EUROPEAN COMMISSION
Brussels, 27.3.2024
SWD(2024) 79 final
COMMISSION STAFF WORKING DOCUMENT
Accompanying the document
Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions
on the 9th Cohesion Report
{COM(2024) 149 final}
ECONOMIC COHESION
•The 2004 enlargement triggered a remarkable convergence of GDP per head. In central and eastern Europe as a whole, income per head increased from 45 % of the EU average in 1995 to nearly 80 % today. Nevertheless, large differences persist; there is ample room for further upward convergence.
•Across the EU, regional disparities narrowed until the financial crisis but then stagnated, mostly because of slower growth of less developed regions in central and eastern Europe and the divergence of some less developed and transition regions, especially in southern Europe.
•Around a third of EU regions – less developed, transition, and more developed regions alike – have yet to see a return to 2008 levels of GDP per head. These are primarily in Italy, Spain, Greece and France, but also in Germany, Finland and the Netherlands. This poor performance is due to slowing growth of productivity, investment and employment.
•Growth of GDP per head in the EU averaged 1 % a year over the period 2001– 2021, but in many regions it stagnated or even declined. In many cases, stag- nation came along with little or no increase in household income and growing inequalities, fuelling political discontent and a decline in support for democratic values and the EU.
•On the positive side, several regions escaped stagnation, using their local strengths to move to more complex economic activities and become integrated into European and global value chains.
•The recovery from the COVID-19 pandemic has been faster than after the 2009 recession, partly because of swift EU policy action, with the rapid mobilisation of Cohesion Policy and the adoption of NextGenerationEU. More recently, escalating geopolitical tensions, with war erupting on the EU’s doorstep, and the surge in energy, raw materials and food prices have exacted a heavy toll on many EU regions.
•Looking ahead, disparities between EU regions and current candidate coun- tries are large but not unlike those between the EU-15 and accession countries in 2004, suggesting that there is a very large untapped potential for further upward convergence.
Chapter 1
Economic cohesion
1.Introduction
Reducing territorial disparities is a cornerstone of European integration, dating back to the Treaty of Rome, which sets the goal of ‘reducing the differ- ences existing between the various regions and the backwardness of the less-favoured regions’. Accordingly, Cohesion Policy is not only the most visible expression of EU solidarity but also a cen- tral pillar of its Single Market and growth model1. Removing barriers to the free movement of goods, services, capital and workers has promoted a better allocation of resources across the EU and fostered the exchange of ideas and innovation. However, market forces alone do not ensure that everyone benefits from economic integration. By investing in infrastructure, innovation, education and other key areas, Cohesion Policy helps less developed regions directly and all other regions indirectly to reap the benefits and economies of scale created by the Single Market.
This report comes 31 years after the introduction of the EU Single Market, 25 years after the launch of the euro and 20 years after the historic EU east- ern enlargement of 2004. It shows the remarkable economic convergence that eastern regions and countries have achieved since then. GDP per head in central and eastern Europe (shortened to ‘east- ern Europe’ in this report) increased from around 45 % of the EU’s average in 1995 to 52 % at the moment of accession in 2004, to nearly 80 % in 2021. This is an extraordinary achievement of Eu- ropean integration and Cohesion Policy, which has invested nearly EUR 1 trillion to support balanced economic development in the EU since 2000.
Some parts of Europe, however, have found it more difficult to converge. As indicated in previ- ous reports, GDP per head in some transition and less developed regions began to diverge from the EU average after the 2009 recession, revealing an increased likelihood of falling into what can be termed a ‘development trap’2, with implications for social and territorial cohesion (Chapters 2 and 3).
Most recently, the outbreak of the COVID-19 pan- demic and escalating geopolitical tensions, with war erupting on the EU’s doorstep, have tested co- hesion. The disruptions in global supply chains and the surge in energy, raw materials and food prices have exacted a heavy toll on households – espe- cially the most vulnerable ones – and the economy at large. Despite encouraging signs of recovery, the long-term impact of these events on cohesion remains difficult to predict, especially in a context where secular structural challenges linked to the green and digital transitions are set to reshape much of the EU economy (Chapters 4, 5 and 6).
Against this background, this chapter provides an update of the state of economic cohesion in the EU by assessing long-term economic convergence between regions over the past 20–30 years and the short-term impact of the pandemic. Tapping into the growth potential of the 82 regions with GDP per head below 75 % of the EU average is key to fostering convergence and the prosperity of the EU. Accordingly, it indicates how productivity and competitiveness have evolved across regions and how they can contribute to economic cohesion going forward.
1See Box 1.6.
2The likelihood of being in a development trap is measured by a composite indicator capturing a protracted period of low or negative growth, weak productivity increases and low employment creation. See: Diemer et al. (2022) and European Commission (2022a).
Chapter 1: Economic cohesion
2.Long-term trends
in convergence and regional economic cohesion
Differences in regional GDP per head in the EU have steadily diminished over the past two dec- ades but there is ample room for further upward convergence3. Some 20 years after the 2004 en- largement, the regions then entering the EU have achieved a remarkable economic convergence, with GDP per head in eastern Europe increasing from 50 % of the EU average in 2004 to nearly 80 % in 2021. However, there is still substantial scope for further convergence. Over 1 in 4 people in the EU (28 %) still live in regions with GDP per head below 75 % of the EU average in PPS terms4, most of them in eastern Member States, but also in outermost regions and increasingly in southern Europe (Map 1.1 and Chapter 3)5. In Bulgaria, for instance, GDP per head was below 50 % of the
EU average in all regions, except in Yugozapaden, the capital city region. To put this into perspective, differences in GDP per head across US states bot- tom out at about 60 % of the US average and only 1 in 12 people live in a state with GDP per head below 75 % of the US average6. This suggests that there is still a large untapped potential for upward convergence in GDP per head – and in living stand- ards – within the EU. Moreover, around a third of EU regions – with a similar share of EU population, around 155 million people in total – have a GDP per head that is yet to return to its 2008 level. These are equally divided between less devel- oped, transition and more developed regions and are present in 12 Member States: Italy (19), Spain (15), Greece (12), France (9), Germany (5), Finland
(4), the Netherlands (4), Portugal (3), Romania (3),
Austria (2), Belgium (1) and Luxembourg (1).
Figure 1.1 Annual growth in real GDP per head in EU regions by level of development, 2001–2021
6
Annual average % growth of real GDP per head
5
4
3
2
1
0
-1
-2
-3
0
25
50
75
100
125
150
175
200
225
250
GDP per head (PPS, % of EU average) in 2000
Source: Eurostat [nama_10r_2gdp] and DG REGIO calculations.
3European Commission (2023).
4GDP per head in PPS terms is the total value of goods and services produced per inhabitant adjusted for differences in price levels between countries. Regions here and throughout the chapter are defined at the NUTS 2 level.
5The EU includes nine outermost regions: Guadeloupe, La Réunion, Mayotte, Guyane, Martinique and Saint-Martin (France), Madeira and Açores (Portugal) and Canarias (Spain). In the outermost region of Mayotte (France), for instance, GDP in PPS was as low as 28 % of the EU average in 2021.
6Clearly the US is not comparable to the EU in political or historical terms but it remains the most comparable economic area in terms of market size, economic development, geographical area and population. It is therefore a relevant benchmark from an economic cohesion perspective: see Head and Mayer (2021). It should be noted, however, that EU NUTS 2 regions are on average smaller in size than US states, which in itself tends to increase disparities.
Figure 1.2 GDP per head in EU regions, PPS, 1995–2021, % of EU average
150
GDP per head (PPS), index EU-27=100
140
130
120
110
100
90
80
70
60
50
40
Less developed
Transition
More developed
Eastern
North-western
Southern
150
140
130
120
110
100
90
80
70
60
50
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
40
Source: Eurostat.
Growth of GDP per head over the past two decades has been robust in eastern regions but subdued in southern and some outermost ones. Over the 2001–2021 period, GDP per head in real terms in- creased in most EU regions, though by only 1 % a year or less in most north-western and southern regions. In line with standard economic conver- gence theory, regions with low levels of GDP per head experienced higher rates of growth on aver- age (Figure 1.1). Per capita growth was particularly high in eastern regions (around 2.5 % a year on average)7. There are, however, exceptions. In most regions in Greece and Italy, in particular, GDP per head fell over this period. At the same time, growth was very low in transition regions in France and Spain and negative in a few more developed re- gions in north-western Europe (Figure 1.2). In the recent past, for the first time in the post-war peri- od, nearly 1 in 6 regions in the EU, 39 in total with over 60 million people, experienced two decades in which GDP per head declined8. The next section
examines convergence dynamics further using a range of statistical indicators.
2.1Key indicators of economic convergence
There are important differences in convergence dynamics between the EU-27 and the EU-15 (i.e. the 15 Member States before the 2004 enlarge- ment). A commonly used statistical indicator to as- sess disparities in GDP per head is the coefficient of variation, which is a measure of its dispersion across regions (see Box 1.2)9. This indicator shows that disparities in GDP per head across EU regions declined sharply over the period 2000–2021 (Fig- ure 1.3). On the one hand, this was largely driven by strong upward convergence of eastern regions. On the other hand, it is evident that convergence dynamics differ markedly between the EU-27 and the EU-15. In the former, regional disparities de- clined up until 2009 and stabilised afterwards.
7Many of the eastern Member States have witnessed significant outmigration during the past two decades, thereby lowering the denomina- tor. This trend is of great social and economic importance and is analysed more in detail in Chapter 6. However, the results of exceptional economic convergence are confirmed when measured in terms of productivity or GDP per person employed (see Section 2), a measure that is not affected by net migration. It is also confirmed by indicators of household disposable income and investment. Despite the enormous progress made, this report shows that there remains ample room for forward upward convergence, and a large heterogeneity of income within countries and among households.
818 of the regions are in Italy, nine in Greece, four in Spain, three in France and one each in Portugal, Finland, Austria and Belgium. From 2010 to 2021, GDP also fell significantly in some outermost regions – in Canarias from 83 % of the EU average to 62 %; in the Açores from 75 % to 66 %; and in Madeira from 81 % to 70 % (Eurostat).
9The coefficient of variation is a way of quantifying the variability of a dataset in relation to its mean. It is calculated by dividing the standard deviation by the mean and then expressing this as a percentage, allowing for comparisons between datasets with different units or scales.
Figure 1.3 Regional (NUTS 2) disparities, EU-27 and EU-15, GDP per head (PPS)
50
EU-27
EU-15
Coefficient of variation, % of mean
45
40
35
30
25
Source: DG REGIO calculations based on Eurostat data.
In the EU-15, disparities declined up until 2006 and at a much slower pace and began to increase af- ter 2009. The coefficient of variation indicates that regional disparities in the EU-27 were still some 30 % larger in 2021 than those in the EU-15 in 2004, suggesting that ample room for upward con- vergence remains.
Regional disparities are wide in many Member States and have tended to widen further in most
of them since 2000 (see also Chapters 2 and 3). In Member States with more than four regions, regional disparities in GDP per head increased in 11 of the 19 Member States concerned between 2000 and 2021 (Figure 1.4). Increases were larg- est in Bulgaria, Croatia and Czechia, but there were also increases in the EU-15, in Denmark, Greece and France. On the other hand, disparities declined in Portugal, Austria, Belgium and Germany. The drivers of within-country regional disparities are
Figure 1.4 Coefficient of variation within Member States, GDP per head (PPS), NUTS 2 regions, 2000 and 2021
0.6
2000
2021
0.5
0.4
0.3
0.2
0.1
0.0
AT
BE
BG
CZ
DE
DK
EL
ES
FI
FR
HR
HU
IT
NL
PL
PT
RO
SE
SK
Source: DG REGIO calculations based on Eurostat data.
Box 1.1 Household disposable income and economic cohesion
Household income per head can be used to show how convergence in GDP per head is reflected in people’s income (Figure 1.5). As for GDP per head, there are large regional differences in growth rates of household income. Net household disposable in- come (NHDI) per head relative to the EU average in- creased steadily between 2000 and 2020 in eastern regions (from 45 % to 75 %) and, to a lesser extent, in less developed regions as a whole (from 60 % to 70 %). On the other hand, it declined substantially in southern regions between 2000 and 2012 (from 115 % to below 100 %) and remained unchanged up until 2020, when it fell (to 95 %) because of the ef- fect on their economies of the COVID-19 pandemic.
GDP and household income per head are key indi- cators for assessing economic convergence and dis- parities across regions, but do not shed light on the extent to which the benefits of growth are shared among people within regions. There were large re- gional differences in growth rates of mean equiva- lised household income across the EU (Figure 1.6).
Over this period, two thirds of regions experienced growth in mean household income, whereas the rest registered no growth or a decline. Many of the high- growth regions are in eastern Europe, while many of those with no growth or a decline are in southern Europe. However, a number of advanced economies from north-western Europe (France, Austria, Bel- gium and Denmark) also saw mean household in- come stagnate during this period. The largest differ- ences in growth rates occur between and not within countries. An exception is France, with some regions experiencing sustained growth and others a decline, including some of the outermost regions1. Moving beyond average income, the European Commission found that high-income households in the EU have benefited most from income growth in countries where growth was above the EU average over the period 2007–2017 (largely catching-up countries)2. Conversely, in countries where income declined, the decline was more equally distributed.
Figure 1.5 Net households disposable income per head in PPS, % of EU average, by group of NUTS 2 region, 2000–2020
Eastern
North-western
Southern
Less developed
More developed
Transition
140
130
120
110
100
90
80
70
60
50
40
Source: Eurostat.
1Significant differences in disposable income persist between some French outermost regions and mainland regions. In Mayotte, the
yearly median disposable income was EUR 3 140 in 2019, far below the national average of EUR 21 680.
2European Commission (2020).
Figure 1.6 Growth in mean equivalised disposable household income, 2010–2019
National Average
NUTS regions
10
Annual growth rate of mean disposable income
Figure 1.6 Growth in mean equivalised disposable household income, 2010–2019
8
6
4
2
0
-2
-4
-6
LT
EE
LV
PL
IE
CZ
HU
SE
DE
FI
SK
BE
AT
DK
LU
ES
FR
IT
EL
Note: NUTS 3 regions for DK, EE, LT, and SK, NUTS 2 regions for AT, CZ, ES, FI, FR, IE, LU, LV, and PL, and NUTS 1 for the remaining countries. Households are defined as one or more persons living in the same dwelling. Disposable income is defined after taxes and transfers. This is equivalised by dividing the total disposable income of the household by the square root of the number
of household members.
Sources: OECD computations based on microdata from the Luxembourg Income Study (LIS) and EU Statistics on Income and Living Conditions (EU-SILC).
Survey-based data shed light on the distribution of regional income between households. Inequali- ties tend to be persistent and high in EU regions3. The top 20 % of households in EU regions, in terms of income, received on average almost 5 times (4.7) more than the bottom 20 % in 2019, an in- crease of 5 % from 2010. However, increased in- equality was not common to all regions. Only in a
small majority of regions (54 %) did top incomes grow more, or decline less, than bottom incomes, and in the rest income inequality narrowed (Fig- ure 1.7). In regions with increasing household in- come inequality, this was driven by low-income households becoming poorer rather than high-in- come ones becoming richer.
Figure 1.7 Growth in mean equivalised disposable household income for the top and bottom quintiles, 2010–2019
Annual growth rate of the top quintile (top 20 %)
12
10
8
6
4
2
0
-2
-4
-6
-12
-10
-8
-6
-4
-2
0
2
4
6
8
10
Annual growth rate of the bottom quintile (bottom 20 %)
Note: NUTS 3 regions for DK, EE, LT, and SK, NUTS 2 regions for AT, CZ, ES, FI, FR, IE, LU, LV, and PL, and NUTS 1 for the remaining countries. Households are defined as one or more persons living in the same dwelling. Disposable income is defined after taxes and transfers. This is equivalised by dividing the total disposable income of the household by the square root of the number
of household members.
Sources: OECD computations based on microdata from the Luxembourg Income Study (LIS) and EU Statistics on Income and Living Conditions (EU-SILC).
3 OECD (2022).
quite heterogeneous across Member States. More developed regions (typically capital city regions) are generally widely outperforming other regions in eastern Member States such as Bulgaria or Roma- nia. In other Member States, such as Portugal, the decline in regional disparities is due to low growth in some developed, previously dynamic, regions. In France, instead, internal disparities increased be- cause growth of GDP per head in regions with low levels was particularly slow. Differences in GDP per head within Member States are often as large as between Member States, indicating that important regional variations are often hidden by national av- erages. The same holds for disparities in employ- ment rates and in a number of social indicators, including between rural and urban areas (Chapters 2 and 3)10. Convergence trends in household dis- posable income show some similarities with those of GDP per head but also differences (see Box 1.1).
GDP per head in less developed regions grew, on average, faster than in other regions before the 2009 recession but not after it. Another widely
used indicator of convergence is the beta coeffi- cient (see Box 1.2), which shows the tendency for lower-income economies or regions to grow fast- er than higher-income ones, narrowing disparities over time. As seen above, this has indeed hap- pened since 2000, especially among less devel- oped regions in eastern Europe. However, in the EU-15, though regions with lower GDP per head grew faster than those with higher levels over the 12 years 1996–2008, their growth was lower in the 12 years 2009–202111. The estimated beta coefficient of convergence indeed turned from negative (Figure 1.8) to positive after the global recession (Figure 1.9). In the EU-12 (those before 1995), GDP per head in lower-income regions grew faster than in higher-income ones throughout the period, but not to the same extent after the global recession. The estimated beta coefficient, indeed, remained negative, as expected, but declined by a third12. This tendency is consistent with a larger fall than elsewhere in investment and total factor productivity in many of the countries concerned after the global recession13.
10Participation rates, for instance, are very high in some Member States (e.g. 82 % in the Netherlands, and almost 90 % in Åland in Finland), but much lower in Greece (63 %), as low as 44 % in Sicilia, and under 40 % in Mayotte.
11The beta coefficient remained more stable in the NUTS 2 regions in the EU-12. As expected with logarithmic functional forms and standard economic theory, it flattened slightly over time, reflecting assumed decreasing returns to scale and a slowdown in the pace of convergence the closer a region gets to the technological frontier.
12A significant decline is also found for other estimates of the beta coefficient over time (through rolling regressions) for the EU as a whole.
See: Monfort (2020).
13Through an analysis of conditional beta convergence (see Box 1.2), Licchetta and Mattozzi (2022) find that limited productivity catch-up is a major explanation for the lack of convergence, especially of southern regions. However, they also note that capital accumulation was particularly sluggish in the euro area in the decade following the global recession and gross fixed capital formation (GFCF) took 10 years to return to its pre-recession level. This was in sharp contrast to the period before 2008, where growth in GFCF was higher than average in many euro area converging countries, although largely (and arguably excessively) concentrated in the construction sector, where it declined markedly afterwards.
Figure 1.8 Estimated beta-coefficient for NUTS 2 regions in the EU-15 and EU-12, 1996–2008
NUTS 2 regions in EU-15
Average GDP per head growth (%), 1996–2008
8
6
4
2
0
-2
-4
8
9
10
11
12
Log line (Ln) GDP per head in 1995
NUTS 2 regions in EU-12
Average GDP per head growth (%), 1996–2008
14
12
10
8
6
4
2
0
-2
-4
7.5
8.0
8.5
9.0
9.5
10.0 10.5 11.0
Ln GDP per head in 1995
Source: DG REGIO calculations based on Ardeco data.
Figure 1.9 Estimated beta-coefficient for NUTS 2 regions in the EU-15 and EU-12, 2009–2021
NUTS 2 regions in EU-15
Average GDP per head growth (%), 2009–2021
8
6
4
2
0
-2
-4
14
Average GDP per head growth (%), 2009–2021
12
10
8
6
4
2
0
-2
-4
NUTS 2 regions in EU-12
8
9
10
11
12
Ln GDP per head in 2008
7.50.0
8.50.0
9.50.0
10.50.0
11.50.0
Ln GDP per head in 2008
Source: DG REGIO calculations based on Ardeco data.
Differences in economic structure and geograph- ical features can partly explain differences in the pace of convergence. A recent statistical approach is built around the notion of ‘club convergence’14.
The clubs or clusters concerned may have a com- mon economic structure, geographical features or other characteristics that affect the pace of con- vergence. One study15 employs this approach to
14In this context, measures of club convergence, such as pair-wise statistical convergence, enable convergence, or divergence, to be exam- ined between pairs of countries or regions, rather than examining entire groups simultaneously as with sigma and beta convergence: see Pesaran (2007). The measure, therefore, complements these more traditional indicators by allowing for the identification of patterns of convergence within the sample analysed.
15Arvanitopoulos and Lazarou (2023).
Box 1.2 Three indicators of statistical convergence: sigma, beta and club convergence
These three concepts are often used in empirical research to assess dynamics of economic develop- ment and convergence among different countries or regions and to explore whether disparities are di- minishing, how fast convergence is occurring, and whether different types of economies exhibit differ- ent convergence patterns.
Sigma (σ) convergence
Sigma convergence refers to a situation where the dispersion or inequality of income, or other indi- cators, between countries or regions declines over time. Accordingly, it indicates that the standard devi- ation – a measure of dispersion around the mean – is narrowing, pointing to a reduction in disparities. In this report, the coefficient of variation, which ex- presses the standard deviation as a percentage of the mean, is used to examine the presence of sigma convergence.
Beta (β) convergence
Beta convergence is an indicator of the rate at which different economies are approaching a common ‘steady state’ of economic development or income1. It shows whether lower-income countries or regions grow at a faster pace than higher-income ones, leading to a reduction in disparities between them. A related concept is that of conditional beta conver- gence, as used, for instance, in the study by Licchet- ta and Mattozzi referenced above. This starts from beta convergence but enables account to be taken of the influence of specific conditions or features on the rate of convergence in addition to initial levels
1Barro and Sala-i-Martin (1992).
2Quah (1996).
3Pesaran (2007).
of GDP per head. Conditional beta convergence al- lows for a more nuanced analysis of convergence dynamics by recognising that factors such as invest- ment, education or governance can also affect the rate at which economies catch up with others.
Club convergence
Club convergence refers to the notion that groups or ‘clubs’ of countries or regions may exhibit dis- tinct patterns of economic convergence2. These may have a common economic structure, geographical features or other characteristics that can at least partly explain different paces of convergence. Within this, pair-wise statistical convergence is a method that assesses the convergence or divergence be- tween pairs of countries or regions, rather than look- ing at entire groups simultaneously as with sigma and beta convergence3. The method is often used to identify and analyse distinct groups of economies that exhibit similar convergence patterns (club con- vergence). It allows researchers to determine which countries or regions are moving closer together and which are not, so increasing understanding of dif- ferences in convergence patterns within a broad- er group of economies. Overall, the results for EU regions found by Arvanitopoulos and Lazarou are broadly in line with those obtained by Pesaran for the world economy. While technological progress seems to have been spreading reasonably widely across economies, there are important geographical and structural factors that mean there are differ- ences in GDP per head that remain persistent.
EUROPEAN COMMISSION
Brussels, 27.3.2024
SWD(2024) 79 final
COMMISSION STAFF WORKING DOCUMENT
Accompanying the document
Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions
on the 9th Cohesion Report
{COM(2024) 149 final}
Box 1.3 Cohesion cycles in the 2000s: a regional snapshot
In broad terms, four cohesion sub-periods can be distinguished in the two decades 2001–2022.
The ‘convergence years’ (2001–2008)
Between 2001 and 2008, nearly all regions experi- enced growth in GDP per head. Overall, growth was above average in both the less developed and the transition regions, with rates of over 5 % a year in many eastern Member States. This is in line with traditional economic growth theories, which predict that growth will tend to be higher the lower the in- itial level of GDP per head. Most of these regions are in less developed and moderately developed Member States, where for the most part growth was faster than the EU average. In Romania and Bulgar- ia, where growth was particularly high, catching-up was not uniform across the country but was driven by the capital city region. Regions in southern Italy, however, did not follow this pattern of catching up. They already experienced a decline in GDP per head in the 2000s even though their GDP per head was well below the EU average.
The ‘low employment period’ (2009–2013)
The global recession of 2009 led to GDP per head in the EU declining between 2009 and 2013, with many of the less developed and transition regions growing more slowly (or shrinking more quickly) than the EU average, so reversing the earlier ten- dency towards convergence. Around 60 % of the EU population lived in regions with a declining GDP per head. The regions hit hardest were mainly in the southern EU, though also in Romania, Ireland and Finland. In most Greek regions, the reduction in GDP per head averaged over 3 % a year. Notable excep- tions were most regions in Poland and some in Bul- garia and Romania.
The ‘delayed recovery’ (2014–2019)
The 2014–2019 period shows a clear recovery from the Great Recession. Almost all regions experienced growth in GDP per head, though at a lower rate than in the pre-recession period. High growth rates were restored in most eastern regions, so leading again to convergence. Growth in many north-western regions also remained below pre-crisis rates, Ireland being the main exception. In many regions in the hard-hit southern Member States, especially in Portugal and Spain, growth rates recovered, but in Greece and many regions in Italy growth remained low. Overall, 10 years after the 2009 financial crisis, over a quar- ter of the EU population still lived in regions where real GDP per head had not returned to pre-crisis levels. This includes the entire population of Greece and Cyprus, 80 % of the population of Italy and a third of that of Spain, but also 75 % of the popula- tion of Finland and over a third of that of Austria. In most of the eastern Member States, GDP per head had returned to pre-crisis levels in all or nearly all regions. However, in Romania and Croatia, 40 % and 25 % of the population, respectively, lived in regions where this was not the case.
