Ever since it was founded, the EU has made the raising of national living standards to a high level a core objective of European integration (European Economic Community, 1957; European Union, 1992). The financing of European cohesion policy, which accounts for 34 percent of the European budget (Darvas/Wolff, 2018), is also based on this objective. Cohesion funds are particularly important for less economically advanced countries: for example, the allocation of EU funds to Romania alone in the 2014–2020 funding period amounted to around 31 billion euros – compared with a Romanian gross domestic product of around 170 billion euros.

The high importance of the prospect of convergence for poorer member states and the large payments of the richer countries justify devoting considerable attention to evaluation. At the same time, a causal analysis of cohesion policy has proved to be extraordinarily difficult – first of all, because the question of a measure for the standardisation of living conditions must be answered. An important indicator for the evaluation of convergence efforts is GDP adjusted for purchasing power. Starting from the idea that poorer countries need above-average growth to catch up to richer countries, the figure represents the so-called beta-convergence. According to this measure, by and large, convergence within the EU takes place when the poorer South and East experience faster growth than the richer West and North.
Viewed historically, in terms of real GDP per capita, such a convergence process has indeed taken place within the EU-15. Since the 1950s, economic output in the relatively poor countries of southern Europe – Greece, Portugal and Spain – has risen sharply, with the lowest growth being in prosperous Denmark. In particular, until the 1980s, the EU-15 has moved closer together economically (Goecke, 2013).

With the entry of the Eastern European countries – whose income levels to some extent lag far below those of Southern Europe – into the EU family, the frame of reference for analysis has fundamentally shifted. And indeed, at first glance, between 1999 and 2016 a solid recovery in the Baltic and Eastern European economies can be seen compared with the EU average, which however is driven by very high growth rates in the pre-crisis period. The disastrous performance of Italy is conspicuous, in addition to the weak performance of Greece, both caused by the financial and economic crisis. In a comparison between 1999 and 2016 adjusted for purchasing power, Italy has lost an entire 25 percent compared to the EU-28 average (del Hoyo et al., 2017).

At the same time, income prospects have not deteriorated everywhere in Italy. The fact that the industrially strong regions of northern Italy have fared relatively well is hardly taken into consideration when looking at the country’s aggregate GDP. Therefore, in order to take into account the differences within the countries, a differentiated consideration of the regional economic situation is needed. For the regional level, it is useful to consider the so-called NUTS-3 classification of the EU countries (see Goecke/Hüther, 2016). In Germany, this is determined from the division into 402 districts and independent cities, in Italy from the 110 provincias, in France from the 101 départements. Overall, the EU is subdivided into 1,341 NUTS-3 regions (NUTS-2013 classification).