(12): 3491-3545, available at 

Darvas, Z. (2019) ‘Global interpersonal income inequality decline: The role of China and India’, World Development 121: 16-32, available at

European Commission (2009) ‘Five years of an enlarged EU: Economic achievements and challenges’, European Economy 1/2009, Directorate-General for Economic and Financial Affairs, 

Oberhofer, H. and H. Winner (2015) ‘Handelseffekte der österreichischen EU-Integration’, FIW Policy Brief 28, Forschungsschwerpunkt Internationale Wirtschaft (FIW), available at 

Data availability and missing data approximations

GDP per capita at purchasing power standards (PPS) is available for all countries for 1995-2023 in Eurostat’s ‘Main GDP aggregates per capita [nama_10_pc]’ dataset. 

GNI per capita at PPS is available only for 2018-2022 in Eurostat’s ‘GNI (gross national income) per capita in PPS [nama_10_pp]’ dataset. 

Household mean income, as well as information about the within-country income distributions, are available mostly for 2004-2022 in Eurostat’s ‘Mean and median income by age and sex - EU-SILC and ECHP surveys [ilc_di03]’ and ‘Distribution of income by quantiles - EU-SILC and ECHP surveys [ilc_di01]’ datasets, but there are a few data gaps 5 Note that Eurostat reports the various EU-SILC statistics for the survey year, and not for the underlying income reference period. The income reference period in EU-SILC is a fixed 12-month period (such as the previous calendar or tax year) for all countries except Ireland, for which the survey is continuous and income is collected for the last twelve months. That is, except for Ireland, a value for a particular year in the Eurostat’s EU-SILC database refers to the previous year, eg the latest 2023 values refer to 2022. . For 2022, data is available for all CEE12 countries except Malta, for which I approximated mean household income at PPS by increasing its 2021 value with the growth of GDP per capita at PPS, adjusted by the average growth difference between mean income and GDP per capita over the preceding decade. The same approximation was made for Luxembourg, for which 2022 mean household income data was not available. For 2023, mean income of all twelve CEE12 countries and ten EU10 countries were approximated the same way.

For Bulgaria, the first observation for mean income and the income distribution is for 2005, while for Romania, the first observation is for 2006. For Bulgaria for 2004, I reduced the 2005 mean income by the growth rate of GDP per capita in 2005 adjusted by the average difference in the growth rates of mean income and GDP per capita in the subsequent ten years. An analogous approximation was made for Romania’s mean income in 2005 and 2004.

Whenever income distribution data was not available (Bulgaria for 2004; Romania for 2004-2005; Malta and Luxembourg for 2022; all countries for 2023), I assumed unchanged within-country distribution of the closest available year.

Approximating the full income distribution from limited income shares data

Eurostat publishes income shares data for the following income groups in each country: the 1st, 2nd, 3rd, 4th, 5th, 95th, 96th, 97th, 98th, 99th and 100th percentiles, the ten deciles and the four quartiles in its dataset ‘Distribution of income by quantiles - EU-SILC survey [ilc_di01]’ 6 Income surveys are the main source for such income-distribution data, which might be subject to incorrect reporting. Top incomes are generally incorrectly measured in income surveys, due to the under-representation of rich people in such surveys and the under-reporting of the income of those who participate. Poorer people might also under-report their income, especially if it is earned on the black market. Barter transactions (especially in underdeveloped rural areas) might further distort the income measurement of the poorest segment of society. Statistics offices often adjust the data to limit the scope for such anomalies. . Data for other percentiles are not made public. Income share data is rounded to one digit after the decimal, which severely limits the usefulness of data for lower income percentiles. For example, for Bulgaria, the reported income shares of the lowest percentile for 2011-2015 are 0.0, 0.1, 0.0, 0.1, 0.0, respectively, which, literally, would imply an infinite increase in mean income from 2011 to 2012 and from 2013 to 2014, and complete disappearance of income from 2012 to 2013 and from 2014 to 2015. It is unfortunate that Eurostat does not publish at least three significant digits for all income share indicators.

Therefore, I use an approximation, the so-called Lorenz Curve regression method of Bhalla (2002), to estimate the income shares of imprecisely reported percentiles and those percentiles of the income distribution that are not available at all. In Darvas (2019), I found that the Lorenz Curve regression method is very precise.

Income distribution of country groups

Using the estimated income shares for each percentile and mean income, I calculate the mean income (in current-price PPS) for each percentile of each country in each year. With the help of population size in each country, I calculate the number of people belonging to each percentile (1/100th of the total population) and I assume that all people in each percentile in a given country have the same income. To calculate the combined income distribution of CEE12 for a particular year, I include the number of people (with their associated income) of each percentile of each of the twelve countries for that year and calculate income distribution statistics from this combined distribution. Similar calculations are used to approximate the combined income distribution of EU10.

Still, since my estimates are approximations, I do not report percentile data on Figures 2 and 3, but ventile data, which is less prone to approximation errors.

Constant price incomes

Price levels differ between countries, which calls for the use of PPS-adjusted income indicators. Eurostat publishes only current-price PPS income variables, which are ideal for cross-country comparison in a particular year, but not for comparison across time, because it reflects inflation. Therefore, for the calculation of real income growth in Figure 3, I adjusted the current-price PPS income growth with EU27 inflation. The reason for using EU27 inflation is that PPS mean income values are measured at EU27 average prices.

About the authors

  • Zsolt Darvas

    Zsolt Darvas is a Senior Fellow at Bruegel and part-time Senior Research Fellow at the Corvinus University of Budapest. He joined Bruegel in 2008 as a Visiting Fellow, and became a Research Fellow in 2009 and a Senior Fellow in 2013.

    From 2005 to 2008, he was the Research Advisor of the Argenta Financial Research Group in Budapest. Before that, he worked at the research unit of the Central Bank of Hungary (1994-2005) where he served as Deputy Head.

    Zsolt holds a Ph.D. in Economics from Corvinus University of Budapest where he teaches courses in Econometrics but also at other institutions since 1994. His research interests include macroeconomics, international economics, central banking and time series analysis.

Related content