The ‘quick rebound’ (2020–2022)
The 2020–2022 period is characterised by the dou- ble shock of the COVID-19 pandemic and Russia’s war of aggression in Ukraine. Due to the nature of these shocks, they affected some regions more than others and – within them – some workers and sectors more than others (e.g. tourism, cultural ac- tivities, and industries affected by supply chain dis- ruptions and high energy prices). Again, southern Europe was on average more heavily affected. How- ever, as discussed below, the ensuing economic re- covery was faster and more broad-based than after the 2009 recession.
1.The short-term impact on economic cohesion of the COVID-19 pandemic
The COVID-19 outbreak had a severe impact on the EU economy and society, but GDP rebound- ed strongly in 2021 after a massive downturn in 2020. GDP fell in all but three EU regions. The unprecedented, bold and co-ordinated economic policy actions taken, including through Cohesion Policy, mitigated the economic and social impact of the pandemic. GDP at EU level already exceed- ed the pre-pandemic level by the last quarter of 2021, whereas it took seven years for it to exceed the pre-recession level after 2009. The regional data also indicate a more broad-based recovery in 2021, with less developed, transition and more developed regions all rebounding (Figure 1.12).
Southern Europe, however, was more heavily af- fected by the 2020 recession, with GDP falling by 10 %. Despite a stronger rebound, GDP in 2021 was still 5 % below the pre-COVID peak. North-western and, more especially, eastern regions have fared
significantly better than southern ones in terms of GDP in the wake of the two crises. However, this has not prevented GDP in the EU as a whole falling behind that of the US and other advanced econo- mies (Figure 1.13).
It is too early to be able to fully assess the longer- term impact of the COVID-19 outbreak on eco- nomic cohesion, but so far less developed regions have recovered more quickly than from the 2009 recession. The data available confirm the substan- tial size of the shock in 2020. Overall, the fall in GDP was much larger than during the recession of 2009. As already highlighted in the 8th Cohesion Report31, some regions were hit more than others and – within them – some workers and sectors (such as tourism, cultural activities, and industries affected by supply chain disruptions) more than others. However, the ensuing economic recovery was more broad-based and faster than in 2010, when GDP continued to fall in around a quarter of EU regions (Figure 1.14). In 2021, this was the case in only four regions32. In 2010, the de- cline was largest in less developed and transition
Figure 1.12 Real GDP in NUTS 2 regions by level of development, 2009–2010 (2008=100) and 2020–2021 (2019=100)
a)Real GDP in NUTS 2 regions by level of development, 2009–2010 (2008=100)
Less developed
More developed
Transition
EU
100
b)Real GDP in NUTS 2 regions by level of development, 2020–2021 (2019=100)
Less developed
More developed
Transition
EU
100
98
98
96
96
94
94
92
92
90
2009
2010
90
2020
2021
Source: Eurostat and Ardeco.
1European Commission (2022).
2There is even a slightly negative correlation between regional growth rates in 2020 and 2021, meaning that regions experiencing a deeper fall in GDP in 2020 were, on average, also the ones that experienced a stronger rebound in 2021 (Figure 1.16).
Figure 1.13 GDP at constant prices in the EU, US and OECD, 2008 GDP=100
140
Eastern
North-western
Southern
EU
US
OECD
GDP at constant prices, 2008 GDP = 100
130
120
110
100
90
80
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
Source: Ameco.
regions. In 2021, the regions where GDP fell by most in 2020 were, on average, the ones where the rebound was strongest33.
Despite the broad-based recovery, there are again very large differences in growth rates across re- gions (last panel in Map 1.3). These may reflect differences in the structure of economies, with sectors more heavily affected by restrictions and supply chain disruptions taking longer to recover. Despite the strong rebound, the impact of the cri- sis on economic cohesion was severe and will need to be monitored in the future together with the ef- fect on overall growth in the EU.
The pandemic reduced employment in all regions, but this was largely offset by a strong rebound in 2021. The reduction in the number employed in more developed regions was similar (1–2 %) in both 2009 and 2020 (Figure 1.15 and Figure 1.16). However, eastern, southern and less developed re- gions still had 5 % fewer people in employment one year after the global recession. This was not the case in 2021 and 2022. Employment in the regions most affected began to recover sooner
and it had already reached its pre-crisis peak in 2021 in nearly all of them. Thanks to job-retention schemes and other policy initiatives, the negative impact of the pandemic on employment was much smaller too than in 200934. Indeed, the rapid eco- nomic recovery led to labour shortages reaching or even exceeding pre-pandemic levels in several Member States by the end of the year35. This is in stark contrast with the employment dynamics after the 2009 recession, where employment con- tinued to decline in eastern and southern Europe two years after the recession.
Both the 2009 recession and the 2020 pandem- ic hit household income in southern EU regions in particular (Figure 1.17). Unlike GDP and employ- ment, household income did not decline markedly in the two periods in the EU as whole, suggesting that automatic stabilisers and discretionary meas- ures played an important role in cushioning the im- pact36. However, there are large differences across the EU. Southern regions experienced a significant decline in household disposable income in the two years following the global recession (2010 and 2011). In the rest of the EU, by contrast, it was
3This is suggested by the slightly negative correlation between regional growth rates in 2020 and 2021.
4Giupponi et al. (2022).
5European Commission (2022) and Chapter 2 of this report.
6Bökemeier and Wolski (2022).
Figure 1.14 Real GDP growth rate in 2009 and 2010, 2020 and 2021, NUTS 2 level, year on year
% change
a)Real GDP growth rate in 2009 and 2010
20
b)Real GDP growth rate in 2020 and 2021
20
GDP growth rate in 2010 (% year on year)
GDP growth rate in 2021 (% year on year)
15
15
10
10
5
5
-20
0
-15
-10
-5
0
5
10
-5
0
-20
-15
-10
-5
0
5
10
-5
-10
-10
-15
GDP growth rate in 2009 (% year on year)
Source: Eurostat.
-15
GDP growth rate in 2020 (% year on year)
Note: data for Polish regions are not yet available and not included.
above the pre-recession level. In 2020, the year of the COVID-19 outbreak, household income con- tinued to grow during the recession in eastern and north-western regions. Southern regions, on the other hand, were hit particularly hard, with a larger decline in household income than in 2009, reflect- ing the much larger impact on GDP (5 % in 2009 against 10 % in 2020). The post-pandemic recovery in household income in the southern EU, however, was stronger in 2021, whereas in 2010 income con- tinued to decline. Nevertheless, in 2022 it declined again, largely because of high inflation and a slower adjustment of wages than in the rest of the EU.
The post-pandemic rebound in investment was ex- ceptionally strong, especially in less developed and southern European regions. The fall in investment in 2020, though large (around 5 %), was less than half of that in 2009 (11 %) (Figure 1.18). This con- trasts with the contraction in GDP, which was larg- er in 2020. The difference was even larger in the year following the recession. Investment remained some 11 % below the pre-recession level in 2010, whereas it rebounded to nearly reach the pre-re- cession level in 2021. Significantly, less developed and transition regions performed, on average, bet- ter than more developed regions after the pandem- ic, while the opposite was the case after 2009. The
difference in the two periods partly reflects the ex- ceptional nature of the 2009 recession, when the decline in investment was deeper and more per- sistent than in previous ones (Figure 1.19) and the rebound much slower than in the US and other ad- vanced economies (Figure 1.20).
Both recessions had a substantial adverse im- pact on fiscal balances in the short term, but the COVID-19 pandemic was followed by a more modest increase in public debt over the subse- quent three years (Figure 1.21). During the period 2009–2011, public debt relative to GDP went up by 17 pp in the EU (15 pp in the eastern EU, 13 in the north-western EU, and 24 in the southern EU). By contrast, the increase between 2020 and 2022 was a much smaller 6 pp (6 pp in the eastern EU, 7 in the north-western EU, and 8 in the southern EU). In both periods the US and Japan adopted a more expansionary fiscal stance, resulting in larg- er and more protracted fiscal deficits (Figure 1.22), which ultimately led to an increase in public debt relative to GDP of 51 pp and 78 pp, respectively, between 2008 and 2022 (Figure 1.23). This con- trasts with a more restrained 20 pp increase in the EU over the same period, though in the south- ern EU the increase was 49 pp (as against 12 in the eastern EU and 18 in the north-western EU).
Figure 1.15 Number employed, by geographical area and level of development 2009, 2010 and 2011, 2008=100
a) Number employed by geographical area
b) Number employed by level of development
102
Eastern
North-western
Southern
EU
102
Less developed
More developed
Transition
EU
Number employed, 2008 = 100
Number employed, 2008 = 100
100
100
98
98
96
96
94
94
92
92
90
2009
2010
2011
90
2009
2010
2011
Source: Eurostat and Ardeco.
Figure 1.16 Number employed, by geographical area and level of development, 2020, 2021 and 2022, 2019=100
a) Number employed by geographical area
b) Number employed by level of development
104
Eastern
North-western
Southern
EU
104
Less developed
More developed
Transition
EU
102
102
Number employed, 2019 = 100
Number employed, 2019 = 100
100
100
98
98
96
96
94
94
92
92
90
2020
2021
2022
90
2020
2021
2022
Source: Eurostat and Ardeco.
Although the increase in the southern EU was much the same as in the US, it was not associat- ed with the same economic performance. Follow- ing the 2010 recovery, several EU Member States front-loaded fiscal consolidation measures in an at- tempt to curtail budget deficits. This yielded mixed
results, as GDP often failed to rebound as fore- cast37. However, in the wake of the 2020 COV- ID-19-induced recession, the EU introduced the NextGenerationEU scheme, making available financial aid of some EUR 750 billion to Member States severe- ly affected by the crisis to support cash-strapped
7Blanchard and Leigh (2013).
Figure 1.17 Real gross household disposable income by geographical area, 2009–2011 (2008=100) and 2020–2022 (2019=100)
a)Real gross household disposable income
by geographical area, 2009–2011 (2008=100)
b)Real gross household disposable income
by geographical area, 2020–2022 (2019=100)
104
Eastern
North-western
Southern
EU
104
Eastern
North-western
Southern
EU
Real gross household disposable income, 2008 = 100
Real gross household disposable income, 2019 = 100
102
102
100
100
98
98
96
96
94
94
92
92
90
2009
2010
2011
90
2020
2021
2022
Note: Income is deflated by the harmonised consumer price index; data for MT and BG are missing.
Source: Ameco.
Figure 1.18 Gross fixed capital formation, in real terms, by level of development, 2009–2010
(2008=100) and 2020–2021 (2019=100)
a)Gross fixed capital formation, in real terms, by level of development, 2009–2010 (2008=100)
Less developed
More developed
Transition
EU
105
b)Gross fixed capital formation, in real terms, by level of development, 2020–2021 (2019=100)
Less developed
More developed
Transition
EU
105
GFCF in real terms, 2008 = 100
GFCF in real terms, 2019 = 100
100
100
95
95
90
90
85
85
80
2009
2010
80
2020
2021
Source: Eurostat, Ameco and Ardeco.
national budgets and to stimulate positive expec- tations for the economy. This collective response appears, so far, to have not only spurred a stronger recovery and mitigated any widening of disparities than after previous recessions but also restrained the increase in public debt.
In sum, the immediate impact of the two reces- sions was deep and broadly similar as regards the macro-economic effects. But the recovery of GDP, employment, household income and investment was stronger and more regionally balanced after the pandemic. The main proximate reason for this
Figure 1.19 Gross fixed capital formation in the EU after the five major recessions since 1980, in real terms, by geographical area, year of recession=100
1980 recession
1992 recession
2001 recession
2008 recession
2019 recession
GFCF at constant prices, recession = 100
110
105
100
95
90
85
80
Source: Eurostat, Ameco and Ardeco.
is that the performance of eastern, and more es- pecially southern, regions was more similar to that of north-western ones. This, in turn, is partly due to the different nature of the two shocks. The 2009 recession stemmed from a global financial crisis, with a severe impact on the banking sector ham- pering the credit channel in the midst of a major de-leveraging process from both the private and the public sector. This, in turn, exerted a prolonged drag on real economic activity, investment, prices
and household income. This was the case through- out the EU, especially as compared with the more robust recovery in the US, and especially in EU re- gions most exposed to the twin de-leveraging pro- cess. By contrast, the 2020 recession was triggered by a different kind of external shock, the spread of a pandemic. The restrictions and disruptions to supply chains that ensued proved more transitory than the 2009 financial crisis. In line with the dif- ferent nature of the two shocks, the price dynamics
Figure 1.20 Gross fixed capital formation, in real terms, by geographical area, 2008=100
Eastern
North-western
Southern
EU
USA
OECD
140
130
GFCF in real terms, 2008=100
120
110
100
90
80
70
60
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
Source: Eurostat, Ameco and Ardeco.
Figure 1.21 General Government consolidated gross debt, by geographical area, 2008–2011 and 2019–2022
a)General Government consolidated gross debt, by geographical area, 2008–2011
Eastern
North-western
Southern
EU
140
b)General Government consolidated gross debt, by geographical area, 2019–2022
Eastern
North-western
Southern
EU
140
120
120
100
100
80
80
60
60
40
40
20
20
0
2008
2009
2010
2011
0
2019
2020
2021
2022
Source: Ameco.
during the recovery phase were also different. In addition, novel and swift policy action – the rapid deployment of Cohesion Policy, new instruments such as SURE (Support to Mitigate Unemployment Risks in an Emergency) and the NextGenerationEU recovery fund – helped to prevent a protracted re- duction in investment. Together, they made available up to EUR 750 billion in financial support to Member States severely affected by the 2020 recession.
The longer-term prospects for economic cohesion, however, remain hard to predict. The addition- al shocks that have occurred since the COVID-19 pandemic pose potentially longer-term challenges to the EU growth model. It is too early to fully as- sess the regional dimension of these shocks, partly because of a lack of regional statistics in many of the areas affected. Several regions, economic sec- tors and categories of workers have suffered sig- nificantly and the current situation remains fragile and volatile, with a risky and uncertain economic outlook. But there are also opportunities. For in- stance, regional economic disparities between the EU-27 and current candidate countries point to a large potential for upward convergence in the
future; see Maps 1.5. and 1.6 comparing the 2004 enlargement with the current relative position of candidate countries vis-à-vis EU regions.
2.The geography of growth, stagnation and discontent: high-growth paths and development traps in Europe
Over the past two decades many regions have experienced a prolonged period of economic stag- nation leading to growing popular discontent. The regions concerned seem to have fallen into a de- velopment trap, a state of sub-par performance of GDP, productivity and employment38. Such a state is empirically correlated with an increase in polit- ical discontent and a decline in support for demo- cratic values and the EU39. Regional development traps are not just an economic concern. The sub- par economic performance and lack of job oppor- tunities have social costs and give rise to political resentment towards what is increasingly regarded as a system that leaves many people behind.
8European Commission (2022).
9Dijkstra et al. (2020, 2023b).
Figure 1.22 General Government net lending (+) or net borrowing (-), excluding interest payments, 2008–2022
Eastern EU
North-western EU
Southern EU
EU
US
Japan
4
2
0
-2
-4
-6
-8
-10
-12
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
Source: Ameco.
Figure 1.23 General Government consolidated gross debt
Eastern EU
North-western EU
Southern EU
EU
US
Japan
300
250
200
150
100
50
0
2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022
Source: Ameco.
On the positive side, though many regions have been persistently trapped, several have succeed- ed in moving from a low-growth to a high-growth development path. This has generally coincided with a shift of specialisation towards more com- plex economic activities linked to local strengths and characteristics, often through integrating into global value chains (see Chapter 5). This section
builds on the concept of a development trap pre- sented in the 8th Cohesion Report40 and extends it in three ways. First, it develops a high-growth path index to identify the best regional performers. Second, it presents a novel approach to determin- ing the characteristics of regions stuck in a devel- opment trap and the ways of escaping from it41. Third, it sets out evidence linking the risk, intensity,
10European Commission (2022).
11Balland et al. (2019).
Box 1.4 Regional cohesion and Russia's war of aggression against Ukraine
Russia's war of aggression against Ukraine sent shockwaves throughout the EU. Some of the EU’s poorer regions are likely to be more affected. This box discusses three reasons: the concentration in richer regions of the economic contribution of work- ing-age refugees; the vulnerability of poorer, rural areas to the sharp increase in energy and food pric- es; and the rise in geopolitical uncertainty, which has pushed up military spending particularly in poorer countries in eastern Europe.
The integration of refugees will probably raise av- erage growth in the EU, but not regional cohesion. Immigration tends to benefit host regions that suc- cessfully integrate refugees in local labour markets. Under the Temporary Protections Directive, Ukraini- an refugees can choose in which EU country to work, and most choose countries with an existing Ukraini- an diaspora and dynamic labour markets: Germany, Poland and Czechia. Working-age Ukrainians added on average 2.5 % to the labour force aged 20–65 in eastern Europe, 1 % in western and northern Europe, and 0.5 % in southern Europe1. Taking into account that language barriers inhibit their integration into labour markets – surveys point to employment rates of about one third – Ukrainian refugees are likely to contribute on average about 0.5 % to the GDP of eastern countries in the short term, and somewhat less in the rest of the EU. The longer these refugees stay, and the better the policies facilitating their in- tegration, the more likely their labour market partic- ipation is to rise. For example, as of August 2022, half of the working-age refugees had found em- ployment in Poland, which currently hosts close to a million Ukrainian refugees, who can benefit from a particularly large existing diaspora and relatively low language barriers.
Even though eastern countries’ living standards tend to lie below the EU average, it is mostly the richer regions that are likely to benefit from their integra- tion into local labour markets. Refugees tend to set- tle in the dynamic regions with better employment
prospects within those countries, such as Prague or Warsaw, whose GDP per capita already substantially exceeds the EU average.
The energy and food price shocks triggered by the war have lowered wealth throughout the EU, but poorer, rural areas were more affected. Prices for energy and food have declined from their peaks, but have had a significant impact on real disposable in- come. Since rural regions within the EU tend to be poorer than urban ones, households living in rural areas tend to spend relatively more on transport, and those that are poorer spend relatively more on energy and food. For example, households in rural areas in Bulgaria spend 35 % of their consumption on food, those in Bulgarian cities 23 %.
Finally, eastern countries bordering Russia, Ukraine or Belarus have raised their military spending more than other Member States since Russia’s invasion of Crimea. With a GDP per head about half that of countries in the north and west, these countries raised their military spending by 0.7 % of GDP be- tween 2014 and 2022, twice as much as those in the west and north. This increase risks crowding out spending that could have been used to advance regional cohesion. Being more intertwined with the Russian economy before the war, these economies are more affected by the sanctions imposed on Russia. The war has been a major disruption to the implementation of cohesion programmes, notably Interreg programmes. External border regions, in Finland and the Baltic States, as well as some Polish border regions, have lost their cross-border co-oper- ation partners. Previous exchanges and cross-border flows have been replaced by closed borders and no co-operation. The Commission introduced changes allowing for the integration of these regions into other co-operation programmes, but the negative border effect is stronger than ever and they must be further supported to look for other co-operation and development opportunities.
1All figures referenced in this box stem from Eurostat as well as various reports from the International Organization for Migration
(https://dtm.iom.int/reports?search=ukraine).
and length of regional development traps to the rise of political discontent in the EU42.
2.1Regions on high-growth trajectories
The picture of convergence shown by the indicators above gives an overall view of macro-regional de- velopments, but it does not lend itself to identify- ing specific features and success stories at a more detailed level. To shed light on these, the meth- odology used to determine the regions stuck in a development trap also enables us to calculate an economic development index (EDI) for regions that have persistently outperformed others43. A large number of EU regions, defined here at the NUTS 3 level, have been on a high-growth trajectory (EDI above 0.5 in Map 1.7) over the past two decades. As expected, these are disproportionally located in eastern Europe, reflecting higher growth during the catching-up phase noted above (beta conver- gence). However, regional success stories are not limited to this broad area of the EU. Indeed, most EU Member States have at least one NUTS 3 region on a high-growth path over the period 2001–2021 (EDI higher than 0.5). This is true not only of most capital city regions, but also of some regions in centre-north Portugal, north-western Spain, coastal France and, to a lesser extent, Italy and Greece, as well as some more developed regions in Germany, Belgium, the Netherlands and Sweden. Overall, this confirms that economic performance has varied substantially across the EU and within countries44.
2.2Regions in a development trap
A novel approach to determining the character- istics of regions in a development trap has shed light on possible links with a new typology of eco- nomic complexity traps45. In addition to the stand- ard characteristics of regions in a development trap46, self-reinforcing dynamics could limit the capacity of regions to innovate and develop new growth paths47. Regions might become trapped in low-complexity activities because of a lack of capability to develop highly complex products48. An analysis of the structural evolution of develop- ment traps over a long period of time has provided systematic empirical evidence on how many re- gions in the EU fail to overcome a ‘low-complexity’ structure, on the extent to which these are high- or low-income regions, and the kinds of traps they have fallen into. The definition of ‘evolutionary traps’ centres around the structural inability of re- gions to develop new activities, because their ca- pabilities prevent them from moving into new and more complex activities that could increase their prosperity. Based on this, it identifies regions that once performed well but have become trapped, as well as those that have managed to escape from being so and how.
The characteristics of regions in a development trap are highly varied in terms of development levels, but the limited capacity of a region to edu- cate people and retain them is a common feature across all levels of development. The reasons for falling into a development trap differ between re- gions depending on the initial level of development,
12Dijkstra et al. (2023b).
13Using the methodology to measure the likelihood of being in a development trap developed by Iammarino et al. (2020), high-growth paths are identified when regions have outperformed their peers in terms of GDP, productivity and employment growth (when the likelihood of so doing is greater than 50 %). The conventional development trap indicator denotes when a region’s growth of GDP per head, productivity and employment is lower than that of the EU, its country, or the region itself over the previous five years. A region scores 1 for each time its growth is higher than the three benchmarks. The score between 0 and 9 is then rescaled to 0 and 1. To identify regions on high-growth paths, the inverse of the average yearly development trap score of each region is taken over the period 2001–2021. This ensures consis- tency and symmetry with the analysis based on the development trap indicator, while pointing to regions outperforming their peers.
14In eastern Member States, economic performance has been strong in capital regions but also across the majority of other regions. In south- ern Europe, regions outperforming their peers are mostly located in Spain and Portugal – cases of catching up again because they were relatively poor regions – but there are positive examples also in Greece and Italy. Coastal regions in France have also generally performed much better than central ones (except for the capital city region). In the rest of Europe, there is a broadly balanced presence of regions in terms of their economic performance.
15Balland et al. (2019).
16Iammarino et al. (2022).
17Arthur (1994).
18Pinheiro et al. (2022)
EUROPEAN COMMISSION
Brussels, 27.3.2024
SWD(2024) 79 final
COMMISSION STAFF WORKING DOCUMENT
[…]
Accompanying the document
Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions
on the 9th Cohesion Report
{COM(2024) 149 final}
Canarias
Guadeloupe Martinique
Guyane
Mayotte Réunion
REGIOgis
Map 1.7 Economic development index at NUTS 3 level, 2001–2021
Likelihood of being in a high-growth trajectory
< 0.3
0.3 – 0.4
0.4 – 0.5
0.5 – 0.6
> 0.6
no data
This index measures if a region's growth is higher than that of the EU, of its country, or of the region itself during the previous five years.
It considers growth of GDP per head, productivity, and employment per head over a five-year period.
A region scores 1 for each time its growth is higher. This score between 0 and 9 is then rescaled to 0 and 1.
Source: DG REGIO calculations based on JRC and Eurostat data.
0
500 km
© EuroGeographics Association for the administrative boundaries
Table 1.3 Socio-economic characteristics of ‘development-trapped’ and other regions, average 2003–2021, by level of GDP per head, 2003
|
Development trapped?
|
GDP/head (PPS) in 2003, index EU-27 = 100
|
|
|
< 75 %
|
75 - 100 %
|
>= 100 %
|
All
|
% of industry in GVA
|
Yes
|
21.5
|
14.8
|
18.8
|
18.1
|
|
No
|
26.3
|
18.1
|
20.9
|
21.0
|
R&D expenditure as % of GDP
|
Yes
|
0.4
|
1.2
|
2.0
|
1.8
|
|
No
|
0.9
|
1.5
|
2.5
|
2.1
|
% of population 25–64 with tertiary education
|
Yes
|
12.1
|
20.2
|
27.0
|
23.9
|
|
No
|
20.9
|
27.7
|
30.9
|
27.2
|
Institutional quality index
|
Yes
|
-1.6
|
-0.5
|
0.3
|
-0.1
|
|
No
|
-0.8
|
0.1
|
0.6
|
0.1
|
% of population (2021) by GDP/head level
|
|
23.3
|
22.5
|
54.2
|
100.0
|
% of population (2021) in trapped regions
|
|
2.4
|
7.3
|
18.6
|
28.4
|
Note: Socio-economic characteristics are average values of all available reference years in period 2003–2021. Source: Eurostat [rd_e_gerdreg, lfst_r_lfsd2pop], JRC (ARDECO), University of Gothenburg, DG REGIO calculations.
geographical features, the macro-economic envi- ronment, the global economic context and struc- tural characteristics. However, there are a num- ber of common traits in terms of the quality of institutions, innovation capacity and importance of manufacturing that vary between trapped and non-trapped regions to differing degrees depend- ing on the level of development. As indicated in the previous section, geographical characteristics, sec- toral specialisation, productivity and investment dynamics affect beta or ‘club’ convergence. How- ever, one common feature of persistently trapped regions at all levels of economic development is lack of human capital (Table 1.3).
This suggests that having in place the conditions and opportunities for investing, attracting and re- taining people with tertiary education is a consist- ent feature of regions that have managed not to fall into a development trap for a large number of years and can reduce the likelihood of becoming trapped (see Chapter 6)49. Past performance is no guarantee of future performance. And not all re- gions can have a large share of tertiary-educated
workers, but – at any level of development – a peo- ple-centred differentiated place-based approach in line with the potential and characteristics of the region may reduce the likelihood of experiencing a persistent period of stagnation (see Chapter 5).
1.1Regions in a development trap and the geography of discontent
Regional development traps are not just an eco- nomic matter. Sub-par economic performance and lack of employment opportunities give rise to so- cial costs and can cause political resentment to- wards what is increasingly regarded as a system that leaves people behind, leading to a growing geography of discontent50. An econometric anal- ysis of the link between the risk, intensity and length of regional development traps and the rise of discontent in the EU, proxied by the support for Eurosceptic parties in national elections between 2014 and 2022, found a strong connection be- tween being stuck in a development trap and sup- port for Eurosceptic parties51. It also found that the longer the period of stagnation, the stronger the
1This is also the case for regions in a ‘talent development trap’, a composite indicator related to the development trap but in the demographic domain. European Commission (2023) shows that 46 regions in the EU with over 70 million inhabitants are in a talent development trap. These regions had an accelerating decline of their working-age population, and a low and unchanging number of people with tertiary edu- cation between 2015 and 2020. It also identifies a second group of 36 regions (with nearly 60 million inhabitants) that are at risk of falling into a talent development trap in the future, because they are strongly affected by the outward movement of people aged 15–39. This group accounts for 13 % of the EU population.
2See Dijkstra et al. (2021 and 2023), who show that political discontent with the EU in Member States and regions is linked to an important extent to economic and industrial decline and being in a development trap.
3Dijkstra et al. (2023b).
Box 1.5 The geography of EU discontent and the regional development trap
In recent years, popular discontent has been brewing in many parts of the world, including in many countries in Europe1. This rising wave of dissatisfaction with a ‘system’ that many feel no longer benefits them is manifested in differ- ent ways, from declining levels of participation in elections to low levels of engagement in civil so- ciety. The dissatisfaction can also be seen in a growing tendency to support more extreme, often
populist, options at the ballot box; and in increas- ing signs of distress and outright revolt by those dis- affected by the system2. In the EU, this disaffection is reflected in the rise of Euroscepticism3. Since the 2008 financial crisis, the share of votes in national legislative elections for ‘hard’ Eurosceptic4 parties has risen from under 5 % to 14 % in 2022, and for all Eurosceptic parties from 7 % to 27 %.
Map 1.8 Development trap index 1 at NUTS-3 level, 2001–2018
Likelihood of being in a development trap
< 0.4
0.4 – 0.5
0.5 – 0.6
0.6 – 0.7
> 0.7
This index measures if a region’s growth is lower than that of the EU, of its country or of the same region during the previous 5 years.
It considers growth in GDP per head, productivity and employment over a five-year period.
A region scores 1 for each time its growth is lower. This score between 0 and 9 is then rescaled to 0–1.
Source: DG REGIO calculations based on JRC and Eurostat data.
© EuroGeographics Association for the administrative boundaries
1Greven (2016); Zakaria (2016); Hawkins et al. (2019); Hopkin (2020).
2Rodríguez-Pose (2018); Kitschelt (2022).
3Torreblanca and Leonard (2013); Dijkstra et al. (2020).
4Eurosceptic parties are defined based on the Chapel Hill Expert Survey.
Figure 1.24 Correlation between being development trapped and the hard Eurosceptic vote for NUTS 3 regions, 2018–2022
60
Minimum share of valid votes in national parliamentary election, in %
50
40
30
20
10
0
-10
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Development trap index 1
Note: Bubble area size reflects population in 2021. Source: Dijkstra et al. (2023b).
The rise of Euroscepticism is not an isolated phe- nomenon. It is instead part of a broader recent in- crease in the popularity of anti-system, or populist, parties5. Explanations can be classified as cultural or economic, or both6. People living in places in de- cline frequently feel trapped in regions they think no longer matter and where they perceive they have no future7. They feel ignored, neglected and marginal- ised by a distant and aloof elite8, and are ill at ease with a changing world that threatens their identity and security.
A study9 finds that much of the rise in discontent is concentrated in places that have been in a devel- opment trap10. The classic example of a region in a development trap is one that initially experienced a spurt in growth allowing it to attain middle-income
levels, but subsequently got stuck without manag- ing to reach high income levels11. However, many regions in Europe have stagnated – and even de- clined – at all levels of development. The risk of be- coming stuck in a development trap is higher in mid- dle-income regions, but can occur in all regions. The same study finds that falling into a development trap is a major factor in understanding why Euro- sceptic voting in national elections has been on the rise across EU regions. People living in regions in a development trap are far more likely to be tempted by Eurosceptic political parties and to support them in elections. The authors also show that factors such as the risk, intensity and length of time spent in a development trap significantly increase the share of the Eurosceptic vote.
5 Hopkin (2020).
6Noury and Roland (2020); Schmid (2022).
7Rodríguez-Pose (2018 and 2020); Lenzi and Perucca (2021).
8McKay et al. (2021).
9Dijkstra et al. (2023).
10The methodology to calculate the development trap is the same as that used in European Commission (2022).
11Kharas and Kohli (2011).
support for parties opposing European integration. Since development traps can occur at different lev- els of development, but appear to be a particular risk for transition regions, they may require policy responses that go beyond support for less devel- oped regions. Assisting all regions that are devel- opment-trapped to become more dynamic should help to reduce regional inequalities and counter the threat of rising discontent in EU societies.
2.Economic cohesion and competitiveness to harness the benefits the Single Market
The productivity dynamics examined above are reflected in a broader measure of sub-national performance, the RCI. This is a composite indica- tor designed to capture the 11 main dimensions of competitiveness of EU NUTS 2 regions: insti- tutions; macro-economic stability; infrastructure; health; basic education; higher education; training and lifelong learning; labour-market efficiency; market size; technological readiness; business so- phistication; and innovation52. The 2022 RCI shows a polycentric pattern, with strong performance of regions with large urban areas, which benefit from agglomeration economies, better connectivity and higher levels of human capital. The index is above the EU average in all regions in Austria, the Benelux countries, Germany and the Nordic Member States. (Map 1.9, left panel). By contrast, all eastern re- gions, except most capital city ones, score below the EU average. Southern regions also score below the average, except for Cataluña, Madrid and País Vasco in Spain, Lombardia in Italy and Lisboa in Portugal. Ireland and, especially, France have a mix of regions above and below the EU average.
Less developed regions, however, have improved markedly over time. In the six years since the indi- cator was first developed in 2016, there has been a clear process of catching up in eastern regions combined with an improvement in southern ones, as they recovered from the economic and financial crisis (Map 1.9, right panel).
Between 2019 and 2022, the RCI improved by 10 index points or more in the capital city region in Lithuania (+20 points), Norte in Portugal (+14), the capital city region in Poland (+13), the Portu- guese outermost region of Madeira (+13), and Illes Baleares in Spain and Śląskie in Poland (both +10).
Within Member States, capital city regions tend to be the most competitive ones. The gap between the capital city region and the others is particularly wide in France, Spain, Portugal and many of the eastern EU Member States. This can be a reason for concern as it increases pressure on resources in the capital city region while possibly leaving them under-utilised elsewhere. In three countries, how- ever, the Netherlands, Italy and Germany, the cap- ital city region is not the most competitive. In the Netherlands, Utrecht remains the best-performing region (at 151, the EU average being 100), fol- lowed by Zuid-Holland which includes Rotterdam and The Hague (at 142). In Italy, Lombardia, which includes Milan, continues to be the best-perform- ing Italian region (at 103), while in Germany this remains Oberbayern, which includes Munich (at 130), and several other regions also outperform Berlin and Brandenburg.
4See Dijkstra et al. (2023a).
Ninth Report on economic, social and territorial cohesion
Map 1.9 RCI: latest values (2022) and change since the first edition in 2016
Box 1.6 Competitiveness, the EU Single Market and Cohesion Policy
The Single Market is a cornerstone of EU integra- tion and competitiveness and goes hand in hand with Cohesion Policy. Removing barriers to the free movement of goods, services, capital and workers has promoted a better allocation of resources across the EU and fostered the exchange of ideas and inno- vation. However, market forces alone do not ensure that everyone benefits from economic integration. In fact, this report highlights significant territorial dis- parities linked to the different levels of development of countries and regions, their specific geographical features and their economic structure. These dispar- ities, though tending to diminish, translate into dif- ferent levels of competitiveness – as captured, for instance, by the RCI – which in turn may lead to frag- mentation within the Single Market. Left alone, the free mobility of labour and capital in the context of uneven levels of competitiveness risks damaging co- hesion. Cohesion Policy, along with other policies, no- tably State-aid rules, helps to create a level playing field essential for the Single Market to function fairly, while supporting less developed regions to develop.
By investing in infrastructure, innovation, education and other key areas, Cohesion Policy helps less de- veloped regions directly and all other regions indi- rectly to reap the benefits of the Single Market. The latter occurs because of spill-over and scale effects linked to the policy and the Single Market1. A more competitive and integrated Single Market gives busi- nesses access to a larger customer base and enables economies of scale to be realised. The proper func- tioning of the Single Market, however, requires that producers and consumers throughout Europe have equal access to it, so that it can ensure the effective matching of supply and demand and the efficient al- location of resources across the EU as a whole, in the long as well as the short term. But access cannot be taken for granted – thus need to support investment where access is limited, especially in the less com- petitive and less developed regions.
1Crucitti et al. 2023.
References
Alvaredo, F., Atkinson, A., Chancel, L., Piketty, T., Saez, E. and Zucman, G. (2016), Distributional National Accounts (DINA) guidelines: Concepts and methods used in WID. world.
Arvanitopoulos, T. and Lazarou, N. (2023), The determinants and dynamics of regional convergence in the EU, JRC Working Papers on Territorial Modelling and Analysis, 06/2023, European Commission, JRC133733.
Barro, R. J. (2001), ‘Human Capital and Growth’, American Economic Review, 91(2), pp. 12–17.
Barro, R.J. and Sala–i–Martin, X. (1992), Convergence, Journal of Political Economy, 100(2), pp. 223– 251.
Bauluz, L., Bukowski, P., Fransham, M., Lee, A., Lopez Forero, M., Novokmet, F., Breau, S., Lee, N., Malgouyres, C., Schularick, M. and Verdugo, G. (2023), ‘Spatial wage inequality in North America and Western Europe: changes between and within local labour markets 1975–2019’, CEPR Discussion Paper No 18381.
Blanchard, O.J. and Leigh, D. (2013), ‘Growth Forecast Errors and Fiscal Multipliers’, American Economic Review, 103 (3), pp. 117–120.
Bökemeier, B. and Wolski, M. (2022), ‘This Time is Different: Fiscal Response to the Covid–19 Pandemic
among EU Countries’, International Economics, 172, pp. 217–226.
Brynjolfsson, E. and McAfee, A. (2014), The Second Machine Age: Work, Progress, and Prosperity in a Time of Brilliant Technologies, W. W. Norton & Company.
Cavallaro, E. and Villani, I. (2021), ‘Club Convergence in EU Countries: A Sectoral Perspective’, Journal of Economic Integration, 36(1), pp. 125–161.
Cervellati, M. and Sunde, U. (2013), ‘Life Expectancy, Schooling, and Lifetime Labor Supply: Theory and Evidence Revisited’, Econometrica, 81(5), pp. 2055–86.
Crucitti, F., Lazarou, N.J., Monfort, P. and Salotti, S. (2023), ‘Where does the EU cohesion policy produce its benefits? A model analysis of the international spillovers generated by the policy’, Economic Systems, 47(3), p. 101076.
Diemer, A., Iammarino, S., Rodríguez–Pose, A. and Storper M. (2022), ‘The Regional Development Trap in Europe’, Economic Geography, 98(5), pp. 487–509.
Dijkstra, L., Papadimitriou, E., Cabeza Martínez, B., De Dominicis, L. and Kovacic, M. (2023a),
‘EU Regional Competitiveness Index 2.0 2022 edition’, European Commission, DG Regional and Urban
Policy, Working Paper No 01/2023.
Dijkstra L., Poelman, H. and Rodríguez–Pose, A. (2020), ‘The geography of discontent’, Regional Studies, 54(6), pp. 737–753.
Dijkstra, L., Poelman, H. and Rodríguez–Pose, A. (2023b), The geography of EU discontent and the regional development trap in Europe, DG Regional and Urban Policy, Working Paper 03/2023, European Commission, Brussels.
Eichengreen, B. (2023), The European Economy since 1945: Coordinated Capitalism and Beyond, Princeton University Press.
European Commission (2017), Reflection Paper on Harnessing Globalisation, Publications Office
of the European Union, Luxembourg.
European Commission (2020), ‘Chapter 3: Inclusive growth and solidarity in the EU: challenges, policy levers and the way forward’, Employment and Social Developments in Europe, Publications Office
of the European Union, Luxembourg.
European Commission (2020), Science, research and innovation performance of the EU: A fair green and digital Europe, Publications Office of the European Union, Luxembourg.
European Commission (2022a), 8th Cohesion Report on Economic, Social and Territorial Cohesion,
Publications Office of the European Union, Luxembourg.
European Commission (2022b), Labour market and wage developments in Europe – Annual review 2022, Publications Office of the European Union, Luxembourg.
European Commission (2023), Regional Trends for Growth and Convergence in the European Union,
SWD(2023) 173 final.
Giupponi, G., Landais, C. and Lapeyre, A. (2022), ‘Should We Insure Workers or Jobs during Recessions?’,
Journal of Economic Perspectives, 36(2), pp. 29–54.
Gordon, R. J. (2015), ‘Secular Stagnation: A Supply–Side View’, American Economic Review, 105(5),
pp. 54–59.
Gordon, R. J. and Sayed, H. (2019), ‘The Industry Anatomy of the Transatlantic Productivity Growth Slowdown: Europe Chasing the American Frontier’, International Productivity Monitor, 37, pp. 3–38.
Greven, T. (2016), The rise of right–wing populism in Europe and the United States: A Comparative Perspective, Friedrich Ebert Foundation, Washington DC, pp. 1–8.
Guzzo, F. and Gianelle, C. (2021), Assessing Smart Specialisation: Governance, Publications Office
of the European Union, Luxembourg.
Hawkins, K. A., Aguilar, R., Silva, B. C., Jenne, E. K., Kocijan, B. and Rovira Kaltwasser, C. (2019), Measuring populist discourse: The global populism database, EPSA Annual Conference in Belfast, UK, June.
Head, K. and Mayer, T. (2021), ‘The United States of Europe: A Gravity Model Evaluation of the Four Freedoms’, The Journal of Economic Perspectives, 35(2), pp. 23–48.
Hegyi, F.B., Guzzo, F., Perianez–Forte, I. and Gianelle C. (2021), The Smart Specialisation Policy Experience: Perspective of National and Regional Authorities, Publications Office of the European Union, Luxembourg.
Hopkin, J. (2020), Anti–system politics: The crisis of market liberalism in rich democracies, Oxford University Press, Oxford.
Iammarino, S., Rodriguez–Pose, A. and Storper, M. (2017), Why Regional Development Matters for Europe’s Economic Future, DG Regional and Urban Policy, Working Paper 07/2017, European Commission.
Iammarino, S., Rodriguez–Pose, A. and Storper, M. (2020), Falling into the Middle–Income Trap? A Study on the Risks for EU Regions to be Caught in a Middle–Income Trap, Publications Office of the European Union, Luxembourg.
Kitschelt, H. (2002), ‘Popular Dissatisfaction with Democracy: Populism and Party Systems’. In: Mény, Y. and Surel, Y. (eds), Democracies and the Populist Challenge, Palgrave Macmillan, London.
Königs, S. et al. (forthcoming), The geography of income inequalities in OECD countries: Evidence from national register data, OECD Publishing, Paris.
Krugman, P. (1991), ‘Increasing returns and economic geography’, Journal of political economy, 99(3),
pp. 483–499.
Lee, N., Rodriguez–Pose, A. and Terrero–Davila, J. (2023), ‘Left–behind vs. unequal places: interpersonal inequality, economic decline, and the rise of populism in the US and Europe’, Papers in Evolutionary Economic Geography (PEEG) 2306, Utrecht University, Department of Human Geography and Spatial Planning, Group Economic Geography, revised March 2023.
Lenzi, C. and Perucca, G. (2021), ‘People or places that don’t matter? Individual and contextual determinants of the geography of discontent’, Economic Geography, 97(5), pp. 415–445.
Licchetta, M. and Mattozzi, G. (2022), ‘Convergence in GDP per capita in the euro area and the EU at the time of COVID–19’, Quarterly Report of the Euro Area (QREA), 21(3), pp. 7–18.
Maddison, A. (2007), Contours of the World Economy, 1–2030 AD: Essays in Macro–Economic History, Oxford University Press, Oxford.
McKay, L., Jennings, W. and Stoker, G. (2021), ‘Political Trust in the Places That Don’t Matter’, Frontiers in Political Science, 3, 642236.
Monfort, P. (2020), Convergence of EU regions redux: Recent trends in regional disparities, WP 02/2020, Publications Office of the European Union, Luxembourg.
Noury, A. and Roland, G. (2020), ‘Identity politics and populism in Europe’, Annual Review of Political Science, 23, pp. 421–439.
OECD (2022), OECD Regions and Cities at a Glance 2022, OECD Publishing, Paris.
OECD (2023), OECD Regional Outlook 2023: The Longstanding Geography of Inequalities, OECD Publishing, Paris.
Pesaran, M. H. (2007), ‘A pair–wise approach to testing for output and growth convergence’,
Journal of Econometrics, 138(1), pp. 312–355.
Prognos and CSIL (2021, forthcoming), Study on Prioritisation in Smart Specialisation Strategies in the EU.
Quah, D. (1996), ‘Convergence Empirics Across Economies with (Some) Capital Mobility’, Journal of Economic Growth, 1(1), pp. 95–124.
Rodríguez–Pose, A. (2018), ‘The revenge of the places that don’t matter (and what to do about it)’,
Cambridge Journal of Regions, Economy and Society, 11(1), pp. 189–209.
Schmidt, V. A. (2022), ‘The discursive construction of discontent: varieties of populist anti–system ideas and discursive interactions in Europe’, Journal of European Integration, 44(2), pp. 165–182.
Torreblanca, J. I. and Leonard, M. (2013), The continent–wide rise of Euroscepticism, European Council on Foreign Relations.
Zakaria, F. (2016), ‘Populism on the March: Why the West is in Trouble’, Foreign Affairs, 95(9).
EUROPEAN COMMISSION
Brussels, 27.3.2024
SWD(2024) 79 final
COMMISSION STAFF WORKING DOCUMENT
[…]
Accompanying the document
Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions
on the 9th Cohesion Report
{COM(2024) 149 final}
SOCIAL COHESION
•EU labour markets have shown resilience in the face of the COVID-19 pandemic and Russian aggression towards Ukraine. With both national government and EU support, employment in most regions rebounded from the reduction in 2020 in just one year. In 2022, the employment rate of those aged 20 to 64 in the EU reached a record high of nearly 75 %.
•Nevertheless, challenges persist and need to be addressed. Despite a reduction in regional disparities, labour markets remain more robust and social conditions better in north-western EU regions than in southern and eastern ones.
•Increased labour market participation of under-represented groups played a key role in reducing employment disparities and tackling labour shortages. The em- ployment rate of women in the EU increased from 61 % in 2013 to 69 % in 2022, helped by improved access to childcare and long-term care and more flexible working arrangements. Nevertheless, the employment gap between men and women still averaged 11 pp in 2022 in the EU and 15 pp in southern Mem- ber States.
•Labour and skill shortages pose potential challenges to cohesion. Recent com- munications from the Commission highlight the need to tackle these shortages. This has become crucial to ensuring that all individuals are equipped with the right skills to take up opportunities and tackle the challenges the green and dig- ital transitions bring about in such a way that no-one is left behind.
•There has been a continuing increase in education levels across all regions, with the tertiary rate in the EU for those aged 25 to 64 reaching 34 % in 2022. But regional disparities persist, notably because of a concentration of graduates in large cities, and rates remain higher in more developed and transition regions (36–38 %) than in less developed ones (26 %).
•The at-risk-of-poverty-or-social-exclusion rate declined from 35 % to 28 % in less developed regions between 2013 and 2019, while it remained unchanged at 19 % in more developed regions. Some 95 million Europeans were still affected in 2022 and achieving the 2030 goal of reducing the number by at least 15 mil- lion may prove difficult if stagnation persists.
Chapter 2
Social cohesion
1.Introduction
This chapter examines progress towards a more social EU. It focuses on cohesion across the main areas covered by the European Pillar of Social Rights action plan, namely employment, skill de- velopment, and poverty reduction (Box 2.1). A sep- arate section considers gender equality and equal opportunities and attitudes towards migrants and other minorities.
The analysis indicates that while the EU is advanc- ing towards a more inclusive and fairer society, in some areas progress has stalled. Labour markets have shown resilience and regional disparities in employment have narrowed. Increased labour mar- ket participation of under-represented groups has been important in furthering convergence and re- ducing labour shortages. There has been a gener- al increase in education levels and participation in adult education and training, especially in less de- veloped regions. However, disparities persist, nota- bly because of a marked concentration of graduates in large cities. A tendency for the at-risk-of-pover- ty-or-social-exclusion (AROPE) rate to decline till 2019 was evident especially in eastern EU regions and rural areas in the southern EU. Nevertheless, some 95 million Europeans remain AROPE, includ- ing 20 million children and people in disadvantaged situations, such as people with disabilities.
Any analysis of labour market and social develop- ments in the EU needs to start from one dimension of change in particular, the shrinking population of working age, which is projected to be some 7 % smaller by 2040, a reduction of 15 million. This has a potential macro-economic impact and affects regions and cities differentially. It emphasises the importance of increases in labour productivity for growth, closely tied to education attainment lev-
els and the skills needed by the labour market. In addition, while capital accumulation was a major driver for growth up to the 1990s, now ideas or innovation that lead to new services and prod- ucts have become more important. Education and training together with creativity are pivotal in this evolving landscape, especially with regard to the skills needed to support workers and businesses in the context of the green and digital transitions.
Labour shortages linked to a limited supply of certain skills, poor working conditions and human resource management, the ageing of the work- force and gender segregation, together with skill shortages and mismatches, continue to hold back growth, competitiveness and cohesion.
2.Impact of COVID-19 and post-COVID years on social situation in the EU
EU labour markets remained resilient in the af- termath of the COVID-19 pandemic, despite the uncertainty created by the Russian war of aggres- sion against Ukraine and significant inflationary pressures. Overall, more people than ever are em- ployed in the EU, and fewer people are unemployed or looking to work longer hours.
The upward trend in employment from 2013 to 2019 resumed after a dip (of 1 pp) in 2020 when the COVID-19 pandemic hit. The employment rate of those aged 20 to 64 reached 74.6 % in 2022, 1.9 pp higher than in 20191, while over- all unemployment of those aged 15 to 74 went down to 6.2 % in 2022 from 7.2 % in 2020. The response of regional labour markets during the pandemic and the subsequent recovery saw nar- rowing differences in employment rates between
1In 2021, due to the introduction of new legislation, there was a break in the EU labour force survey (LFS) time series, which involved, among other revisions, a change in the definition of employment. Selected series of main indicators were retroactively corrected for the break. However, regional series were not included in these adjustments. For this report, regional employment rates from 2008 to 2020 are extrap- olated to be consistent with the country-level break-corrected time series.
Box 2.1 European Pillar of Social Rights and its action plan
The European Pillar of Social Rights was proclaimed by the European Parliament, the Council and the European Commission at the Social Summit for fair jobs and growth in Gothenburg on 17 Novem- ber 2017. Then the President-elect of the European Commission, Ursula von der Leyen, committed to the Pillar in her speech before the European Parlia- ment in Strasbourg in July 2019 and in her political guidelines for the mandate of the next European Commission, announcing further action to imple- ment the associated principles and rights.
The Pillar sets out key principles and rights to sup- port fair and well functioning labour markets and welfare systems. It supports the convergence to- wards better working and living conditions among participating Member States. The principles are grouped into three broad categories:
•equal opportunities and access to the labour market, which includes equal access to edu- cation and training, gender equality and active support for employment;
•fair working conditions, namely the right to se- cure and adaptable employment, fair wages, in- formation on working conditions and protection in case of dismissal, consultation with social partners, support in achieving a suitable work- life balance, and a healthy and safe working environment;
•social protection and inclusion, which includes access to childcare and support for children’s ed- ucation, unemployment benefits and access to activation measures, minimum-income support, old-age pensions, affordable healthcare, support for people with disabilities, affordable long-term care, housing and assistance for the homeless and access to essential services.
The Pillar reaffirms rights already present in the EU but complements them by taking account of new realities arising from societal, technological and economic developments. As such, it does not affect
principles and rights already contained in the bind- ing provisions of EU legislation. By putting together rights and principles set at different times, in dif- ferent ways and in different forms, it aims to make them more visible, understandable and explicit. On 4 March 2021, the European Commission adopt- ed the European Pillar of Social Rights action plan1, and proposed three headline targets for the EU to reach by 2030, welcomed by EU leaders at the Porto Social Summit in May 2021 and at the European Council of June 2021:
1.at least 78 % of the population aged 20 to 64 to be in employment, supported by halving the gender employment gap;
2.at least 60 % of all adults aged 25 to 64 to participate in training every year; and
3.a reduction of at least 15 million in the num- ber of people identified as AROPE, including at least 5 million children.
Member States have set national targets for each of the targets, and progress towards both the EU-level and national targets is monitored through the Euro- pean Semester.
The action plan establishes principles and rights to foster a fairer and more just society within the EU. It encompasses initiatives to combat poverty and social exclusion, which include increasing the ade- quacy and coverage of minimum wage protection, support for social benefits, policies aiming at labour market activation, active inclusion for minimum in- come recipients, adequate social protection, long- term care and pensions, the child guarantee and investment in education and training.
The action plan also includes a proposal for a revised social scoreboard, to track progress towards the Pil- lar principles more comprehensively. The yearly joint employment report2 provides regional breakdowns (at NUTS 2 level) of the social scoreboard headline indicators for which data are available.
1European Commission (2021b).
2European Commission (2023h).
more and less developed regions and between north-western and eastern and southern Member States. Given the exogenous nature of the shock and with support from national and EU measures, it took just one year, after the decline in 2020, for the employment rate in nearly all regions to return to, or surpass, the 2019 level. By contrast, during the previous economic crisis, reductions in employ- ment, which began in 2009, persisted until 2013, and the employment rate returned to pre-crisis levels only by 2015–2017 and only by 2019 in southern countries.
After a small fall (of 0.8 pp) in 2020, the pro- portion of women in employment continued to expand, helped by improved access to childcare, more flexible working arrangements and increas- ing education levels. Despite this, progress in clos- ing the gender employment gap has slowed down in recent years in most regions (except those in eastern countries) and in the EU as a whole still stood at 11 pp in 2022. The employment rate of migrants (i.e. those born outside the EU), after a significant fall (of 2.5 pp) in 2020, increased fast- er than for other groups between 2020 and 2022 (by 4.0 pp), confirming their adaptability to chang- ing economic conditions and their contribution to meeting labour shortages in particular sectors and regions.
The positive trend in tertiary education continued across all regions during the pandemic. The pro- portion of people aged 25 to 64 with tertiary edu- cation in the EU even increased in 2020 (by 1.2 pp), reaching 34.3 % in 2022. By contrast, adult partic- ipation in education and training (in the previous four weeks) decreased (by 1.7 pp) when COVID-19 hit, but rebounded the following year, especial- ly in less developed regions and eastern Member States. Almost 12 % of those aged 25–64 partici-
pated in education and training (in the four weeks preceding the survey) in the EU in 2022, 1.1 pp more than in 20192.
After two decades of low inflation, the COVID-19 pandemic was followed by a surge inflation as reduced supply chains struggled to keep up with increasing demand and as the Russian war in Ukraine in early 2022 reduced energy and food supplies3. As a result, inflationary pressures accen- tuated concerns about the effects on lower-income households that spend a larger share of their in- come on energy, food and transport, on which price increases were especially large4. Accordingly, the proportion of households reporting financial dis- tress increased from 12.5 % in December 2021 to
15.8 % in December 20225.
The proportion of the population experiencing se- vere material and social deprivation (see Box 2.4 for the definition) increased marginally in the EU from 6.3 % in 2021 to 6.7 % in 2022, but by more (by 1.2 pp) in Latvia, Estonia, Romania, Germany and France. There were also large increases (from 6.8 % in 2019 to 8.3 % in 2022) in those reporting an inability to afford a decent meal (with meat, chicken, fish or a vegetarian equivalent) every sec- ond day and an inability to keep their home ade- quately warm (from 6.9 % to 9.3 %) – an indicator of energy poverty reversing the reduction between 2016 and 2019.
Overall, perhaps partly as a result of the policy re- sponses at EU and Member State level, the AROPE rate, which declined consistently between 2016 and 2019 in most types of regions, has remained unchanged since 2019. Also in 2022, relative pov- erty and income inequality, as measured by the ratio of the income of the top 20 % of households to that of the bottom 20 %, remained unchanged6.
2Note that the EU target of achieving at least 60 % of adults participating in training each year by 2030 is based on a different indicator,
covering the last 12 months rather than just the previous four weeks.
3European Commission (2023a) and Fulvimari et al. (2023).
4OECD (2023).
5European Commission (2023a). The financial distress indicator is based on the business and consumer survey and is composed of the share
of adults reporting the need to draw on savings and the share of adults reporting the need to run into debt.
6Eurostat’s flash estimate for 2022. The EU-SILC (EU statistics on income and living conditions) AROPE and at-risk-of-poverty (AROP) rates for year N are based on the accrual income from the previous year, N-1. Eurostat’s flash estimates complement EU-SILC indicators with estimates for the latest income changes and are based on modelling and micro-simulation techniques that consider the interaction between labour market developments, economic and monetary policies, and the implementation of social reforms for income year N.
3.Labour market developments
The EU is well on track to meeting its headline tar- get of at least 78 % of people aged 20–64 being in employment by 20307 (Box 2.2). Overall, the rate increased by around 8 pp from the end of the re- cession in 2013 to 74.6 % in 20228. Notably, in the Netherlands, Sweden, Estonia, Czechia, Germany, Malta, Hungary and Denmark, the rate was 80 % or more, with increases of 15 pp or more in Malta and Hungary. In Greece, Croatia, Spain and Roma- nia, countries with less robust labour markets, the increase was also large (over 10 pp). In Italy, the increase was more modest (5 pp) to 65 % in 2022, the lowest in the EU. At the same time, the unem- ployment rate in the EU fell from 11.4 % in 2013 to 6.2 % in 2022.
Despite these positive trends, regional dispari- ties persist, especially among some population groups9. Untapped labour potential includes young people not in employment, education or training (‘NEETs’) (11.7 % of those aged 15 to 29 in 2022), the long-term unemployed (2.4 %), large numbers of women (the labour market participation rate of women as a whole being 74 %, almost 11 pp less than for men), and people with disabilities (with a participation rate of just 55.8 %).
3.1Narrowing disparities in EU labour markets continue
The response of regional labour markets during the COVID-19 pandemic and the subsequent re- covery was marked by some convergence of less developed regions. Between 2019 and 2020, em- ployment rates declined more in more developed regions than in transition and less developed ones (by 1.5 pp as against 0.8 pp and 0.6 pp). The re- gional variations reflect the severity of the meas- ures implemented to restrict economic activity, which varied between countries, and the nature of these measures – such as to preserve jobs as against supporting those losing their jobs. The economy was disrupted in each region differently,
and losses in some sectors (such as wholesaling and retailing; arts, entertainment, and recreation activities) in transition and less developed regions were offset to some extent by an expansion in ICT. Subsequently, over the two years of post-COVID recovery, employment increased faster than in the pre-crisis period in all three types of region (by around 1.5 pp a year on average).
Southern Member States, as a group, suffered the biggest fall in the employment rate (by 1.9 pp) in 2020, almost twice as much as in north-western ones (1.0 pp), while in eastern ones the reduction was negligible (0.2 pp). However, the rate also re- bounded more quickly in southern Member States (Table 2.1, upper part).
In part, perhaps because of national and EU sup- port measures and due to the exogenous nature of the pandemic, developments since 2020 con- trast with those experienced during the earlier fi- nancial and economic crisis. From 2009, employ- ment rates declined over a five-year period, with the largest falls in less developed regions. It took six to eight years for rates to return to pre-crisis levels (Figure 2.1). The biggest fall was in southern countries (of 7 pp), with the rate recovering to the pre-crisis level only after 10 years (Figure 2.2).
The developments since 2013 have seen a reduc- tion in disparities between less developed regions and others, the difference in the employment rate narrowing from 15 pp to 10 pp in 2022. The gap between north-western countries and southern ones narrowed by the same amount, while be- tween the former and eastern countries, the gap was reduced from 10 pp to only 2 pp.
Narrowing disparities are also evident across NUTS 2 regions. In several regions in Poland (5), Hungary (5), Portugal (3), Greece (Attiki), Bulgar- ia (Severoiztochen) and Romania (Bucureşti-Ilfov), the employment rate increased by 15 pp or more between 2013 and 2022, to over 78 % in some cases. Nevertheless, marked regional disparities
7European Commission (2023h). Progress towards the target is measured through the Joint Employment Report and the Employment Com- mittee monitoring tools.
8The reference year for time series comparison in further analysis of the labour market is limited to 2013, marking the end of the previous recession. 2013 represents the lows, not the start, as depicted in Figure 2.1 and Figure 2.2.
9European Commission (2022a).
Ninth Report on economic, social and territorial cohesion
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
40
- 3
How to read the chart: In 2008, the employment rate in less developed regions was 61 % (red line). As a result of the economic recession, it started to decline in 2009 (red bars - RHS), hitting a low of 58 % in 2013 and surpassed the 2009 level in 2016, reaching 62 % - after 7 years. By contrast, as a result of COVID-19, the rate fell to 65 % in 2020, and returned to the 2019 level of 67 % in 2021 – just one year later. It continued to rise in 2022, reaching 68 %.
Source: Eurostat [lfst_r_lfsd2pwc] and DG REGIO calculations (employment 2008-2020 extrapolated to be consistent with country-level break-corrected data).
Figure 2.2 Employment rates and changes by geographical area, 2008–2022
North-western EU
Southern EU
Eastern EU
80
9
change compared to previous year
75
70
6
65
60
3
55
50
0
45
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
40
- 3
How to read the chart: In 2008, the employment rate in southern EU countries was 66 % (brown line). As a result of the economic recession,
it started to decline in 2009 (brown bars - RHS), hitting a low of 59 % in 2013 and surpassed the 2009 level only in 2019, reaching 66 %. As a result of COVID-19, the rate fell to 64 % in 2020 and returned to the 2019 level of 66 % in 2021 – just one year later. It continued to rise in 2022, reaching 68 %
Source: Eurostat [lfst_r_lfsd2pwc] and DG REGIO calculations (employment 2008-2020 extrapolated to be consistent with country-level break-corrected data).
remain within Member States. In many regions in Greece (8), Romania (4), Italy (8), Spain (6), the out- ermost regions in France, Belgium (3) and Croatia (Panonska Hrvatska), the rate was still below 66 % in 2022 (Map 2.1 and Map 2.2). Some of the low- est employment rates in the EU are in the outer- most regions with some having rates below 50 %.
3.2Unemployment at record lows in many regions
Mirroring employment developments, the decline in overall unemployment, youth unemployment and NEETs resumed in 2021 and 2022 after increasing in 2020. The overall unemployment rate of those aged 20 to 64 fell to 6.2 % in 2022 0.4 pp lower than in 2019 and a substantial 5.4 pp lower than in 2013 (Table 2.1, lower part). After the recession in 2009, unemployment took until 2017–2018 to
Ninth Report on economic, social and territorial cohesion
return to pre-recession levels in north-western and eastern Member States and it was still above its 2008 level in southern ones in 2022.
The youth unemployment rate for those aged 15 to 24 declined from 25.7 % in 2013 to 14.4 % in 2022, while the NEET rate for those aged 15 to 29 fell from 16.1 % to 11.7 %. Regional dispari- ties diminished between 2013 and 2022, primar- ily because of larger reductions than elsewhere in less developed regions and in southern countries. While, however, the youth unemployment rate re- mains lower in more developed regions than in others, it was still the case till 2022 that 5–6 % of young people aged 15–24 (the youth unem- ployment ratio in Table 2.2) were unemployed, the same as in other types of regions (Table 2.2). Youth unemployment remains particularly high in the outermost regions10.
Reductions in unemployment are evident across almost all NUTS 2 regions. In a number of regions, many in Greece and Spain, both the overall and youth unemployment rates declined by more than
10pp between 2013 and 2022. Nevertheless, many of these regions, as well as some (the out- ermost ones) in France11 and Italy, still have both overall and youth unemployment rates that are more than double the EU average (Map 2.3 and Map 2.4).
The downward trend in labour market slack12 has also resumed after the increase in 2020. In 2022, the rate of slack in the EU fell to 12 % of the ex- tended labour force, 2.6 pp lower than in 2019 and
7.3 pp lower than in 2013.
3.3Labour market challenges include skill shortages
The unemployment rate fell to record lows in the EU in 2022, while the number of job vacancies reached record highs. In north-western Member States, job vacancy rates have been consistently high in the recent past in ‘professional, scientific and technical activities; administrative and support service activities’ (5.5 %), ‘construction’ (5 %) and ‘ICT’ (4.7 %). Rates have also been higher in these sectors than others in eastern countries (2.3 %, 2.4 % and 2.1 %, respectively) and they have been increasing in southern countries. There is a con- sistent pattern of high job vacancies, along with a substantial wage premium, in the ‘ICT’ sector in all three groups of regions, suggesting a shortage of supply of the relevant skills. The high job vacancy rate in the ‘professionals’ and ‘construction’ sec- tors might imply a need to adjust wages to attract and retain workers (Figure 2.3).
Although there are signs of some cooling down, with job vacancy rates declining in north-western and eastern countries13, skill shortages and a mis- match between available jobs and available work- ers have become a major issue for labour mar- kets across EU regions. This might intensify with ongoing demographic trends (see Chapter 6), and the effects of the green and digital transitions14 (see Chapters 4 and 5) on selected occupations and across all skills levels15. The 2023 demogra- phy toolbox16 (Box 2.2) outlines a comprehensive approach that empowers all generations to realise their talents and personal aspirations, also with a view to filling labour shortages. This Communica- tion on Skills and Talent Mobility enhance the EU’s
10Youth unemployment reached levels as high as 55.4 % in Mayotte in 2020, and 43.9 % in Canarias, 41.9 % in La Reunion, 38.7 % in Mar- tinique and 37.8 % in Guadeloupe (all 2022). Source: Eurostat.
11Mayotte has one of the highest unemployment rates in the EU (27.8 % in 2020, the latest year for which there are data).
12Eurostat refers to four groups of individuals as labour market slack: unemployed people according to the International Labour Organi- zation definition, those actively seeking a job but not immediately available for work, those available for work but not seeking it, and under-employed part-time workers. The extended labour force includes the labour force (unemployed and employed) and the potential additional labour force (the two categories outside the labour force, i.e. those available but not seeking, and those seeking but not available). Eurostat (2023).
13European Commission (2023b). The share of recent job starters fell significantly in summer 2022 and remained unchanged to the first half
of 2023, implying that employers were less active in recruiting new personnel.
14European Commission (2023b). Growing demand for skilled workers and occupational mismatches could affect the efficient functioning of
the labour market and lead to simultaneous increases in vacancies and unemployment.
15European Commission (2023a).
16European Commission (2023c).
Ninth Report on economic, social and territorial cohesion
Box 2.2 Demography toolbox, and addressing labour shortages
In October 2023, the Commission put forward a Communication outlining a comprehensive set of policy tools available at the EU level to support Member States in managing demographic change and its impacts. The toolbox encompasses nota- bly regulatory instruments, policy frameworks, and funding, which can be combined with national and regional policies. It stresses that gender equality, non-discrimination and inter-generational fairness must be at the heart of policy choices.
The toolbox draws on the practices and experience of Member States and sets out a comprehensive ap- proach with four pillars:
1)better reconciling family aspirations and paid work, notably by ensuring access to high-quality childcare and work-life balance, with a view to fos- tering gender equality.
2)supporting and empowering younger generations to thrive, develop their skills, and facilitate their ac- cess to the labour market and to affordable housing;
3)empowering older generations and sustaining their welfare, through reforms combined with ap- propriate labour market and workplace policies;
4)where necessary, helping to fill labour shortag- es through managed legal migration in full com-
plementarity to harnessing talents from within the Union.
The toolbox acknowledges the need to consider the territorial aspect of demographic shifts, particular- ly in regions facing population decline and a ‘brain drain’ of young workers.
The fourth pillar of the toolbox highlights the fact that demographic change, if unaddressed, could increase labour shortages, leading to economic bottlenecks. The EU is already experiencing record labour shortages, particularly in ICT, construction, care, and transport. As ‘baby boomers’ retire by the mid-2030s, shortages in both high- and low-skilled jobs are expected to increase unless countered by increased labour force participation and wage ad- justments. However, without productivity increases, higher labour costs could affect the competitiveness of EU firms in global markets.
The toolbox emphasises that to fill skill gaps, legal migration from non-EU countries is crucial, especial- ly for skills that are critical to the green and digital transitions. Despite its large labour market, the EU has relatively low inward labour migration, especial- ly of high-skilled workers, compared with other des- tinations, such as the US.
attractiveness to talent across occupations where skill shortages may exist and boost intra-EU mo- bility17. The annual sustainable growth survey for 2024 also stresses that skill shortages, namely in healthcare and long-term care, STEM18 (particularly ICT, see Maps 2.5 and 2.6), green and certain ser- vice occupations, are major bottlenecks for innova- tion and competitiveness and, so, for sustainable growth.
As regards the future of work, major trends, spe- cifically in platform and tele- working and artificial intelligence (AI)19, are likely to affect labour mar- kets in all regions. They both offer opportunities (access to flexible employment, participation in the labour market irrespective of location) and pose risks (exacerbating existing regional disparities in the necessary infrastructure). In this regard, the challenge is to respond to current regional labour and skills shortages and anticipate future ones, making use of reliable intelligence on skills, includ- ing that provided by public services.
17European Commission (2023d).
18Science, technology, engineering, and mathematics.
19European Commission (2021c). The European Commission has been working on several initiatives on the future of work. The proposed directive on platform work aims to classify digital platform workers more meaningfully and establish the first set of EU rules governing the use of AI in the workplace. The Commission is examining the implications of teleworking and the right to disconnect within the broader digitalisation of the workplace and is currently assessing the next steps in light of the European Parliament’s legislative resolution on these issues. The EU’s approach to AI centres on excellence and trust, with a focus on enhancing research and industrial capacity while ensuring safety and fundamental rights.
Ninth Report on economic, social and territorial cohesion
Table 2.2 The labour market situation of young people by level of development and by geographical area EU regions, 2013 and 2022
Note: 2021 break in LFS series.
Source: Eurostat [lfst_r_lfsd2pwc, edat_lfse_22], DG REGIO calculations.
Chapter 2: Social cohesion
How to read the maps: Two sectors – information and communication (J), and professional, scientific & technical activities and administrative & support service activities (M_N) – registered
double-digit employment growth in the EU (22 % and 15 %) between 2013 and 2020.
Excess growth is the difference between growth in the selected sector (J or M_N) and total employment growth (excluding agriculture, which broadly declined). For instance, in southern regions of Poland employment growth in sector (J) was 50 % higher than total employment growth in these regions. In all regions of Greece, employment growth in sector (M_N) was either negative or lower than total employment growth in these regions.
In cases where total employment growth (excluding agriculture) is negative, the excess growth is set to growth in the selected sector (J, M_N)
EUROPEAN COMMISSION
Brussels, 27.3.2024
SWD(2024) 79 final
COMMISSION STAFF WORKING DOCUMENT
Accompanying the document
Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions
on the 9th Cohesion Report
{COM(2024) 149 final}
Chapter 2: Social cohesion
1.Advancing equality for migrants and minorities
Migrants, Roma and other minority groups face specific challenges such as discrimination and barriers to accessing the labour market and quality education. Improving their inclusion in the labour market can help to ad- dress labour and skill shortages in the context of a declining working-age population (see Chapter 6). The EU values of equality and non-discrimination highlight the importance of having inclusive policies and practices in place so that all members of society can thrive.
1.1Migrants support regional labour markets, while facing challenges
to integrate
Migrants (in this report defined in terms of the country of birth rather than nationality), including people moving within the EU, tend to settle in re- gions of north-western Member States, especially in larger cities where there are more economic op- portunities and support networks are well estab- lished (Maps 2.19 and 2.20).
The employment of migrants, especially non- EU migrants, increased markedly between 2015 and 2019. The ‘demography toolbox’45 and the E (Employment) and social developments in Europe (ESDE) 2023 report underline the role of migrants in meeting labour shortages, particularly in low- and medium-skilled occupations46. In ad- dition, the COVID-19 pandemic has highlighted the adaptability of migrant employment to chang- ing economic conditions. The employment rate of migrants in the EU fell substantially in 2020 (by 2.5 pp), by much more than for native-born people (just 0.6 pp), but also recovered more over the next two years to 2022 (by 4.0 pp), increasing by almost twice as much as for native-born people (2.1 pp). The extent of the fall in employment in 2020 and the subsequent rebound was particu-
larly large in southern Member States and in less developed regions.
People born in another Member State are most- ly mobile EU citizens who benefit from the rights guaranteed by the free movement of workers47. As a result, they have similar, or even slightly higher, employment rates in most types of regions (Figure 2.10a), particularly in eastern Member States. Their risk of poverty or social exclusion is also much the same as for native-born people.
In contrast, migrants from outside the EU tend to have significantly lower employment rates, some 10 pp lower than the native-born in north-west- ern and southern Member States (Figure 2.10b). The disparity partly arises from a more substantial employment gap for women (15 pp) than for men (4 pp). A complex set of factors influences where non-EU migrants go and where they perform well in the labour market. They are most numerous in the more robust labour markets in north-western countries. Their employment rates are lowest in the less developed regions, though they appear to play an important role in meeting labour shortag- es, and the difference in the average rate com- pared with the native-born is less than in transition and more developed regions (8 pp lower as against 13–14 pp lower).
Despite the growth in their employment, migrants face social challenges48. Their AROPE rate in 2022 was more than double that of the native-born (40 % against 19 %), as was their rate of materi- al and social deprivation (24 % against 11 % and reaching half of the Roma population).
A recent OECD report49 has assessed the uneven impact of migrants on regions and cities, point- ing to their positive impact on regional develop- ment through innovation, international trade, re- ducing labour and skill shortages and boosting economic growth. The ‘migration outlook 2023’ of
45European Commission (2023c).
46The ESDE report 2023 highlighted that workers born outside the EU are more often employed in occupations facing persistent labour short- ages, in particular in low-skilled occupations.
47European Union (2011).
48European Commission (2022b).
49OECD (2022).
Chapter 2: Social cohesion
Figure 2.10 Employment rates and changes for migrants as against native-born, and by geographical area, 2017–2022
a) Migrants versus native-born
b) By geographical area
Native
Other EU-born
Non-EU born
80
15
75
12
North-western EU
Southern EU
Eastern EU
80
15
75
12
70
9
70
9
65
6
65
6
60
3
60
3
55
0
55
0
50
- 3
50
- 3
45
- 6
45
- 6
Note: 2021 break in LFS series.
Source: Eurostat [lfst_r_lfsd2pwc], DG REGIO calculation.
the International Centre for Migration Policy Devel- opment (ICMPD)50 and the recent Frontex report51 highlighted the pressure of a significant rise in ir- regular migration in 2022 and 2023, the highest since 2016. As regards Ukrainian refugees, the statistical evidence on their impact is not yet clear or consistent across EU regions. As of December 2023, more than 4.2 million displaced people from Ukraine had received protection under the Tempo- rary Protection Directive, which provides the right to enter the EU labour market. Cohesion funds have provided support to Member States to ensure Ukrainian refugees can access their rights under temporary protection, for example through lan- guage teaching, childcare, the certification of skills and on-the-job training.
1.2Most EU regions are friendly places for minorities to live in, though progress is needed in eastern and southern parts
Several factors can affect the labour market pros- pects of different groups and create a more in- clusive environment for them to contribute to the economy and society. These include the extent of discrimination, ease of access to education and training, and social attitudes.
Though carried out some time ago, the EU LGBT survey52 showed that lesbians, gays, bisexual and transgender people (LGBTQ+) face obstacles to enjoying their fundamental rights, particularly in employment and education.
The European Agency for Fundamental Rights 2021 Roma survey53 indicates that 25 % of Roma across the EU have experienced discrimination over the last 12 months.
50ICMPD (2023).
51Frontex, the European Border and Coast Guard Agency (2023).
52European Agency for Fundamental Rights (2014).
53European Agency for Fundamental Rights (2022).
Chapter 2: Social cohesion
Map 2.21 Living conditions for minorities, 2022
Immigrants from outside the EU
Racial and ethnic minorities
Gay and lesbian people
The more recent Gallup survey in 2022 provided in- sights into attitudes towards migrants, ethnic and racial minorities and the LGBTQ+ community in 140 EU regions (Map 2.21). It revealed that regions in north-western Member States are generally seen, by all respondents and not only migrants or minor- ities, as more friendly places for minority groups than those in eastern and southern countries.
•A significant majority of all respondents report- ed that their city or area was a ‘good place’ for racial and ethnic minorities to live, the propor- tion varying (from 50 % to 95 % across regions and being over 80 % in 80 regions). On the oth- er hand, it was less than 60 % in 10 regions in southern and eastern countries.
•Around two thirds of all respondents believed their city or area was a ‘good place’ for mi- grants to live, the proportion varying from 30 % to 97 % across regions. The figure was over 80 % in nearly 50 regions, though under 50 % in 15 regions, mainly in Hungary and Bulgaria.
•The smallest proportion of respondents consid- ered their city or area was a ‘good place’ for gay and lesbian people to live, though again the figure varied widely across EU regions, from 10 % to 95 %. It was over 80 % in around 60 regions, but under 40 % in 20 regions, pri- marily in Bulgaria and Romania.
Generally, regional differences were less pro- nounced (less than 10 pp) in countries where the overall perception of minority groups was positive, and more pronounced where the reverse was the case, with capital city regions showing the widest differences with the rest of the country.
The Gallup results are in line with the distribution of migrants across regions, most concentrating in the north-western parts of the EU, where eco- nomic conditions and social support, but also atti- tudes to migrants, are more favourable. Attitudes to migrants, therefore, tend to be most favourable where they are most numerous.
2.Summary of spatial developments
More developed regions
As indicated above, there has been a continuing increase in employment rates in more developed regions over the past decade, although less than in other parts of the EU. The average employment rate exceeded 78 % in 2022, with unemployment of only 5 %. Though youth unemployment was still 12 % and 9 % of 15–29 year-olds were classi- fied as NEETs, these figures remained less than in other regions. Several factors have contribut- ed to this relatively favourable situation. Many 25–64 year-olds have tertiary education (38 %) or upper-secondary or post-secondary vocation- al education (32 %). There seems to have been progress in upskilling and reskilling, essential for the green and digital transitions, with increased participation of adults in training. The situation of women has been constantly improving, while more women have tertiary education than men (40 % against 37 %), the gap in employment rates per- sists (74 % against 83 %). Continuing improve- ments in access to childcare (93 % of children aged 3 to compulsory school age being in ECEC) has helped to narrow this.
Transition regions
The employment rate in transition regions in- creased markedly over the period 2013 to 2022, from 67 % to 75 %, while the unemployment rate almost halved to 7 %. Nevertheless, youth unem- ployment still stood at 16 % in 2022, and 11 % of 15–29 year-olds were classified as NEETs. The fac- tors underlying the general improvement over the past decade include the relatively large proportion of 25–64 year-olds with either tertiary education (36 %) or with upper-secondary vocational qualifi- cations (35 %). There has been some rise in adult participation in education and training after the significant fall in 2020 and the situation of women has constantly improved. However, although even more women than men have tertiary education as compared with more developed regions (40 % against 32 %), the gap in the employment rate remains almost as large (71 % against 79 %), de- spite 95 % of children between 3 and compulsory school age attending pre-school education.
Less developed regions
Employment rates in less developed regions (NUTS 2) increased more than in others between 2013 and 2022, from 58 % to 69 %, and the average difference with more developed regions narrowed from 15 pp to 10 pp. The unemploy- ment rate halved to 8 % over the period, still high- er than in other regions, and though the youth unemployment rate fell by 16 pp, it remained at
22 %; and while the proportion of those aged 15–29 who were NEET also declined, it was still 16 % in 2022. Several factors underlie the worse labour market situation than elsewhere. Tertiary education rates for those aged 25 to 64 remain relatively low (26 % in 2022), though the propor- tion with upper-secondary vocational education is slightly higher (40 %). While adult participation in education and training has increased lately, it was still only 8 % in 2022. The situation of women im- proved consistently, but although the gap in tertiary education rates with men is wide (30 % against 21 %), the employment rate of women remains much lower than for men (61 % against 76 %). While some 87 % of children between 3 and com- pulsory school age attend pre-school education, this is less than in other regions. A larger proportion of people were also AROPE than in other regions (28 % in 2022 as against 19 % in more developed regions and 22 % in transition ones), though this is less than in 2016 (34 %) and the gap with more developed regions narrowed appreciably over these six years (from 14 pp to 9 pp).
References
Cedefop (2021), The green employment and skills transformation: insights from a European Green Deal skills forecast scenario, Publications Office of the European Union, Luxembourg.
EIGE (2023), Gender Equality Index 2023, European Institute for Gender Equality.
European Agency for Fundamental Rights (2014), EU LGBT survey – Main results, Publications Office of the European Union, Luxembourg.
European Agency for Fundamental Rights (2022), Roma Survey 2021 – Main results, Publications Office of the European Union, Luxembourg.
European Commission (2020a), European Skills Agenda for sustainable competitiveness, social fairness and resilience.
European Commission (2020b), A Union of Equality: Gender Equality Strategy 2020–2025, Communication from the Commission to the Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions, COM(2020) 152 final.
European Commission (2021a), Proposal for a Council Recommendation on individual learning accounts, COM(2021) 773 final.
European Commission (2021b), European Pillar of Social Rights Action Plan, Publications Office of the European Union, Luxembourg.
European Commission (2021c), Proposal for a Directive of the European Parliament and of the Council on improving working conditions in platform work, COM(2021) 762 final.
European Commission (2021d), Proposal for a Directive of the European Parliament and of the Council on combating violence against women and domestic violence, COM(2022) 105 final.
European Commission (2022a), Employment and Social Developments in Europe 2022, Publications
Office of the European Union, Luxembourg.
European Commission (2022b), 8th Cohesion Report on Economic, Social and Territorial Cohesion,
Publications Office of the European Union, Luxembourg.
European Commission (2023a), Employment and Social Developments in Europe 2023, Publications
Office of the European Union, Luxembourg.
European Commission (2023b), Labour market and wage developments in Europe – Annual review 2023, Publications Office of the European Union, Luxembourg.
European Commission (2023c), Demographic change in Europe: a toolbox for action, Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions, COM(2023) 577 final.
European Commission (2023d), Skills and Talent Mobility, Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions, COM(2023) 715 final.
European Commission (2023e), Annual Sustainable Growth Survey 2024, Communication from the Commission to the Parliament, the Council, the European Central Bank, the European Economic, Social Committee and the Committee of the Regions and the European Investment Bank, COM(2023) 901 final.
European Commission (2023f), Synthesis Report, State of Health in the EU, Publications Office of the European Union, Luxembourg.
European Commission (2023g), 2023 report on gender equality in the EU, Publications Office of the European Union, Luxembourg.
European Commission (2023h), Joint Employment Report 2024 – Commission Proposal, Publications
Office of the European Union, Luxembourg.
European Labour Authority (2023), Report on labour shortages and surpluses – 2022,
Publications Office of the European Union, Luxembourg.
European Union (2011), Regulation (EU) 492/2011 of the European Parliament and of the Council of 5 April 2011 on freedom of movement for workers within the Union.
European Union (2019), Directive (EU) 2019/1158 of the European Parliament and of the Council of
20 June 2019 on work–life balance for parents and carers and repealing Council Directive 2010/18/EU.
European Union (2020), Council Recommendation of 24 November 2020 on vocational education and training (VET) for sustainable competitiveness, social fairness and resilience (2020/C 417/01).
European Union (2022a), Council Recommendation of 16 June 2022 on a European approach to micro‑credentials for lifelong learning and employability (2022/C 243/02).
European Union (2022b), Council Recommendation of 8 December 2022 on early childhood education and care: the Barcelona targets for 2030 (2022/C 484/01).
European Union (2022c), Directive (EU) 2022/2381 of the European Parliament and of the Council of 23 November 2022 on improving the gender balance among directors of listed companies and related measures (‘Gender Balance on Boards Directive’).
European Union (2023), Directive (EU) 2023/970 of the European Parliament and of the Council of 10 May 2023 to strengthen the application of the principle of equal pay for equal work or work of equal value between men and women through pay transparency and enforcement mechanism (‘Pay Transparency Directive’).
Eurostat (2023), ‘Labour market slack – employment supply and demand mismatch’, Statistics Explained. https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Labour_market_slack_-_ employment_supply_and_demand_mismatch
Frontex, the European Border and Coast Guard Agency (2023), Irregular border crossings into EU so far this year highest since 2016, News Release.
Fulvimari et al. (2023), Economic and distributional effects of higher energy prices on households in the
EU, Fair Transition Working Paper, Publications Office of the European Union, Luxembourg.
Hazelkorn, E. and Edwards, J. (2019), Skills and Smart Specialisation: The role of Vocational Education and Training in Smart Specialisation Strategies, Publications Office of the European Union, Luxembourg.
ICMPD (2023), Migration Outlook 2023.
Menyhert, B. (2022), The effect of rising energy and consumer prices on household finances, poverty and social exclusion in the EU, EUR 31257 EN, Publications Office of the European Union, Luxembourg, JRC130650.
Menyhert, B. (2023), Energy poverty – New insights and analysis for improved measurement and policy,
Publications Office of the European Union, Luxembourg, JRC133804.
OECD (2022), The Contribution of Migration to Regional Development, OECD Publishing, Paris. OECD (2023), The uneven impact of high inflation, OECD Publishing, Paris.
EUROPEAN COMMISSION
Brussels, 27.3.2024
SWD(2024) 79 final
COMMISSION STAFF WORKING DOCUMENT
[…]
Accompanying the document
Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions
on the 9th Cohesion Report
{COM(2024) 149 final}
2.1Employment rates are higher in cities in southern and eastern Member States, and in thinly populated areas in north-western ones
As noted above, in the EU as a whole, employ- ment rates in cities, towns and suburbs, and thin- ly populated areas are similar – around 75 % in 2022. There are, however, marked differences be- tween different geographical areas (Figure 3.1a).
In north-western Member States, the employment rate for those aged 20 to 64 was 80 % in thinly populated areas and towns and suburbs in 2022, as opposed to 76 % in cities. The difference largely reflects differences in Germany, Austria, France and especially Belgium (of 10 percentage points – pp) (Figure 3.2). In southern countries, the employment rate in thinly populated areas increased markedly between 2013 and 2022 (by 14 pp) to almost the same level as in cities (to 67 % as against 69 %).
Figure 3.1 Employment, education and social indicators in regions by degree of urbanisation, 2013 (2015 for AROPE) and 2022
a)Employment rate
90
80
70
60
Cities Town and suburbs Thinly populated areas
50
b)Unemployment rate
20
16
12
8
4
Cities Town and suburbs Thinly populated areas
Cities Town and suburbs Thinly populated areas
Cities Town and suburbs Thinly populated areas
Cities Town and suburbs Thinly populated areas
Cities Town and suburbs Thinly populated areas
0
c)Tertiary education rate
50
40
30
20
10
Cities
Town and suburbs Thinly populated areas
Cities
Town and suburbs Thinly populated areas
0
d)At risk of poverty or social exclusion rate
35
30
25
20
15
Cities
Town and suburbs Thinly populated areas
Cities Town and suburbs Thinly populated areas
Cities Town and suburbs Thinly populated areas
Cities Town and suburbs Thinly populated areas
Cities Town and suburbs Thinly populated areas
10
Cities
Town and suburbs Thinly populated areas
Note: For employment and tertiary education rates: lighter parts of bars are for 2013, darker parts for the increase in 2013–2022, bar heights show the % for 2022. For unemployment and AROPE rates: the heights of bars denote % for 2013 (2015 for AROPE), lighter parts of bars show the reduction 2013–2022 (2015–2022 for AROPE), darker parts and % figures are for 2022. 2021 break in LFS series, 2020 break in EU-SILC series. Source: Eurostat [lfst_r_pgauwsc, edat_lfs_9915, ilc_peps13n] and DG REGIO calculations.
Figure 3.2 Employment rate by degree of urbanisation in EU Member States, 2022
Cities
Towns and suburbs
Thinly populated areas
National average
90
85
80
75
70
65
60
55
50
NL SE EE CZ MT DE HU DK LT FI IE CY SI PT AT LV SK PL BG LU FR BE HR ES RO EL IT
Source: Eurostat [lfst_r_ergau].
In eastern countries, the employment rate in rural areas also increased over the period (by 10 pp to 72 %) but by less than in cities (by 14 pp to 80 %), so the gap between the two widened (to 8 pp from 4 pp). In Bulgaria and Romania, the employment rate in cities was higher than the EU average and much higher than in thinly populated areas (13 pp higher in Bulgaria, 17 pp in Romania).
Unemployment rates to a large extent mirror these differences. In north-western and southern Mem- ber States, rates are lower in thinly populated ar- eas than in cities, while the opposite is the case in eastern Member States (Figure 3.1b).
2.2Tertiary education favours cities, especially in eastern Member States
Around 34 % of people aged 25 to 64 in the EU had tertiary education in 2022. However, there are substantial differences between different types of regions. The proportion was much higher in cities (44 %) than in towns and suburbs (30 %) and thinly populated areas (25 %), reflecting the strong demand for workers with tertiary education there. The average difference, moreover, widened between 2013 and 2022 (from 11 to 14 pp in towns and suburbs, and from 17 to 19 pp in thinly populated areas). The difference was substantial- ly wider in eastern Member States (46 % in cities
against 18 % in rural areas), giving rise to a large difference in employment and social outcomes (Figure 3.1c).
This pattern of difference was common across all Member States. In 10 EU Member States, over 50 % of the population aged 25 to 64 in cities – and over 60 % in Luxembourg, Lithuania, Ireland and Sweden – had tertiary education. Conversely, the proportion was below 20 % in thinly populat- ed areas in 10 Member States and around 10 % or below in Bulgaria and Romania. The disparities between cities and thinly populated areas were particularly pronounced in these two countries, as well as in Hungary, Luxembourg and Slovakia (Fig- ure 3.3). To some degree, these disparities reflect the difference in the structure of economic activity and the consequent difference in the mix of skills demanded, though they also act as a constraint on the extent to which activity can shift into higher value-added sectors in rural areas.
Vocational education and training (VET) comple- ments tertiary education and equips the economy with high skills that are essential to address la- bour shortages and deliver on the green and dig- ital transitions (see Chapter 2). Its contribution is evident in thinly populated areas, where those with VET qualifications accounted for 46 % of the pop-
Figure 3.3 Tertiary education attainment by degree of urbanisation in EU Member States, 2022
Cities
Towns and suburbs
Thinly populated areas
National average
80
70
60
50
40
30
20
10
0
IE LU SE CY LT BE NL FI DK EE FR ES SI LV AT EL PL DE PT MT BG HU SK CZ HR IT RO
Source: Eurostat [edat_lfs_9915].
ulation aged 25–64, compared with 27 % in cities and 38 % in towns and suburbs.
A low level of tertiary education coupled with a limited increase in this between 2015 and 2020 and an accelerating decline in the working-age population are features of regions in a ‘talent de- velopment’ trap, as discussed in Chapter 5. This affects 16 % of the population in the EU, main- ly in eastern Member States, especially Bulgaria, Romania, Hungary and Croatia, as well as in the south of Italy, eastern Germany and the north-east of France.
2.3Poverty and social exclusion are more prevalent in thinly populated areas of eastern and southern Member States and in cities in north-western ones
The AROPE rate declined in the EU over the period 2015–2019 and remained unchanged from then until 2022 in cities, towns and suburbs, and thinly populated areas alike. The reduction in the rate, down on average by 2.4 pp to 22 % over the seven years to 2022, was especially large in rural are- as (4.3 pp), particularly in eastern Member States (7.4 pp).
At EU level, the difference between cities, towns and suburbs, and thinly populated areas is nota- bly smaller than between more developed and less developed regions (11 pp) or between north-west- ern and southern Member States (5 pp) (as de- scribed in Chapter 2). Indeed, the difference in the rate between cities, towns and suburbs, and thinly populated areas in the EU narrowed over the pe- riod, largely as a result of the reduction in rural areas (of 4 pp to 22 %) (Figure 3.1d).
The geographical breakdown highlights the rela- tively high AROPE rates in thinly populated areas in eastern Member States, despite a large reduction over the 2015–2022 period (of 7 pp to 28 %). In Romania and Bulgaria in particular, the difference in the AROPE rate between thinly populated areas and cities was especially wide (29 pp in the for- mer, 19 pp in the latter). In Austria and Belgium, by contrast, the difference was especially wide in the opposite direction (15 pp and 11 pp, respectively) (Figure 3.4).
10
0
CZ SI FI SK NL PL SE DK CY AT BE FR HU IE MT HR DE LU EE PT LT IT LV ES EL BG RO
Source: Eurostat [ilc_peps13n].
1.Connecting territories
Mobility is important for both the economy and social life. Cohesion Policy is aimed at improving links between Member States and regions in the EU, in part by supporting the development of the trans-European transport network (TEN-T), espe- cially in regions where transport infrastructure remains under-developed7. Promoting sustainable transport and removing transport bottlenecks was one of 11 thematic objectives for Cohesion Policy in the 2014–2020 period and is part of one of the five Policy Objectives for the 2021–2027 period.
Well targeted infrastructure investment and net- work design are crucial for a transport system that provides accessibility to people and businesses and reduces regional disparities in connectivity. Public transport (especially railways) tends to be less developed outside cities in terms of network density and service frequency. Distances travelled are typically too great to use a bicycle or to walk. As a result, dependency on road transport tends to be higher.
1.1Road networks are sparser in eastern Member States and
infrastructure needs per head are higher in thinly populated areas regions8
Road accessibility depends on a sufficiently dense and fast road network that connects places and people. Various other factors also affect accessibil- ity, including the distribution of the population, the efficiency of the layout of the road network, and geophysical features such as mountains, rivers and lakes. Nevertheless, all other things being equal, greater road length per head and more roads that are motorways can be expected to result in greater accessibility and better road performance.
Over the past decade, public investment in trans- port amounted to around EUR 112 billion a year, accounting for roughly a quarter of total public in- vestment9. According to data from the Internation- al Transport Forum, the greater part of this went on roads.
2European Commission (2021).
3This sub-section is largely based on Brons et al. (2022).
4This concerns total gross fixed capital formation (Eurostat GOV_10A_EXP).
Figure 3.5 Total road length by road class in the EU (km), 2019
Local roads
Secondary roads
Motorways
0
1 000 000
2 000 000
3 000 000
4 000 000
Source: DG REGIO and JRC.
km
Two thirds of the road network in the EU consists of local roads in terms of length, just under a third of secondary roads, and only 2 % of motorways (Figure 3.5). This breakdown is much the same in all Member States.
Despite the very small part of the network made up of motorways, they are important in providing fast road connections, particularly for intermediate and long-distance journeys. The motorway net- work is well developed in most north-western and southern Member States, but much less developed in Romania, Bulgaria, Estonia and Latvia, especial- ly in the more rural parts (Map 3.3). Although these areas are served by secondary and local roads, the lack of motorways tends to imply lower speeds and so lower accessibility.
The length of roads per head differs according to the degree of urbanisation. Because of the dis- persed nature of the settlements in thinly populat- ed areas, much greater road lengths per head are required to connect them (Table 3.2). For example, local road length per head is 10 times greater in thinly populated areas than in cities (19 versus 1.8 km per inh), with towns and suburbs in an in- termediate position (just under 3 times the length per head in cities, but a quarter of the length in ru- ral areas). The length of motorways and secondary roads per head is also greater in thinly populated areas (though these roads are frequently used by people living outside these areas).
Table 3.2 Road length per inhabitant by road class and degree of urbanisation, 2018
|
Thinly populated areas
|
Towns/suburbs
|
Cities
|
All roads (m/inh)
|
31.0
|
5.5
|
2.1
|
Motorways (m/inh)
|
0.78
|
0.10
|
0.07
|
Secondary roads (m/inh)
|
11.3
|
1.00
|
0.3
|
Local roads (m/inh)
|
19.1
|
4.4
|
1.8
|
Note: Data presented here are based on grid-level classification by degree of urbanisation. Source: DG REGIO, JRC.
Canarias
Guadeloupe Martinique
Guyane
Mayotte Réunion
Açores
Madeira
REGIOgis
Map 3.3 Motorways and major roads
Roads No data
Source: JRC based on Tom Tom data.
0
500 km
© EuroGeographics Association for the administrative boundaries
1.2Road performance remains
low in some eastern Member States and thinly populated areas
Transport performance by car, defined here as the share of population within 120 km that can be reached within 90 minutes10, varied substantially between Member States in 2021. It is highest in Cyprus and only slightly lower in Malta, both rela- tively small islands, where most destinations can be reached within 90 minutes. It is also high in Belgium and the Netherlands, countries that are also relatively small and highly urbanised, with dense road networks. In Portugal and Spain, where there have been several decades of substan- tial investment in transport infrastructure11, road performance has increased markedly as a result and is now above the EU average and higher than Germany and France. Road performance is lowest in Slovakia and Romania, where road networks remain underdeveloped, and mountainous areas make road construction difficult and costly.
Road performance by car also varies substantial- ly between regions within Member States, both in less developed (especially in Greece, Bulgaria and Slovakia), moderately developed (Portugal) and more developed (Austria) ones (Map 3.4).
Road performance tends to be low in thinly popu- lated areas, especially in eastern Europe, and high in more densely populated regions, particularly in the Netherlands and Belgium, but also in many Spanish regions. In several of the latter, the pop- ulation is concentrated in densely populated cit- ies – decent road networks, accordingly, providing access to large populations within 90 minutes of driving. Most of the capital city regions have high road transport performance, including in Bulgaria, Croatia, Romania and Slovakia, where overall road performance is low.
1.3Passenger rail performance is poor compared with road, particularly in thinly populated areas
For journeys between urban areas, trains tend to be the main alternative to cars, provided there is a railway station within easy reach and the journey is affordable. As a sustainable means of transport, rail is pivotal in the design and construction of the TEN-T, because it is integral to EU climate policy. Besides the costs involved, the extent to which trav- ellers are willing to consider using trains depends in large measure on the time journeys take as com- pared with using a car. It also depends on the ease of reaching the departure station and of reaching the final destination from the arrival station12.
5For a description of the transport performance indicator see Box 3.3.
6European Commission (2016); cohesion open data platform
(https://cohesiondata.ec.europa.eu/).
7The focus of the analysis here is on accessibility and travel times and does not take account of other factors determining travel choice, including the cost – i.e. ticket price – safety and comfort.
Canarias
Guadeloupe Martinique
Guyane
Mayotte Réunion
Map 3.4 Road transport performance (% of population within a 120-km radius that can be reached in 90 minutes) by NUTS 3, 2021
%
0 – 20
20 – 40
40 – 50
50 – 60
60 – 70
70 – 80
80 – 90
90 – 100
EU-27 = 77.2
Share of population within a 120-km radius that can be reached within 90 minutes by car. Source: DG REGIO, based on Eurostat and TomTom data (FR (RUP): JRC and IGN-F).
0
500 km
© EuroGeographics Association for the administrative boundaries
Box 3.4 Estimating the impact of traffic congestion on car travel time in the EU
(I)within national borders; and
(II)within 60 minutes driving in
Canarias
Guadeloupe Martinique
Guyane
free-flow conditions, i.e. with- out congestion. As a next step, the free-flow speed2 and trav- el time on the quickest routes
Açores
55 – 60
60 – 65
> 65
no data
Source: JRC based on TomTom data.
Mayotte Réunion
Madeira
Map 3.5 Estimated average free-flow travel speed by functional urban and rural area (km/h)
km/h
<= 50
50 – 55
REGIOgis
from an origin to all destina- tions are considered. In order to track changes in speed and travel time in the morning commute, the analysis calcu- lates the travel time on the same route when the network speeds reflect those of a reg- ular weekday at 8:30 in the morning3.
Map 3.5 and Map 3.6 show, for FRAs and FUAs4, the estimated average speed of travelling in free-flow conditions and the loss in average travel speeds in weekday 8:30 am driving conditions. Free-flow speeds depend inter alia on national regulations, which explains the fact that some of the variation shows up at the country level (Map 3.5).
© EuroGeographics Association for the administrative boundaries
For example, in areas of Ger- many, Italy, Spain and Latvia speeds tend to be higher than in most other Member States. Nevertheless, there are signifi-
A recent analysis by the JRC estimates the reduction in speed and increase in travel time on the Europe- an road network due to congestion. As a first step, the approach1 uses an ‘origin-constrained spatial interaction model’, which produces a distribution of passenger car trips from every inhabited 1-km origin grid cell to all inhabited grid cells that are:
cant regional variations in most countries, indicating in particular lower free-flow speeds in urban areas. The loss in travel speed in morning peak conditions is largest in FUAs in Spain, Germany, Finland and Latvia (Map 3.6). As a general rule, reductions in speed tend to be larger in areas where the free-flow speed is higher.
1The approach is based on Jacobs-Crisioni et al. (2015), using data from Batista e Silva et al. (2021).
2Travel speeds are obtained from speed profiles recorded in the TomTom data.
38:30 in the morning is selected because, across Europe, this is when most time is lost (Christodoulou et al., 2020).
4FUAs are defined using the provisional boundaries of the 2021 Geostat grid. The specification of FRAs is an ongoing task. The defi- nition used here is the currently preferred one but is provisional.
Canarias
Guadeloupe
Guyane Martinique
Lower car travel speeds during the morning rush hour lead to losses in travel time5. Figure 3.6 shows, by Member State and ur- ban audit zone, the amount of travel time lost. This is calculated as the total estimated amount of time residents would lose when travelling their modelled jour- neys at 8:30 am travel speeds instead of free-flow speeds, rel- ative to the kilometres of road in a specific zone. In all Member States, the impact of traffic con- gestion on travel time is much greater in urban centres than in other areas. Outside urban cen- tres, the impact of congestion in commuting zones is only slightly higher than in non-commuting ones.
5Time losses need to be measured appropriately, as they depend among other things on factors such as av- erage travel speeds and lengths of travel, which vary considerably across the EU. To indicate the territorial scale of time loss, hours lost are therefore normalised by road lengths per urban audit zone.
Figure 3.6 Travel time hours lost due to morning peak traffic per km of road length
Outside commuting zone
In commuting zone
Urban centre
30
Total hours travel time loss per km
20
10
0
CY MT DK IE LU EE HR SI SE EL BG LT FI FR LV PT HU NL AT IT SK BE DE RO CZ PL ES
Source: Batista e Silva and Dijkstra (2024), JRC based on TomTom.
Rail performance is defined here as the proportion of the population living within a 120-km radius that can be reached by rail within 90 minutes (see also Box 3.3). This proportion lies between 0 and 100 % but has positive values only for people liv- ing in locations where they have access to a rail station (see Box 3.5).
In all NUTS 3 regions, transport performance by rail remains lower than by road, which hardly en- courages people to travel by train, especially if they need to travel frequently or quickly.
At the EU level the average rail performance is 15.7, which means that, on average, around just under 16 % of the population living within a 120-km radius can be reached within 90 min- utes by rail. However, there is substantial variation across EU regions (Map 3.7). Around a quarter of people in the EU have access to a reasonable rail service (rail performance indicator above 20). Most of these live in urban areas. Only some 6 % of peo- ple, all living in capital city or other metro regions, can reach over half of the population living in a 120-km radius within 90 minutes. The top-per- forming regions include Paris and surrounding re- gions, Berlin, Copenhagen and the surrounding re- gion, and Barcelona, where more people live close to a station and where there are more, and faster, train connections. In thinly populated areas, rail performance tends to be lower because the pop- ulation is more dispersed and stations are fewer
Box 3.5 Determining who has access to a rail station
To assess whether or not a person has access to a rail station, the approach followed is, first, to determine the area that can be reached within 15 minutes by:
All people living in a 200 x 200 m grid cell that has its centre in the area reachable within 15 minutes are considered to have access to the station for the purpose of this analysis.
•walking at a moderate speed;
•a bike ride at a realistic speed;
•a car ride, including time for parking and allow- ing for possible congestion; or
•a short trip by public transport.
and farther between. Indeed, many people in rural regions do not have access to a rail station at all.
Rail performance also tends to be lower in eastern EU regions, particularly in Lithuania and Romania. This is partly linked to the fact that eastern re- gions tend to be less densely populated and have a larger proportion of people living in rural regions. However, rail performance is also low in urban regions as compared with urban regions in other parts of the EU, which reflects the low investment in the rail network before EU accession.
Table 3.3 Access to primary schools (2018), universities (2020) and healthcare centres (2021–2022) by urban-rural typology including closeness to a city
|
Primary school
< 15 min walking
|
University
< 45 min driving
|
Distance to nearest healthcare centre
|
Urban
|
77.9
|
98.6
|
6.4
|
|
|
|
|
Intermediate
|
58.0
|
89.8
|
10.3
|
Intermediate – close
|
58.6
|
91.7
|
10.1
|
Intermediate – remote
|
48.6
|
61.9
|
13.6
|
|
|
|
|
Rural
|
45.3
|
69.1
|
14.0
|
Rural – close
|
44.7
|
73.9
|
13.0
|
Rural – remote
|
47.3
|
55.6
|
16.8
|
Source: DG REGIO calculations based on data from Eurostat, JRC and TomTom.
EUROPEAN COMMISSION
Brussels, 27.3.2024
SWD(2024) 79 final
COMMISSION STAFF WORKING DOCUMENT
[…]
Accompanying the document
Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions
on the 9th Cohesion Report
{COM(2024) 149 final}
Canarias
Guadeloupe Martinique
Guyane
Mayotte Réunion
REGIOgis
Map 3.7 Rail transport performance (% of population within a 120-km radius that can be reached in 90 minutes) by NUTS 3, 2019
%
<= 5
5 – 10
10 – 20
20 – 30
30 – 40
40 – 50
50 – 75
> 75
EU-27 = 15.7
Taking into account population living within 15 minutes at 15 km/h around stations.
Sources: REGIO-GIS, International Union of Railways, railway operators, JRC, TomTom.
0
500 km
© EuroGeographics Association for the administrative boundaries
1.1Urban regions have better access to education and healthcare services13
If transport networks provide poor connectivity, this typically translates into poor access to es- sential services such as education and healthcare (Map 3.8).
For children in primary education, access to school varies considerably across regions. The proportion of the population living within a 15-minute walk of a primary school is over 80 % in several re- gions in the south and east of Spain, south and north-west of Italy, north of France and the Neth- erlands. It also tends to be higher in capital city regions than others. The smallest proportions are in southern and eastern regions of Germany, and in Croatia, Latvia and Lithuania. While the average proportion is 80 % in urban areas across the EU, in rural regions and in remote intermediate regions it is less than half (Table 3.3). This might well reduce the attractiveness of such regions as places to live for families with young children.
Access to universities tends to follow a similar pat- tern. The share of the population that can reach a university within a 45-minute drive is close to 100 % in many regions in most Member States. On average, access is less in eastern Member States, but not markedly so. Regions with low access are mostly in Finland, Romania and Poland. More gen- erally, access is better in more densely populated areas. In urban regions, close to 100 % of the pop- ulation can reach a university within a 45-minute drive. In rural regions, it is only 69 %, and in re- mote rural regions, only just over half. Proximity to a university may affect the number of students needing to leave their home region to follow a uni- versity course of study, which may be reflected in higher outward migration of young people from remote rural regions than others.
Access to healthcare centres varies substantially across regions, but this partly seems to be be- cause of differences at Member State level. Re- gions where the distance to the nearest healthcare centres is on average longest, over 35 km, are in Greece, Sweden and Romania. Most centres are lo- cated in or near cities, the average distance in ur- ban regions being 6.4 km. In rural regions, the av- erage distance is over twice as long, and 16.8 km in remote ones. At the same time, the proportion of the population aged over 65, who are those most often in need of medical treatment, is largest in these regions (see Chapter 5).
2.Border regions and cross-border co-operation
Border regions account for more than 40 % of the EU’s landmass, 30 % of its GDP and 30 % of its population, some 150 million people. Almost 2 million people live in one country in the Schengen area and work in another, and some 3.5 million people cross one of the 38 internal borders of the EU every day. Many border regions are peripher- al, distant from metropolitan centres, with more limited access to healthcare and other essential services than others. Border regions can also face specific challenges in times of crises, whether linked to restrictions on cross-border movement during pandemics or a sudden influx of refugees from a conflict zone on the other side of the bor- der. Disaster prevention and precautionary action tend to be more difficult because of differences in governance, and administrative and legal sys- tems. Co-operation across borders may be a way of escaping a development trap or demographic decline. Additionally, border areas are places with high growth potential, where cultural and linguistic diversity encourages intense social and econom- ic interaction, where many people carry out daily activities on both sides of the border and where cross-border co-operation between towns and cit- ies provides opportunities for multipolar growth14.
2This subsection uses the urban-rural typology. This typology classifies NUTS 3 regions in three types: (i) urban regions: more than 80 % of the population live in an urban cluster, (ii) intermediate regions: 50–80 % live in urban clusters; (iii) rural regions: less than 50 % live in urban clusters. For a definition of urban clusters see Box 3.2.
3Strasbourgh-Kehl, Gorizia-Nova Gorica, Cieszyn-Český Těšín, Tui-Valenca, Frankfurt an der Oder-Slubice, etc.
Map 3.8 Access to education and healthcare services in EU regions by NUTS 3 region
Primary schools
Universities
Healthcare centres
Ninth Report on economic, social and territorial cohesion
These opportunities are behind the logic of Inter- reg15 intervention, both at the cross-border and transnational level. Interreg intervention supports co-operation by linking resources and people and helping to remove barriers to interaction, and building trust and a common identity.
Towards citizen-driven and people-to-people projects
Interreg has been pioneering closer involvement of citizens in Cohesion Policy. There is an increas- ing number of programmes promoting citizen-led initiatives and participation, through cross-border
4Interreg is a key EU instrument that strengthens co-operation between regions and countries within the EU. As part of the EU’s Cohesion Policy, Interreg plays a vital role in promoting regional development and cohesion, and reducing economic disparities. For the 2021–2027 period, Interreg runs with a budget of EUR 10 billion and is focused on addressing current challenges such as climate change, digital trans- formation, and social inclusion.
‘people-to-people’ projects and civil society en- gagement16. At the same time, these projects help to build solidarity and change attitudes towards neighbours living on the other side of the border. This is particularly true of projects under the first Interreg specific objective (‘a better cooperation governance’) introduced in the 2021–2027 pe- riod, to improve governance for better territorial co-operation.
Removing obstacles to co-operation
While Interreg support for cross-border interac- tion increases, co-operation encounters obstacles because of legal and administrative differences on the two sides of the border, which, inter alia, affect the functioning of the Single Market. The removal of these barriers requires decisions well beyond programme management but has potential ben- efits. It has been estimated that removing 20 % of the obstacles would generate a gain of 2 % in GDP and over 1 million jobs in border regions17. On the other hand, the economic impact of bor- der restrictions introduced because of COVID-19 was for border regions more than twice the aver- age in other regions. In 2020, 44 % of respondents in border regions identified legal and administra- tive differences as the most important obstacle to cross-border co-operation18. The Commission has recently adopted a Regulation on facilitating cross-border solutions19 to reduce the effect of these differences.
Still missing transport links
While Interreg is not designed for funding large infrastructure projects, there is a clear gap in small-scale cross-border transport connections, as illustrated by an inventory of 57 legal and ad- ministrative obstacles affecting public transport20. Not all of these take the form of missing infra- structure – in many cases they involve lack of co- ordination in timetables or ticketing.
Paving the way for enlargement
The EU has land borders with 23 countries, includ- ing the candidate countries. Participation in Inter- reg programmes, in which they are equal partners, and in macro-regional strategies gives the coun- tries concerned an opportunity to build their capac- ity to participate in Cohesion Policy programmes not only at the central but also at the local and regional level, so preparing them for accession.
3.Regions with specific
geographical features
This section examines the socio-economic perfor- mance of areas with specific geographical charac- teristics, such as island regions, outermost regions, border regions, mountain and coastal regions, and northern sparsely populated regions.
The unique features of these regions can have a significant effect on their economic development, requiring a more specific approach than other re- gions at a similar level of development. Islands, for example, may have higher transport costs, which affect the competitiveness of their industries. Mountainous regions tend to be limited in terms of available arable land and transport infrastructure. Coastal regions have issues arising from climate change, such as rising sea levels and increased vul- nerability to natural disasters. Outermost regions, geographically distant from the European main- land, have issues of isolation and reduced access to markets. Sparsely populated northern regions have problems of connectivity and accessibility.
Examining the economic dynamics of these re- gions enables a fuller assessment to be made of regional disparities across the EU. Differences in economic performance between regions can be significant, and disparities can lead to outward mi- gration, social inequalities and political tension. By comparing these regions with others, a deeper un- derstanding can be gained of the factors affecting regional development.
5Ninka et al. (2024).
6Camagni et al. (2017).
7European Commission (2020).
8European Commission (2023).
9European Commission (2022).
Box 3.7 Regional typologies based on specific geographical features
The different types of regions examined in this sec- tion are defined as follows.
•Border regions are NUTS 3 statistical regions with an international land border, or regions where more than half of the population live within 25 km of such a border. Two categories can be distinguished: external border regions – those sharing a border with countries that are not in the EU, which are mostly located along its eastern border and the border with the west- ern Balkans; and internal border regions – those sharing a border with other EU Member States or the four members of EFTA, Iceland, Liechtenstein, Norway and Switzerland. These categories are not mutually exclusive in that a region may have both an internal and an external border.
•Island regions are NUTS 3 statistical regions that consist entirely of one or more islands, islands being defined here as having: (i) a minimum sur- face area of 1 square km; (ii) a minimum dis- tance of 1 km between the island and the main- land; (iii) a resident population of more than 50; and (iv) no fixed link (e.g. bridge, tunnel or dam) with the mainland.
•Mountain regions are NUTS 3 statistical re- gions in which more than half of the land area is mountain or in which more than half of the population live in mountain areas1.
•Coastal regions are defined as NUTS 3 statis- tical regions that have a coastline, or in which more than half of their population live less than 50 km from the sea.
•Outermost regions are defined in Articles 349 and 355 of the Treaty on the Functioning of the Euro- pean Union and are Guadeloupe, Guyane, Réunion, Martinique, Mayotte and Saint-Martin (France), Açores and Madeira (Portugal) and Canarias (Spain). In the outermost regions the NUTS 2 and NUTS 3 levels coincide, except for Canarias, which are comprised of six NUTS 3 regions.
•Northern sparsely populated regions are 11 NUTS 3 statistical regions covering the four north- ernmost counties of Sweden (Norrbotten, Väster- botten, Jämtland and Västernorrland) and the seven northernmost and easternmost regions of Finland (Lapland, Northern Ostrobothnia, Central Ostrobothnia, Kainuu, North Karelia, Pohjois-Savo and Etelä-Savo). Together with the northernmost regions of Norway, they formed the ‘northern sparsely populated areas’ network in 2004.
1 The definition of topographic mountain areas is largely based on Nordregio (2004).
At the same time, the specific characteristics of these regions are a source economic potential that can be harnessed for sustainable development not only of the regions themselves but also of the wider EU. Coastal areas, for example, as well as islands and mountainous regions, can capitalise on their natural resources and tourism potential.
Table 3.4 summarises the number of NUTS 3 re- gions included in each of these types of regions as well as the share of the EU population living in them, GDP at current prices in 2021 and GDP per head in purchasing power standards (PPS) in 2021.
It should be noted that several regions are in fact included simultaneously in different categories. For example, the number of regions with internal and external borders does not add up to the total number of border regions. Mountain regions and
sparsely populated ones are often border regions. In several cases, island regions are also mountain regions, and more than half of their population live in a border region; in some cases, island regions are also outermost regions, all of the latter, except Guyane, being islands.
In terms of population, the group of coastal re- gions is by far the largest, with almost 37 % of the EU population in 2021. This is followed by border regions (28 %) and mountain regions (26 %). The remaining groups have much smaller proportions of EU the population: only 5 % in island regions, 1 % in outermost regions, and 0.5 % in northern sparsely populated regions. Between 2008 and 2021, the proportion of the population living in these regions remained remarkably stable, except for coastal and mountain regions, in which it in- creased (by 3 pp and 1 pp, respectively).
Table 3.4 Main characteristics of regions with specific territorial characteristics, 2021
EU-27
1166
(100)
|
446.5
(100)
|
14 524 809
(100)
|
32 524
(100)
|
Border regions
384
|
124.6
|
3 412 107
|
27 923
|
(33.0)
|
(27.9)
|
(23.5)
|
-85.9
|
Internal border
332
|
108.7
|
3 147 885
|
28 998
|
(28.5)
|
(24.3)
|
(21.7)
|
(89.2)
|
External border
81
|
25
|
392 579
|
20 059
|
(7.0)
|
(5.6)
|
(2.7)
|
(61.7)
|
Island regions
58
|
20.6
|
748 688
|
33 578
|
(5.0)
|
(4.6)
|
(5.2)
|
(103.2)
|
Coastal regions
339
|
163.7
|
5 337 003
|
31 014
|
(29.1)
|
(36.7)
|
(36.7)
|
(95.4)
|
Mountain regions
309
|
115.7
|
2 915 947
|
26 741
|
(26.5)
|
(25.9)
|
(20.1)
|
(82.2)
|
Outermost regions
14
|
5
|
98 368
|
19 947
|
(1.2)
|
(1.1)
|
(0.7)
|
(61.3)
|
Northern sparsely populated regions
11
|
2.2
|
93 898
|
33 995
|
(0.9)
|
(0.5)
|
(0.6)
|
(104.5)
|
Source: DG REGIO calculations based on ARDECO.
In 2021, coastal regions accounted for the same share of EU GDP as their population, while border, mountain and outermost regions accounted for smaller shares, and island and northern sparsely populated regions larger shares.
GDP per head in PPS in island regions and sparsely populated northern regions was higher than the EU
average in 2021 (3.2 % and 4.5 % higher, respec- tively), while in the other regions it was below the average, most especially in external border regions and outermost regions (both 38–39 % below).
In terms of growth of GDP per head in real terms, border regions, islands and northern sparsely pop- ulated regions had average growth rates higher
Figure 3.8 Growth rates of GDP per head (at constant prices) in regions with specific territorial characteristics in different time periods during 2001–2021
External border
Internal border
Island
Coastal
Northern sparsely populated
Mountain
Outermost
EU-27
4
Average annual change in GDP per head, %
3
2
1
0
-1
-2
-3
-4
2001–2008
2009–2019
2020–2021
2001–2021
Source: DG REGIO calculations based on Ardeco.
than the EU average over the period 2001–2021 (Figure 3.8). In the external border regions, the growth rate averaged 2.3 % a year, twice the EU average (1.1 %). This is in part because of the re- gions concerned being mostly less developed re- gions with higher growth potential than others.
The figures for the island regions must be treated with caution, as they are distorted by the fact that Ireland had a significantly higher growth rate than the EU average, especially after 2014, because of the presence of large multinational companies, whose profits form a significant share of GDP. In all island regions apart from Ireland, GDP per head declined slightly in real terms over the 20-year pe- riod, especially after 2008, which clearly reflects structural weaknesses. GDP per head in the out- ermost regions was also less than the EU average after 2008.
Dividing the period before and after the COV- ID-19 pandemic, i.e. 2009–2019 and 2020–2021, growth of GDP per head was above the EU aver- age in both sub-periods in external border regions and island regions. The latter, however, is because of Ireland. In the other island regions, GDP per head fell in both the years before the pandemic and the years after (by 2.7 % between 2019 and 2021). The outermost regions were affected most
by the pandemic, with GDP per head falling by
3.8 % between 2019 and 2021, while mountain regions also experienced a decline (of 1.5 %). The northern sparsely populated regions had higher growth than the EU average in both the 2001– 2008 and 2020–2021 periods.
GDP per head in PPS was above the EU average in northern sparsely populated regions in 2021 and for most of the 2001–2021 period (Figure 3.9). In island regions, it converged to the average after 2014 and exceeded it in 2021, again solely because of Ireland. In the other island regions, there was a steady and progressive reduction in GDP per head relative to the EU average over the period (from 84 % in 2001 to 66 % in 2021). In coastal regions, GDP per head declined relative to the average from 2010 onwards, in the aftermath of the Great Recession of 2008–2009. The same is the case for mountain regions, though at a lower level. In the outermost regions, GDP per head began to fall relative to the EU average from 2006, and in the following 15 years it fell by 17 % of the average. In internal and especially external border regions, on the other hand, GDP per head increased continu- ously relative to the EU average – especially in the latter, the level rising from 44 % of the average to 62 % over the period.
Figure 3.9 GDP per head in PPS, EU=100 in regions with specific territorial characteristics, 2001–2021
Outermost
Northern sparsely populated
External border
Internal border
Coastal
Mountain Island
110
GDP per head in PPS (EU-27=100)
100
90
80
70
60
50
40
Source: DG REGIO calculations based on Ardeco.
Figure 3.10 Change in social indicators in regions with specific territorial characteristics,
2011–2021
a)Employment rate
80
75
70
65
60
55
b)Unemployment rate
25
20
15
10
5
c)Tertiary education rate
40
30
20
10
50
0
0
Note: For employment rate and tertiary education rate: lighter bar parts are for 2011, darker parts for increase 2011–2022, and bar heights show the percentage for 2021. For unemployment rate: the bar heights show the percentage for 2011, lighter bar parts show the reduction 2011–2022 and darker parts the percentage for 2022.
Source: DG REGIO calculations based on Eurostat [urt_lfe3emp].
The different indicators of the socio-economic sit- uation in regions with specific territorial character- istics help to give a better understanding of their performance and situation relative to that of other parts of the EU21. Figure 3.10a shows that border regions (including both internal and external bor- der regions) performed slightly better than the EU average in terms of the employment rate, in terms of both the level in 2021 (76 % compared with 75 %) and the growth over the period 2011–2021 (9 pp compared with 8 pp). Coastal and mountain regions had a lower employment rate of around 70 %, but while the former have seen a substantial increase over the decade, the latter have seen only a slight rise. Island and outermost regions lag be- hind the other categories, with employment rates of 65 % and 62 % respectively, although both showed a marked improvement over the decade.
All categories of regions show a reduction in the unemployment rate over the period 2011–2021, ranging from a third to a half (Figure 3.10b).
In 2021, the border regions had a lower rate of unemployment (5 %) than the EU average, while in coastal and mountain regions it was above the average (8 %), and in the islands further above (10 %). The outermost regions had the highest rate in 2011, and although it fell by 10 pp over the following decade, it still stood at 16 % in 2021.
The share of the population aged 25–64 with ter- tiary education also varies between these catego- ries of regions and others (Figure 3.10c). In 2021, the average share was marginally larger than the EU average in coastal regions, though small- er than the average in all the other categories, if only slightly so in island regions. Mountain regions had the smallest share (29 %). Between 2011 and 2021, the share of the population with tertiary ed- ucation increased in all categories of regions and by much the same as the EU average, by slightly less in mountain and border regions, and by mar- ginally more in coastal, island and outermost ones.
10Data on these indicators were not available for the categories of northern sparsely populated regions and internal and external border regions.
References
Batista e Silva, F. and Dijkstra L. and (eds.) (2024, forthcoming), Challenges and opportunities for territorial cohesion in Europe – contributions to the 9th Cohesion Report, Joint Research Centre Science for Policy report.
Batista e Silva, F., Poelman, H., Dijkstra, L. (2021), The JRC–GEOSTAT 2018: A novel small–area population estimation for Europe, Presented at European Forum for Geography and Statistics, 7–8 September 2021.
Brons, M., Dijkstra, L., Ibáñez, J.N., Poelman, H. (2022), Road infrastructure in Europe: Road length and its impact on road performance, European Commission, DG Regional and Urban Policy, Working Paper 03/2022, Publications Office of the European Union, Luxembourg.
Camagni, R., Capello, R., Caragliu, A., Toppeta, A. (2017), Quantification of the effects of legal and
administrative border obstacles in land border regions, ABC Department Politecnico Milano, Milan.
Christodoulou, A., Dijkstra, L., Christidis, P., Bolsi, P., Poelman, H. (2020), ‘A fine resolution dataset of accessibility in European cities’, Scientific data 7, Article 279.
Dijkstra, L. and Jacobs–Crisioni, C. (2023), Developing a definition of Functional Rural Areas in the EU,
JRC135599.
Dijkstra, L., Poelma, H., Veneri, P. (2019), The EU–OECD definition of a functional urban area, OECD Regional Development Working Papers, 2019/11, OECD Publishing, Paris.
European Commission (2008), Green Paper on Territorial Cohesion. Turning territorial diversity into strength, Communication from the Commission to the Council, the European Parliament, the Committee of the Regions and the European Economic and Social Committee, COM(2008) 616 final.
European Commission (2016), Transport. Work Package 5. Ex post evaluation of Cohesion Policy programmes 2007–2013, focusing on the European Regional Development Fund (ERDF) and the Cohesion Fund (CF), Final Report, Publications Office of the European Union, Luxembourg.
European Commission (2020), Cross–border cooperation in the EU, Gallup International, Publications Office of the European Union, Luxembourg.
European Commission (2021b), A long–term Vision for the EU’s Rural Areas – Towards stronger, connected, resilient and prosperous rural areas by 2040, Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the regions, COM (2021) 345 final.
European Commission (2022), Study on providing public transport in cross–border regions – Mapping of existing services and legal obstacles – Inventory of administrative and legal obstacles to cross–border public transport, DG Regional and Urban Policy, Publications Office of the European Union, Luxembourg.
European Commission (2023), Amended proposal for a Regulation of the European Parliament and of the Council on a mechanism to resolve legal and administrative obstacles in a cross–border context. COM (2023) 790 final.
European Commission, Directorate for Regional and Urban Policy, Cohesion Open Data Platform, https:// cohesiondata.ec.europa.eu/.
Eurostat (2019), Methodological manual on territorial typologies: 2018 edition, Publications Office of the European Union, Luxembourg.
Ninka, B., Schausberger, B., Minichberger, D. (2024), Civic and civil society. Engagement in Interreg. Interact Programme.
Jacobs–Crisioni, C., Kompil, M., Baranzelli, C., Lavalle, C. (2015), Indicators of urban form and sustainable transport. Introducing simulation–based indicators for the LUISA modelling platform, JRC Technical Reports, EUR 27708 EN.
Koprić, I. (2012), ‘Consolidation, fragmentation, and special statuses of local authorities in Europe’,
Croatian and Comparative Public Administration, 12(4), pp. 1175–1196.
Ninka, B., Schausberger, B., Minichberger, D. (2024), Civic and civil society: Engagement in Interreg,
Interact Programme.
Nordregio (2004), Mountain Areas in Europe: Analysis of mountain areas in EU member states,
acceding and other European countries, Nordregio Nordic Centre for Spatial Development.
Teles, F. (2016), Local Governance and Inter–municipal Cooperation, Palgrave Macmillan, Basingstoke.
EUROPEAN COMMISSION
Brussels, 27.3.2024
SWD(2024) 79 final
COMMISSION STAFF WORKING DOCUMENT
Accompanying the document
Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions
on the 9th Cohesion Report
{COM(2024) 149 final}
THE GREEN TRANSITION
•The effects of climate change in the EU are exacerbating regional disparities, particularly in coastal, Mediterranean, and south-eastern regions. These regions are at risk of losing over 1 % of GDP annually as a result and their ageing popu- lations are more exposed to the harmful effects of climate change.
•The EU has reduced its total greenhouse gas (GHG) emissions by 27 % since 1990 while GDP has increased by 65 %. There is, however, significant regional variation. Capital city regions with high population density have the lowest emis- sions per head while regions with heavy industry have the highest. Meeting the 2030 target requires a comprehensive effort to decarbonise all sectors.
•The green energy transition offers opportunities for rural, less developed regions rich in untapped wind and solar energy potential. These regions, however, require a higher level of competitiveness and innovation as well as a skilled workforce to develop and produce the necessary clean technologies.
•The conservation status of most protected habitats and species, which are in danger of disappearing, remains unfavourable. A regional assessment of the health of forests shows that they are productive and well connected but have levels of organic carbon in their soils that are too low, and too few threatened bird species.
•Concerns persist over air, water and soil quality. Air pollution, especially in east- ern Europe and urban areas, creates health inequalities. Wastewater treatment gaps exist in south and south-eastern Europe. In rural regions built-up areas per person are increasing faster than in urban ones, weakening the capacity of soil to retain water.
•Rail has the potential to outperform flights for journeys up to 500 kilometres, provided speeds reach 175 kilometres an hour. Electric vehicle recharging points doubled in the EU between 2020 and 2022, but availability is concentrated in certain regions, creating disparities.
•6 million people work in carbon-intensive industries in the EU. Shifts to green employment favour more developed regions, so widening regional disparities.
•Extending the EU’s emissions trading system to fuels for heating buildings and transport will reduce GHG emissions but create problems for low-income, rural households and micro-enterprises that spend proportionately more on fuel.
Chapter 4
The green transition
1.Introduction
Europe has experienced unprecedented droughts, floods, forest fires and heatwaves in recent years, in line with the expected increase in frequency of these extreme weather events as a consequence of climate change. Together with biodiversity loss and environmental pollution, they underscore the urgent need for sustainable practices to protect our planet’s delicate ecosystems and ensure the exist- ence of a healthy environment for future genera- tions. The European Green Deal addresses these challenges in a co-ordinated way by providing a comprehensive framework to integrate environ- mental, economic and social dimensions to tackle ecological degradation and foster a sustainable and resilient EU. It serves as the guiding policy for the EU’s efforts to transition to a greener and more sustainable future. Its central objective is to trans- form Europe into the world’s first climate-neutral continent by 2050.
Cohesion Policy, which has been supporting the pursuit of environmental objectives, will continue to play a key role in implementing the Green Deal, notably by providing financial support and guiding regional development in a sustainable direction. The policy, with its long-standing focus on reducing socio-economic disparities between EU regions, is in line with the Green Deal’s goals of achieving a sustainable, fair and inclusive transition. In the 2021–2027 period, over EUR 100 billion is pro- grammed to go to supporting the green transition through projects on renewable energy infrastruc- ture, energy-efficiency, sustainable transport, cli- mate adaptation, and initiatives on disaster risk management, circular economy, water manage- ment, and nature conservation. Additionally, Co- hesion Policy promotes research and innovation, helping regions to develop and implement green technologies and practices1.
This chapter examines the main regional trends with respect to climate change and the environ- ment. The focus is on assessing the extent to which the impacts of climate change, biodiversi- ty loss and environmental pollution are unevenly distributed across the EU and therefore have the potential to widen inequalities between regions and the people living there. Moreover, this chap- ter examines the regional contribution to achieving climate targets and describes the challenges and opportunities of the green transition.
2.The climate and energy transition
In 2015, countries agreed in Paris on a global framework to limit global warming to below 2°C and to continue efforts to limit it to 1.5°C above pre-industrial levels. Parties also agreed to in- crease the ability to adapt to the impacts of climate change and increase climate resilience. The Euro- pean Climate Law establishes the legal framework for achieving these goals, of the EU becoming cli- mate-neutral by 2050, with an interim target of reducing net greenhouse gas (GHG) emissions by at least 55 % from 1990 levels by 2030.
The ‘Fit for 55’ package of measures is aimed at achieving this goal by revising and updating the EU’s climate legislation and policies. The main el- ements are a revised emissions trading system (ETS), including fuel use in buildings and road transport, a social climate fund, binding emission reductions for each Member State, new emission rules for cars and vans, a new carbon border ad- justment mechanism, and a target for carbon stor- age in natural ecosystems and agricultural soils. In addition, in response to the global geopolitical situation, the EU has decided to reduce its depend- ence on Russian fossil fuels, save energy, and ac- celerate the use of renewable energy while also
1At least 30 % of the European Regional Development Fund (ERDF), 37 % of the Cohesion Fund (CF), and 35 % of ‘horizon Europe’ needs to go to support climate action (mitigation and adaptation). The 2021–2027 inter-institutional agreement sets the goal of allocating at least
7.5 % of annual spending to biodiversity objectives in 2024 and 2025 and 10 % in both 2026 and 2027.
scaling up the production of clean technologies, such as batteries, wind turbines, heat pumps, pho- tovoltaics, electrolysers, and carbon capture and storage.
This section assesses current and future territorial climate effects and estimates the costs of inaction to regions. It examines the current emissions path- ways by sector and region and identifies challeng- es to achieving the 2030 emissions reduction tar- get. It also sets out trends in energy-efficiency and highlights the potential for regions to contribute to the transition from fossil fuels to renewable ener- gy generation. It addresses, in addition, the issues of sustainable mobility and a fair transition from the perspective of employment in carbon-intensive sectors and household energy costs.
2.1Regions in the frontline of climate change
The 2021 floods in the regions along the Bel- gian-German border caused direct damage of EUR
34.5 billion, while the costs resulting from the 2023 floods in Emilia-Romagna (Italy) amounted to EUR 8.5 billion. These costs show the vulner- ability of both national and regional economies to extreme weather events2. 2022 was the sec- ond-worst year in the EU as regards area burned by wildfires3. Nearly 900 000 hectares of natural land were affected by the fires. About 43 % of the total burnt area burned within ‘Natura 2000’ sites. The frequency of these events is expected to increase with climate change. These examples underscore the importance of preparing regions against the impacts of climate change.
This section reports the effects of climate change on people, ecosystems and economies at NUTS 3 level using a data-driven framework4. Historical climate data, socio-economic factors, and reported effects were combined to establish impact rela- tionships. High-resolution climate projections were used to estimate climate hazards in the EU for var-
ious global warming scenarios. The corresponding effects were determined at the regional level in 2050. These were calculated under three differ- ent scenarios for global warming levels by 2050 (of 1.5, 2 and 3°C), assuming no climate adapta- tion. The present-day baseline represents the av- erage global climate observed between 1991 and 2020, which was already 0.9°C warmer than the pre-industrial temperature. The economic costs of climate change are based on the estimated dam- age from river and coastal flooding, droughts and storms to buildings, infrastructure, agriculture, and water and energy supply. Costs resulting from en- ergy demand for climate regulation of buildings, losses in labour productivity because of high sum- mer temperatures and heatwaves, and increased maintenance of roads and railways are also in- cluded. Human exposure to climate extremes is quantified as the number or proportion of people exposed to river or coastal flooding, storms, wa- ter stress and wildfires. Finally, human mortali- ty is calculated as the number of excess deaths caused by less-than-optimal temperatures, both low and high. Not all possible impacts are included, so the total damage is therefore probably under- estimated. Table 4.1 describes the climate effects of the different impact categories used in the re- gional assessment.
The various effects of climate change impose ad- ditional costs on the EU economy. Global warm- ing of 2°C by 2050 – the most plausible scenario given current global commitments to reduce GHG emissions5 – would imply an estimated additional cost of EUR 203 billion by 2050 (0.44 % of to- tal GDP) compared with the present-day baseline. The largest economic effect comes from the en- ergy required for air conditioning in buildings and the losses in labour productivity from excessively high temperatures (Figure 4.1). These additional costs are on top of the already large effects of climate extremes on the economy at present. For instance, under the baseline scenario, the costs of damage from storms, coastal and inland flooding,
2Source: DG REGIO, data from the EU Solidarity Fund, which supports Member States with post-disaster relief –
https://cohesiondata.ec.eu-
ropa.eu/stories/s/An-overview-of-the-EU-Solidarity-Fund-2002-2020/qpif-qzyn/.
3San-Miguel-Ayanz et al. (2023).
4Based on preliminary results of an ongoing study by the Joint Research Centre (JRC), building on the ‘PESETA IV’ project:
https://joint-re-
search-centre.ec.europa.eu/peseta-projects/jrc-peseta-iv_en.
5Intergovernmental Panel of Climate Change (2021).
Table 4.1 Socio‑economic characteristics of development‑trapped regions and other regions
Coastal flooding
Coastal Europe faces rising sea levels and more intense storms, increasing economic losses and population exposure. Inadequate flood protection may amplify the damage, varying with coastal features and wealth distribution. Urbanisation exacerbates these threats.
River flooding
In most river basins, floods become more frequent and intense as global warming continues, leading to increased economic losses and population exposure. Urbanisation of river floodplains exacerbates these effects.
Droughts
The effects of drought increase most in southern and western parts of the EU, while in central and eastern European regions they remain relatively unchanged with 2°C warming. The effects in most northern and north-eastern regions will decline because of northern Europe generally becoming wetter with climate change.
Fires
Regions in the southern EU already face a high risk of fire for prolonged periods. 2°C global warming increases and lengthens fire risk in most regions, with the most significant expansion of the population exposed to the risk of wildfires being in western and south-eastern parts of the EU where scrubland and woods are close to urban areas.
Wind and storms
Projections for storms associated with global warming are highly uncertain, with the effects tending to be limited and variable in different regions of the EU. Damage from storms increases as the density of infrastructure and asset values increase.
Water availability Global warming leads to northern Europe becoming wetter and the south drier, causing the availability of water to increase in the former and diminish in the latter. The duration and intensity of water scarcity increases in existing water-scarce areas in southern Europe, along with the number of people exposed.
Labour productivity Labour productivity declines everywhere in Europe with global warming, but the effect is greater in southern regions, which are already more exposed to heat stress.
Transport
In all regions of the EU, higher temperatures increase the risk of roads rutting and rails buckling, raising operating and maintenance costs. The largest effects are projected for eastern regions, where routine maintenance is less frequent, and replacement costs higher than in other parts.
Energy
Warmer climates reduce the need for heating per unit of floor area but this is countered by increasing house sizes with higher income levels, while the need for cooling increases. This results in higher energy costs across most of the EU, most notably in the south and east.
Temperature‑ related mortality
Global warming reduces cold-related deaths because of milder temperatures. However, this is offset by the increased mortality with an ageing population. Heat-related deaths rise in all regions, amplified by population ageing. This leads to higher overall mortality from non-optimal temperatures, with the largest increases in the eastern and southern EU.
and droughts amount to EUR 28 billion a year. This is projected to rise to EUR 73 billion with a rise of 2°C by 2050, a figure well above the esti- mated costs of such damage in 2021 and 2022 (EUR 50– 60 billion)6.
Crucially, the effect is very different across regions (Map 4.1). In the vast majority of NUTS 3 regions (76 %), the additional economic costs in 2050 are estimated to remain below 1 % of regional GDP. In regions of north-eastern Germany, Lithuania and
Finland, costs would be slightly lower than today, mainly because of less risk from drought and low- er energy demand for buildings. By contrast, 42 of the 1 152 regions are estimated to face addition- al costs of over 2 % of regional GDP, 28 regions costs of over 3 %, 17 regions costs of over 4 %, 11 regions costs of over 5 %, and six regions costs of over 6 %. In several of these regions, the high costs mainly come from a large increase in coastal damage.
6European Environment Agency – EEA.
Figure 4.1 Overall estimated effects of climate change in the EU in 2050 under the present-day baseline and different global warming scenarios
Additional economic costs of climate change in
the EU compared with the present‑day baseline
Wind storms
Droughts
Maintenance of roads and railway tracks
Coastal flooding
River flooding
Loss in labour productivity from heat
Energy demand for heating and cooling of buildings
0.9
0.8
Proportion of the EU population exposed to weather
and climate‑related extreme events
Coastal flooding
River flooding
Wind storms
Wild fires
Water shortages
30
25
Excess mortality attributed to heat and cold in the EU
Heat
Cold
500
400
0.7
20
300
Annualised costs as % of GDP
0.6
0.5
15
0.4
200
10
0.3
0.2
5
100
0.1
0.0
1.5°C
in 2050
2°C
in 2050
3°C
in 2050
0
Baseline
1.5°C in 2050
2°C
in 2050
3°C
in 2050
0
Baseline
1.5°C in 2050
2°C
in 2050
3°C
in 2050
Source: JRC.
In addition to economic effects, climate change will increase people’s exposure to coastal and inland flooding, storms, water shortages and wildfires. Already, 97 million people, 21 % of the EU popu- lation, are exposed to these hazards. This number is estimated to increase to 24 % by 2050 under a 2°C global warming scenario and to over 25 % if global warming reaches 3°C. Water scarcity and wildfires have the potential to expose people to risks over a wider geographical area, while coastal and inland flooding and storms have much more localised effects and so result in less exposure. Exposure also varies markedly between the north and south (Map 4.1), with southern regions and the people living there most exposed, especially to for- est fires and water shortages.
Heat and cold are recognised environmental risk factors for human health. The current excess mortality from cold and heat in the EU amounts to 334 000 people, with the majority dying from
the cold. Overall mortality is projected to increase to 438 000, with a larger proportion dying from heat than at present. Mortality is higher in east- ern Europe than elsewhere, mainly because of population ageing more than in the rest of the EU (Map 4.1). (Perhaps unexpectedly, excess mortality from the cold is higher than from the heat, even under global warming scenarios.)
The impact of climate change on tourism, which is responsible for 5 % of total GDP, is also likely to be significant. Global warming will lead to a redi- rection of tourism. According to forecasts, a tem- perature increase of 3°C will reduce the number of summer tourists in southern coastal regions by almost 10 % and increase those in northern coast- al regions by 5 %7.
In summary, the regions that will be most affected by climate change are mainly in the Mediterranean region and in the eastern EU, especially in Bulgaria
7Matei et al. (2023).
Ninth Report on economic, social and territorial cohesion
Map 4.1 The impact of climate change under a 2°C global warming scenario in NUTS 3 regions, 2050
Additional economic costs
Human exposure to harmful climate impacts
Mortality from less-than-optimal temperatures
and Romania. Many of these regions are already poorer than the EU average. Their economies are expected to be disproportionately affected, their populations to be much more exposed to climate risks and, in the case of eastern Europe, their age- ing populations to experience higher mortality.
Climate risk management and adaptation are cru- cial in the EU to prepare for the climate impacts and to mitigate the escalating costs of the effects of extreme weather events, floods, forest fires and water scarcity. By pro-actively preparing for these challenges, EU regions can reduce the impacts on human life as well as the economic costs associ- ated with disaster response, infrastructure repair, and healthcare needs, so safeguarding their finan- cial stability. In addition, effective adaptation strat- egies enhance resilience, ensuring the well-being of both ecosystems and communities in the face of climate change. For every euro invested in risk prevention, the return on investment in terms of lives saved and damage avoided can range from EUR 2 to EUR 10, and sometimes even more8. Im- portantly, these investments can also yield addi- tional economic and social benefits. For example, nature-based solutions help reduce climate-related disaster risks such as floods or wildfires, but they also attract tourism, increase property values, and improve air quality and public health conditions.
2.2Reducing GHG emissions must be accelerated to meet the 2030 target
In 1990, total GHG emissions in the EU were 4.9 gi- gatonnes of CO2 equivalent (GtCO2eq)9. This had fallen to 3.6 GtCO2eq by 2022, a reduction of 27 %. The total amount of GHG emissions corresponds to
11.7 tCO2eq per person in 1990 and 8.0 tCO2eq per person in 202210. This is unevenly distributed across regions (Map 4.2). Capital city regions have the lowest emissions per person, often less than 5 tCO2eq, while regions with heavy industry or gas- and coal-fired power plants emit over 10 tCO2eq per person. It should be noted, however, that these
emissions are production-based and are calculat- ed by dividing the GHG emissions produced in a region by its population. This means that the emis- sions generated by the electricity consumed by a region are accounted for in the region where it is produced rather than where the demand for it aris- es. Moreover, GHG emissions from imports to the EU have not been factored in.
The downward trend in GHG emissions has not pre- vented the EU economy from expanding by 65 % between 1990 and 2022, signifying a decoupling of growth from emissions. This is demonstrated by the carbon intensity of GDP (the tonnes of GHGs emit- ted to produce EUR 1 000 of GDP), which in 2022 averaged 259 kilogrammes of CO2eq, less than half that in 1990 (600 kilogrammes of CO2eq). In sev- eral eastern countries, many regions had both low GDP and high emissions in 1990, but have succeed- ed in achieving high growth while reducing emis- sions since then. As a result, regional disparities in carbon intensity have narrowed across the EU11.
In the EU as a whole, GHG emissions have steadily decreased since 1990 at a rate of 0.1 tCO2eq per person a year. There are pronounced national and regional differences in the pattern of reduction, but three main ‘pathways’ can be distinguished (Figure 4.2). In Belgium, Czechia, Germany, France, the Netherlands, Denmark and Sweden, average emissions peaked well before 2000 and then grad- ually declined. In most of the countries that joined the EU in 2004 and in subsequent years (Estonia, Latvia, Lithuania, Poland, Hungary, Slovakia, Bul- garia and Romania), average emissions declined rapidly in the early 1990s after the collapse of the Soviet Union when GDP fell markedly, but then remained broadly unchanged, though with fluctu- ations up and down, reflecting (in some degree) developments in GDP. In the southern Member States (Spain, Portugal, Italy, Slovenia, Greece and Malta), as well as in Ireland, Austria and Finland, emissions peaked around 2005 and then declined sharply up until 2021. All three pathways show a
8International Bank for Reconstruction and Development / World Bank (2021).
9Crippa et al. (2023); GHG emissions based on the emissions database for global atmospheric research (EDGAR) excluding emissions from shipping, aviation, offshore installations and land use, land-use change, and forestry.
10Population and GDP from the annual regional database of DG REGIO; GDP at constant prices (2015 as reference year).
11European Commission (2023b).
Figure 4.2 Trends in regional greenhouse gas emissions, 1990–2022
Regions from BE, CZ, DE, DK, FR, LU, NL, SE
Regions from AT, CY, EL, ES, FI, IE, IT, MT, PT, SI
Regions from BG, EE, HR, HU, LT, LV, PL, RO, SK
2030 target (-55% from 1990)
15
Average regional emissions grouped by countries (tCO₂eq per person)
14
13
12
11
10
9
8
7
6
5
4
1990
1995
2000
2005
2010
2015
2020
2025
2030
Note: Countries are grouped based on their emission profiles. The 2030 target is at the EU level and represents a reduction in emissions of 55 % compared with 1990.
Source: JRC-EDGAR.
rebound of emissions in 2021 and 2022 as GDP recovered from the effects of the COVID-19-relat- ed restrictions on economic activity in 2020.
Achieving the 2030 target (a 55 % reduction in GHG emissions compared with 1990) means that the average GHG emissions in the EU in 2030 need to fall to 4.7 tCO2eq per person12. To achieve this, emissions will need to fall at a faster rate between 2023 and 2030 than between 1990 and 2022. Power generation and industry together accounted for nearly half of GHG emissions in 2022. For both, emissions were reduced by 37 % over the 1990– 2022 period and by 29 % over the 2005–2022 period. The two are since 2005 covered by the EU ETS, a mechanism that limits the total number of emission allowances each year. Emissions also de- clined from buildings (by 30 %) and agriculture (by 24 %) over the period, whereas emissions from transport increased by 20 %.
The challenges that regions face to reduce emis- sions differ (Map 4.3, which uses a different colour for the sector contributing most to total GHG emis- sions in 2022, indicates some of these). Agriculture contributed most to GHG emissions in the Irish and Danish regions. Transport was the most important source in rural regions in Spain, France, Italy, Aus-
tria and Germany (see also Box 3.5 in Chapter 3). Up to now, it has proved difficult to fully decarbon- ise transport, with oil and petroleum remaining the main source of power, still accounting for nearly 30 % of final energy demand in the EU. To reverse this trend, the Commission has proposed a sepa- rate emissions trading scheme for fuel combustion in buildings and for road transport, the Social Cli- mate Fund providing financial support to vulnera- ble households, transport users and micro-enter- prises in the transition to sustainable energy use.
2.3Rural, less developed regions can drive the energy transition
Achieving the EU’s climate and energy goals re- quires saving energy, increasing the share of re- newable energy, using energy more efficiently, and enhancing carbon sinks. Beyond reducing GHG emissions, these measures also help lower ener- gy bills, protect the environment, and reduce fossil fuel purchases (and hence the EU’s dependence on oil and gas imports).
In 2021, the EU’s primary energy consumption was 1 309 million metric tonnes of oil equiva- lent (Mtoe), down 12.6 % from 2005. The current 2030 target is 992.5 Mtoe. At the country level,
12European Commission (2023a).
Chapter 4: The green transition
Figure 4.3 Energy statistics by country
a)Change in primary energy consumption, 2005 to 2021
%
30
20
18
10
0
-10
-20
-30
-40
-33
EL PT IT LT ES DE DK MT EE FR NL SI LU SE HR RO IE CZ CY SK FI BE HU AT BG LV PL EU
Source: Eurostat [NRG_IND_EFF].
%
b) Share of energy from renewable sources in 2021
70
63
60
50
40
30
20
10
0
LU MT IE NL BE HU PL BG SK CZ CY IT DE FR ES EL RO SI LT HR PT DK AT EE LV FI SE EU
Source: Eurostat [NRG_IND_REN].
c) Installed renewable energy capacity in 2022
GW
12
10
8
6
4
2
0
DE ES PL NL FR SE IT
FI PT EL DK AT BE HU IE CZ RO SI BG EE LT LU CY LV HR MT SK
Source: Wind Europe and Solar Power Europe.
(62.6 %) had by far the largest share coming from renewables in the EU, ahead of Finland (43.1 %) and Latvia (42.1 %). At the other end of the scale, Luxembourg (11.7 %) had the smallest share. For- est biomass is an important source of renewable energy, especially in northern Europe. It should be emphasised that biomass can only contribute effectively to reducing GHG emissions if it is pro- duced in a sustainable way.
Following Russia’s war of aggression against Ukraine and the subsequent rise in energy prices, demand for natural gas in the EU fell by 13 % in 2022, the sharpest decline in history13. While mild- er winter temperatures played a role, policy was also important, particularly record increases in so- lar and wind capacity. Two industry organisations, SolarPower Europe14 and WindEurope15, have esti- mated that 41 GW of new solar photovoltaic (PV) capacity and 16 GW of additional wind capacity, mostly onshore, were installed in the EU in 2022, signifying an increase of 47 % relative to 2021 for solar and 40 % for wind power. Germany and Spain accounted for nearly 35 % of the overall in- crease in renewable capacity.
These numbers suggest that EU policies to reduce reliance on Russian fossil fuels and to accelerate the green energy transition are succeeding. How- ever, achieving a carbon-neutral energy sector requires further upscaling of renewables and there is substantial untapped potential in this regard16.
the largest reductions in energy up to 2021 were achieved in Greece (of 33 %) – where GDP declined substantially after 2002, so depressing energy demand – Portugal (21 %) and Italy (20 %) (Fig- ure 4.3). Poland is the only country that consumed more primary energy than in 2005 (18 % more).
In 2021, renewable energy accounted for 21.8 % of gross energy consumption in the EU, only around half the target for 2030 (42.5 %). Again, there are wide variations between countries. Sweden
In 2023, solar, wind and hydro power installed in the EU together produced 972 terawatt hours (TWh) of electricity. But this represents only a frac- tion of the technically available potential, estimat- ed at 12 485 TWh a year, divided between solar PV (88 %), onshore wind (11 %) and hydro pow- er (1 %). The potential amounts to over 5 times the electricity consumed in 2021 and is mainly concentrated in the EU’s rural areas (9 784 TWh). It would come predominantly from potential
13 IAE (2023).
14SolarPower Europe (2022).
15WindEurope (2022).
16Perpiña Castillo et al. (2024).
Ninth Report on economic, social and territorial cohesion
EUROPEAN COMMISSION
Brussels, 27.3.2024
SWD(2024) 79 final
COMMISSION STAFF WORKING DOCUMENT
[…]
Accompanying the document
Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions
on the 9th Cohesion Report
{COM(2024) 149 final}
and reskilling of workers, investments in small and medium-sized enterprises, creation of new firms, research and innovation, environmental rehabilita- tion, clean energy, job-search assistance and trans- formation of existing carbon-intensive installations.
It is equally essential to prioritise social equity and provide support for workers affected and their households. Investing in retraining programmes through JTF support can help people acquire the skills to take up green economy jobs, while finan- cial support can reduce the burden on low-income households and create a more equitable transition path.
1.1Progress toward a just transition
in fossil and energy‑intensive industries
This section presents regional statistics on current employment in carbon-dependent or carbon-inten- sive sectors in the EU and identifies the areas and activities where the green transition is creating new jobs. It also assesses the territorial impact of extending the ETS to fuels for residential heating and transport. Coal and carbon-intensive regions in the EU that are identified as most severely af- fected by transition process, receive support from the JTF to support the diversification of their econ- omies in the affected sectors.
Almost 340 000 people were directly and indirectly employed in the coal industry in the EU in 2018. The jobs concerned are highly concentrated, with 60 % in just seven regions (Śląskie and Łódzkie in Poland, Sud-Vest Oltenia in Romania, Yugoiztochen in Bul- garia, Severozápad in Czechia, Köln and Branden- burg in Germany, and Dytiki Makedonia in Greece) (Map 4.17). It is estimated that between 54 000 and 112 000 direct jobs could be lost by 203047.
The peat and oil shale industries are smaller. The former is estimated to employ, directly and indirectly, just under 12 000 and the latter almost 7 000, all in Estonia, the only country in the EU with such an industry. Closing down these indus- tries could have a significant impact on local and
regional employment and will require economic restructuring.
More people work in carbon-intensive industries. In 2020, nearly 6 million people were employed in the car, steel, minerals, paper, chemicals, coke and petroleum sectors, 3 % of total employment in the EU. The main employment clusters in these sectors are in central Europe (Map 4.18).
The coal industry and carbon-intensive manufactur- ing face transformational challenges given the EU commitment to becoming climate-neutral by 2050. This means phasing out coal and shifting to low-car- bon technologies, such as those based on hydrogen, and using carbon capture and storage where decar- bonisation is not yet possible. It also means helping to mitigate the socio-economic and environmental impact of the transition on regions and the people living there. Case studies of fossil fuel phase-out (coalmining in the UK, oil refining in Croatia, and peat extraction in Finland) have shown that car- bon-dependent industries are often deeply rooted in local culture and identity48. The industries are con- centrated in a few places and job losses have been shown to have long-term adverse physical, mental and social effects on the people and communities concerned. Attempting to retrain the workers losing their jobs is insufficient. There needs to be long- term cohesive educational, financial and social sup- port to ensure a just transition. The support involved needs to be early and targeted, with collaboration with existing local support networks and alignment of interests among key stakeholders. The case stud- ies highlight the importance of place-based meas- ures, centred on partnership.
1.2Competitiveness and sustainability of sectors in the climate and energy transition
The transition to a competitive green economy is underway, but the pace varies between regions. The regional competitive environmental sustaina- bility indicator49 has been developed to show the share of employment in 56 NACE (nomenclature
32Alves Dias et al. (2021).
33Kaizuka (2022).
34Marques Santos et al. (2023) and update for 2019 and 2020 in Marques Santos et al. (2024).
Ninth Report on economic, social and territorial cohesion
of economic activities) sectors that are systemat- ically more competitive and sustainable than the EU median (Map 4.19). Sectoral competitiveness is measured by labour productivity and sustaina- bility by GHG emissions per worker. The indicator has been calculated for the years 2008–2020 and shows the shift in employment towards greener and more productive sectors over this period.
In 2019, the average region had 17 % of employ- ment in sectors that were both more competitive and more sustainable than the EU median. The share was largest in southern Germany, northern Austria, southern Ireland, and southern Scandi- navia, as well as in capital city regions. Between 2008 and 2020, the share increased by signifi- cantly more in more developed regions than in less developed or transition ones (Figure 4.15), widen- ing the difference between them.
Econometric analysis suggests that the transi- tion to more competitive and sustainable regional economies is positively associated with investment co-funded by the ERDF, CF and European Social Fund50. This is particularly true in respect of com- petitiveness and the restructuring towards higher value-added sectors, which is especially evident in less developed regions that receive most funding.
Improvements in sustainability, however, are much less evident, suggesting that this is more difficult to achieve and that the transition to a low-carbon economy requires more time and effort. Factors such as R&D, the quality of government, and the qualifications of the workforce seem to be impor- tant in this regard. Adequate policy-making, reforms and investment are essential to implement the tran- sition to a low-carbon economy and adjust to new circumstances in a way that spurs employment, competitiveness and economic growth, with a focus on leveraging circular economy principles and de- ploying clean technology solutions to drive innova- tion and efficiency across industries.
1.3Longer‑term impact of the extension of the ETS and the transformation of industrial and service sectors
The ETS is designed to limit emissions of GHGs from power generation and large industrial plants through a cap-and-trade mechanism. In 2021, the ETS covered 40 % of GHGs emitted in the EU. In 2023, the EU approved a new ETS for fuel com- bustion in buildings, road transport and a few oth- er sectors. The emissions concerned account for another 40 % of EU emissions and so are equally
Figure 4.15 Trends in the regional competitive environmental sustainability indicator by category of region for Cohesion Policy, 2008–2020
Less developed
Transition
More developed
30
REGIOgis
Map 4.19 Regional competitive environmental suitability indicator, 2019
% of total employment
<= 10
10 – 20
20 – 30
30 – 40
40 – 50
50 – 60
> 60
Share of employment of 56 NACE sectors that are systematically more competitive and more sustainable than the EU median.
Source: JRC.
0
500 km
© EuroGeographics Association for the administrative boundaries
important for achieving climate objectives. The share of emissions covered varies between coun- tries and regions. The share is largest in Luxem- bourg (Figure 4.16), mainly because of interna- tional through traffic.
While GHG emissions from household energy con- sumption declined by 30 % between 1990 and 2021, those from road transport, which remains highly dependent on oil and petrol, increased by 18 %.
Higher prices for carbon fuels give an incentive for innovation and help to reduce emissions, but they tend to hit poorer households harder. The ex- tension of the ETS means that climate action will become more tangible for people, as they will be directly affected in heating their homes and using their cars as taxes are imposed or increased from 2027 under the system. Across the EU, households spend an average of between 3 % and 10 % of their income on heating and fuel (Figure 4.17). Although household expenditure on heating fu- els in the EU increases with household disposable income51 – for the 20 % of households with the highest income (i.e. in the top quintile of the in- come distribution), expenditure is around twice as
high as for the 20 % with the lowest levels – it in- creases less than in proportion. It, therefore, repre- sents a larger share of overall expenditure for the households in the bottom quintile than for those in the top. Fuel price increases, therefore, affect poorer households more because more of their budget goes on heating, posing increased risks of energy poverty. Households living in densely pop- ulated areas systematically spend less on heating than those in intermediate or sparsely populated areas, irrespective of income levels.
Total expenditure on fuel for transport is highest for all income groups in rural areas, and lowest in urban areas. The share decreases as income increases. As expected, the share of expenditure for transport fuels is larger in rural areas than others because of the greater use of private cars and motorcycles and a lower availability of public transport.
Extending the ETS to include fuel for heating and transport will therefore have a particularly large impact on low-income households in rural areas. The sharp increase in energy prices in 2022 seems to have led households to seek alternatives for heating their homes–firewood and heat pumps in particular.
Figure 4.16 Emissions under the ETS and ETS2
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
LU EE BE CZ PL NL IE DE FI CY SK SI BG LT AT HU IT DK ES EL HR FR RO LV SE PT MT
Source: EDGAR (JRC).
36Koukoufikis and Uihlein (2022); Ozdemir and Koukoufikis (2024).
Figure 4.17 Average expenditure and share of household income going on fuel for heating and transport by income quintile, EU, 2020
Expenditure on heating fuels
Expenditure on heating fuels
1 200
Densely populated
Intermediate populated
Sparsely populated
Densely populated
Intermediate populated
Sparsely populated
6
Share of expenditure in %
1 000
5
800
4
600
3
400
2
200
1
0
1
2
3
4
5
Income quintile
0
1
2
3
4
5
Income quintile
2 000
1 800
1 600
1 400
1 200
1 000
800
600
400
200
0
Expenditure on transport fuel
Densely populated
Intermediate populated
Sparsely populated
1
2
3
4
5
Income quintile
Expenditure on transport fuel
Densely populated
Intermediate populated
Sparsely populated
5
Share of expenditure in %
4
3
2
1
0
1
2
3
4
5
Income quintile
Note: Data for CZ, IE, IT, PL, PT, RO, FI and SE are not yet available for 2020; for CZ in 2015, population weights were adjusted with European Union statistics on income and living conditions (EU-SILC) weighted total number of households.
Source: JRC based on Eurostat.
The price of firewood and pellets52, therefore, was 54 % higher in the EU in November 2022, when it peaked, than the year before, and in Austria, Den- mark, the three Baltic States, and Slovenia, twice as high, while sales of heat pumps in the EU increased by 39 % in 202253.
37According to the Eurostat harmonised index of consumer prices (other solid fuels comprise coke, briquettes, pellets, firewood, charcoal and peat).
38European Heat Pump Association (2023).
2.Key messages
The green transition has the potential to reduce regional inequalities, but it could equally lead to them widening. On the one hand, it is expected to create new jobs, provided it is supported by appro- priate policies, especially in rural, less developed regions that have high potential for the devel- opment of wind and solar power and for carbon capture and storage in natural ecosystems. On the other hand, there is evidence that the green tran- sition favours more developed regions, attracting investment and skilled workers there, while pos- ing challenges for employment and households in low-income rural areas, in particular, and poten- tially exacerbating social inequalities.
Addressing these challenges requires deepening the territorial approach to implementing the green transition in an equitable way. This can be done by supporting vulnerable regions through co-financing investment in renewable energy, energy-efficiency, clean and circular technologies, carbon-free vehicles and the corresponding infrastructure, and retraining and education, taking into account the ‘do no signif- icant harm’ principle to balance trade-offs. This is particularly important in less developed regions, which tend to be less prepared for the transition to a climate-neutral economy and to have more diffi- culty in reaping the potential benefits. It is equally important to prioritise social equity and provide sup- port for the workers affected, through retraining so that they have the skills to take up green jobs, and to help mitigate the burden on low-income households. As the green transition unfolds, minimising the im- pact on energy costs is vital to prevent heightened risks of energy poverty. Also, rural-proofing can help make policies on climate adaptation, energy, trans- port or employment fit for purpose.
Climate risk management and adaptation to climate change is becoming increasingly important to miti- gate the escalating costs of extreme weather events, floods, forest fires and water shortages. Better pre- paredness and increased climate resilience, such as by protecting and restoring ecosystems, depend on pro-active territorial policies to help vulnerable re- gions reduce the economic costs of disaster mitiga- tion, infrastructure repairs and the consequences for healthcare, and so ensure their financial stability.
References
Alves Dias, P., Conte, A., Kanellopoulos, K., Kapetaki, Z., Mandras, G., Medarac, H., Nijs, W., Ruiz Castello, P., Somers, J. and Tarvydas, D. (2021), Recent trends in EU coal, peat and oil shale regions, EUR 30618 EN, Publications Office of the European Union, Luxembourg, JRC123508.
Brons, M., Dijkstra, L., Poelman, H. (2023), How fast are rail trips between EU cities and is rail faster than air?, DG Regional and Urban Policy, Working Paper 03/2022, Publications Office of the European Union, Luxembourg.
Cedefop (2021), The green employment and skills transformation: insights from a European Green Deal skills forecast scenario, Publications Office of the European Union, Luxembourg.
CINTRAN (2023), The regions face individual socio–economic challenges – A primer to the socio–economic risk index, https://storymaps.arcgis.com/ collections/263f2045c6c54b9ab120755369d64756?item=2.
Crippa, M., Guizzardi, D., Pagani, F., Pisoni, E. (2023), GHG Emissions at sub–national level (v2.0), European Commission, Joint Research Centre (JRC) [dataset].
Dougherty, S., Montes Nebreda, A., The multi–level fiscal governance of ecological transition, 2023, OECD Working Papers on Fiscal Federalism, 44, OECD Publishing, Paris.
European Central Bank (2023), The economy and banks need nature to survive,
https://www.ecb.europa.
eu/press/blog/date/2023/html/ecb.blog230608~5cffb7c349.en.html.
European Commission (2020), Sustainable and smart mobility strategy – putting European transport on track for the future, Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions, COM (2020) 789 final.
European Commission (2023a), Climate action progress report 2023.
European Commission (2023b), EU regional GHG emissions pathways to net–zero, https://cohesiondata. ec.europa.eu/stories/s/EU–Regional–emissions–pathways–to–net–zero–decarbo/hb9w–9hyd/.
European Commission, DG for Mobility and Transport, https://transport.ec.europa.eu/transport–themes/ clean–transport/clean–and–energy–efficient–vehicles/green–propulsion–transport/hydrogen–and–fuels– cells–transport_en.
European Environment Agency (2018), Unequal exposure and unequal impacts: social vulnerability to air pollution, noise and extreme temperatures in Europe, EEA Report No 22/2018, Publications Office of the European Union, Luxembourg.
European Environment Agency (2023a), Economic losses from weather– and climate–related extremes in Europe,
https://www.eea.europa.eu/en/analysis/indicators/economic–losses–from–cli
mate–related.
European Environment Agency (2023b), Water scarcity conditions in Europe (Water Exploitation Index Plus),
https://www.eea.europa.eu/en/analysis/indicators/use–of–freshwater–resour
ces–in–europe–1.
European Environment Agency (2023c), European bathing water quality in 2022.
European Heat Pump Association (2023), European heat pump market and statistics report 2023. Eurostat (2020), Attitudes of European citizens towards the environment, Spatial Eurobarometer 501.
Intergovernmental Panel on Climate Change (2021), ‘Summary for Policymakers’. In: Masson–Delmotte, V., Zhai, A. Pirani, S.L. Connors, C. Péan, S. Berger, N. Caud, Y. Chen, L. Goldfarb, M.I. Gomis, M. Huang,
K. Leitzell, E. Lonnoy, J.B.R. Matthews, T.K. Maycock, T. Waterfield, O. Yelekçi, R. Yu and B. Zhou (eds.), Climate Change 2021: The Physical Science Basis, 2021, Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, pp. 3−32.
International Bank for Reconstruction and Development / World Bank (2021), Economics for Disaster Prevention and Preparedness. Investment in Disaster Risk Management in Europe Makes Economic Sense.
International Energy Agency (2023), Europe’s energy crisis: What factors drove the record fall in natural gas demand in 2022?, International Energy Agency, Paris.
Kaizuka, S. (2022), A Just Transition?: How to Leave no one Behind, a Lesson from a Comparative Analysis of the British Coal Mines and the Croatian Oil Refinery, Third Joint EU Cohesion Policy Conference, 14–16 November, Zagreb.
Koukoufikis, G., Uihlein, A. (2022), Energy poverty, transport poverty and living conditions – An analysis of EU data and socioeconomic indicators, EUR 31000 EN, Publications Office of the European Union, Luxembourg, JRC128084.
Maes, J., Bruzón, A.G., Barredo, J.I., Vallecillo, S., Vogt, P., Marí Rivero, I., Santos–Martín, F. (2023), Accounting for forest condition in Europe based on an international statistical standard, Nature Communications 14, 3723.
Maes, J., Teller, A., Erhard, M., Conde, S., Vallecillo Rodriguez, S., Barredo Cano, J.I., Paracchini, M., Abdul Malak, D., Trombetti, M., Vigiak, O., Zulian, G., Addamo, A., Grizzetti, B., Somma, F., Hagyo, A., Vogt, P., Polce, C., Jones, A., Marin, A., Ivits, E., Mauri, A., Rega, C., Czucz, B., Ceccherini, G., Pisoni, E., Ceglar, A., De Palma, P., Cerrani, I., Meroni, M., Caudullo, G., Lugato, E., Vogt, J., Spinoni, J., Cammalleri, C., Bastrup–Birk, A., San–Miguel–Ayanz, J., San Román, S., Kristensen, P., Christiansen, T., Zal, N., De Roo, A., De Jesus Cardoso, A., Pistocchi, A., Del Barrio Alvarellos, I., Tsiamis, K., Gervasini, E., Deriu, I., La Notte, A., Abad Viñas, R., Vizzarri, M., Camia, A., Robert, N., Kakoulaki, G., Garcia Bendito, E., Panagos, P., Ballabio, C., Scarpa, S., Montanarella, L., Orgiazzi, A., Fernandez Ugalde, O. and Santos–Martín, F. (2020), Mapping and Assessment of Ecosystems and their Services: An EU ecosystem assessment, EUR 30161 EN, Publications Office of the European Union, Luxembourg, JRC120383.
Marques Santos, A., Barbero, J., Salotti, S., Diukanova, O. and Pontikakis, D. (2023), ‘On the road to regional ‘Competitive Environmental Sustainability’: the role of the European structural funds’, Industry and Innovation, 30(7), pp. 801–823.
Marques Santos, A., Barbero, J. and Salotti, S. (2024), Measuring transition to a competitive and sustainable economy, Territorial Development Insights Series, JRC Working Papers on Territorial Modelling and Analysis, 07/2022, European Commission, Seville, JRC136629.
Matei, N., Garcia Leon, D., Dosio, A., Batista E Silva, F., Ribeiro Barranco, R. and Ciscar Martinez, J.C. (2023), Regional impact of climate change on European tourism demand, EUR 31519 EN, Publications Office of the European Union, Luxembourg, JRC131508.
Maucorps, A., Römisch, R., Schwab, T., Vujanovic, N. (2022), The Future of EU Cohesion – Effects of the Twin Transition on Disparities across European Regions, The Vienna Institute for International Economic Studies and Bertelsmann Stiftung.
Ozdemir, E., Koukoufikis, G. (2024), Just Energy and Transport Data Inventory, European Commission, Petten, JRC135908.
Perpiña Castillo, C., Hormigos Feliu, C., Dorati, C., Kakoulaki, G., Peeters, L., Quaranta, E., Taylor, N., Uihlein, A., Auteri, D., Dijkstra, L. (2024), Renewable Energy production and potential in EU Rural Areas, Publications Office of the European Union, Luxembourg, JRC135612.
Rodríguez–Pose, R., Bartalucci, F. (2023), Regional vulnerability to the green transition, Single Market Economics Papers, Publications Office of the European Union, Luxembourg.
San–Miguel–Ayanz, J., Durrant, T., Boca, R., Maianti, P., Liberta, G., Jacome Felix Oom, D., Branco, A., De Rigo, D., Suarez–Moreno, M., Ferrari, D., Roglia, E., Scionti, N., Broglia, M., Onida, M., Tistan, A., Loffler, P. (2023), Forest Fires in Europe, Middle East and North Africa 2022, Publications Office of the European Union, Luxembourg, JRC135226.
Sasse, JP., Trutnevyte, E. (2023), ‘A low–carbon electricity sector in Europe risks sustaining regional inequalities in benefits and vulnerabilities’, Nature Communications, 14, 2205.
Sippel, L., Nolte, J., Maarfield, S., Wolff, D., Roux, L. (2018), Comprehensive analysis of the existing cross–border rail transport connections and missing links on the internal EU borders – Final report, Publications Office of the European Union, Luxembourg.
SolarPower Europe (2022), European Market Outlook for Solar Power 2022–2026.
Többen, J., Banning, M., Hembach–Stunden, K., Stöver, B., Ulrich, P., Schwab, T. (2023), Energising EU Cohesion – Powering up lagging regions in the renewable energy transition, Bertelsmann Stiftung, Gütersloh.
Vysna, V., Maes, J., Petersen, J.E., La Notte, A., Vallecillo, S., Aizpurua, N., Ivits, E., Teller, A. (2021), Accounting for ecosystems and their services in the European Union (INCA), Final report from phase II of the INCA project aiming to develop a pilot for an integrated system of ecosystem accounts for the EU. Statistical report, Publications office of the European Union, Luxembourg.
WindEurope (2022), Statistics and the outlook for 2023–2027.
World Economic Forum (2020), Nature Risk Rising: Why the Crisis Engulfing Nature Matters for Business
and the Economy, https://www3.weforum.org/docs/WEF_New_Nature_Economy_Report_2020.pdf.
EUROPEAN COMMISSION
Brussels, 27.3.2024
SWD(2024) 79 final
COMMISSION STAFF WORKING DOCUMENT
Accompanying the document
Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions
on the 9th Cohesion Report
{COM(2024) 149 final}
Regional innovation and the digital transition
205
REGIONAL INNOVATION
AND THE DIGITAL TRANSITION
•Innovation shapes markets, transforms economies, stimulates changes in the quality of public services and is indispensable to achieving the overarching objec- tives of the twin green and digital transitions.
•Innovation is an important driver of long-run productivity growth and a key deter- minant of the competitiveness of firms, especially those in the EU competing in an increasingly competitive and fragmented geopolitical context.
•From a forward-looking perspective, the green and digital transitions have the potential to dramatically redefine production processes and value chains globally, with clear implications for economic geography and with more innovative firms finding it easier to adjust and take advantage of the opportunities that arise.
•There is potential for all EU regions to benefit from the digital transition, but the economic structure of more developed regions suggests that they are better equipped to do so.
•This is in line with the existing indicators of the geography of innovation – meas- ured in terms of skills and education, R&D, patent activity, or composite indicators such as the Regional Innovation Scoreboard – which show a clustering around more developed, often metropolitan, areas and a persistent innovation divide.
•There is evidence pointing to substantial untapped potential for cross-border co-operation across all types of EU region in developing the value chains needed for the twin transitions.
•Place-based approaches can unlock the potential of all regions to innovate in line with their strengths and characteristics.
•Education – from early childhood to tertiary – plays a foundational role in fos- tering innovation. Investment in education is essential for creating the skilled, re- silient and adaptable workforce required for sustained innovation and long-term economic development.
•Investment in R&D that fosters innovation in developed regions can have signif- icant benefits for neighbouring ones, while for less developed regions, policies to improve the quality of institutions are equally important for stimulating innovation.
•The development of digital skills and access to a fast internet connection are key to ensuring that all regions can harness the potential of the digital transition. Over the past few years, there has been a significant improvement in broadband connectivity in connectivity in many regions, but wide disparities across the EU remain as well as a persistent rural-urban gap in access to very-high-capacity networks.
Chapter 5
Regional innovation and the digital transition
206
1.Innovation and competitiveness of EU regions in a new complex global environment
Innovation plays a pivotal role in driving long-term productivity growth and competitiveness1. Innova- tion shapes markets, transforms economies, stim- ulates changes in the quality of public services and is essential for achieving the overarching objectives of the twin green and digital transitions. A substan- tial amount of the European Regional Development Fund (ERDF) (EUR 56 billion for the 2021–2027 period) goes to foster research and innovation (R&I) in the EU through place-based programmes co-managed at the local level (‘smart specialisa- tion’ strategies, see Box 5.2). These programmes play a central role in strengthening regional inno- vation ecosystems so that they are better equipped to stimulate and sustain economic development2.
More skilled and creative workers, increasingly ef- ficient and powerful machines, new products and processes are key dimensions of innovation in an increasingly competitive global environment. Their importance has become evident over time, as EU firms have increasingly had to compete with those from emerging economies rapidly moving-up the value chain. These economies still have the advan- tage of cheaper labour, less stringent environmen- tal regulations, and a rapid pace of technological
advancement3. Moreover, in some areas, such as South-East Asia and China, they have reached the technological frontier in a number of sectors4. In advanced manufacturing and green technolo- gies, the EU is a world leader in innovation. How- ever, more effort is needed to maintain and further build a strong global position in digital technolo- gies, an area where the US is a leader and emerg- ing economies are becoming stronger5.
Prospectively, the green and digital transitions have the potential to dramatically redefine pro- duction processes and value chains globally, with clear implications for economic geography. In this regard, the creation and diffusion of innovation
– and its spatial dimension – are key not only to the competitiveness of the EU in the global econ- omy, but also to its economic, social and territorial cohesion.
Empirical studies support the notion that innova- tion tends to concentrate in specific geographical areas, underlining the importance of understand- ing the spatial, social and economic dimensions of innovation. The link between innovation and spatial agglomeration effects has been extensively stud- ied, and the close proximity of firms, suppliers, and related institutions in a cluster has been shown to foster innovation6. Agglomerations facilitate the sharing of tacit knowledge and collaboration,
1 European Commission (2022a).
2 In regions across the EU, the alignment of support from the ERDF with smart specialisation strategies is supporting place-based innovation and investment in line with regional business needs and opportunities. This has led to the creation of regional innovation hubs and industrial clusters based on the co-location of research infrastructures, universities, research and technology centres, and industry (e.g. Grenoble, Hamburg and Brno). Thematic smart specialisation platforms and partnerships have also become important means of connecting innova- tors with similar or complementary strengths in different parts of the EU, including in technology areas that are key to the twin green and digital transitions. Over the last six years, 37 inter-regional partnerships involving 180 regions in 33 EU and non-EU countries have provid- ed such support in areas such as advanced battery materials, and hydrogen and fuel cell technology.
3World Economic Forum (2019).
4The EU has a strong overall innovation performance but lags behind China in investment in intangibles and patent activities relating to digitalisation (European Commission, 2022b). While the EU is strong in advanced manufacturing and advanced materials (in terms of both publications and patent applications), its production, design and capacity are less strong in other areas, including artificial intelligence (AI), big data, cloud computing, cybersecurity, robotics and micro-electronics (European Commission, 2021b, 2022b).
5European Commission (2022b).
6 Porter (1998).
and attract a pool of skills that serve to increase innovation7. The formation of such a cluster is also influenced by the ‘quality’ of the location, by the amenities available and the business envi- ronment8. The positive externalities generated by innovation clusters tend to have multiplier effects on local employment and income, so reinforcing the benefits of attracting high-skilled jobs and the people to fill them9. In sum, the fact that innova- tion tends to agglomerate in specific areas high- lights the importance of understanding its spatial, social and economic dimensions, with a view to developing a balanced policy mix that promotes economic cohesion as well as innovation.
Place-based approaches can tailor policies to fos- ter the potential of regions to innovate in line with their strengths and characteristics. Investment in research and development (R&D) can stimulate innovation in more developed regions, with im- portant benefits for neighbouring regions. On the other hand, for less developed regions, policies tar- geted at education, skills and training are needed to foster innovation10. The quality of institutions is also important for regions at all stages of devel- opment to successfully participate in competitive research programmes11. Creating collaborative networks between lagging regions and innovation hubs can facilitate knowledge transfer and provide opportunities for shared learning12. For regions struggling to keep pace with innovation hubs, it is important to identify economic sectors where they have a comparative advantage and introduce tailor-made policies that help to develop these13. Such an approach can involve support for the cre- ation of clusters to unleash agglomeration forces and to focus on linked economic activities with ap- propriate degrees of complexity14. All this implies that a differentiated, place-based approach to fos-
tering innovation is essential for promoting eco- nomic convergence across regions and reducing the innovation divide.
This chapter presents an overview of regional in- novation and digital performance across Europe and the future potential. Section 2 sets out indica- tors of innovation, such as education, expenditure on R&D, patent applications and the Regional In- novation Scoreboard. Section 3 gives an overview of digital accessibility across regions. Section 4 indicates how cross-border co-patenting and spe- cialisation in sectors where regions have potential strengths can help them to take advantage of the opportunities offered by the digital transition and reduce the risk of a digital and innovation divide. Section 5 assesses how foreign direct investment (FDI) and access to finance can foster innovation and integration into global value chains.
2.The geography of innovation in Europe: education, R&D, patent applications, and the Regional Innovation Scoreboard
Innovation can take many forms and assessing it requires a holistic approach that covers the main dimensions. Measuring innovation is a widely ac- knowledged challenge15. This is particularly true in respect of the regional context, which highlights the need for better territorial data on innovation. This section provides a snapshot of regional inno- vation in the EU by reviewing the main indicators: tertiary education, expenditure on R&D, patent applications, and the Regional Innovation Score- board, a composite indicator capturing several di- mensions of innovation.
207
7Rosenthal and Strange (2003).
8Chatterjee and Sampson (2015). 9 Moretti (2010).
10 Rodríguez-Pose and Crescenzi (2008).
11 Peiffer-Smadja et al. (2023). 12 Foray (2009).
13McCann and Ortega-Argilés (2015).
14Delgado, Porter and Stern (2010); Boschma (2015).
15OECD and Eurostat (2018).
Figure 5.1 Share of population aged 30–34 with tertiary education, in the EU-27 Member States, and NUTS 2 regions, 2021
Capital region
National average
Other NUTS 2 regions
EU-27
80
70
% of population aged 30–34
60
50
40
30
20
10
LU IE CY LT NL DK SE BE FR SI LV ES PL FI EL PT MT EE AT SK DE CZ HU HR BG IT RO
Source: Eurostat.
2.1Regional education systems and attainment
thrive, so underpinning sustainable and inclusive long-term development18.
208
Education plays a pivotal role in fostering innova- tion. A well educated population is a prerequisite for sustained innovation and long-term economic development. Numerous studies underline the cor- relation between education, creativity, entrepre- neurship and innovative capacity, emphasising the multi-faceted nature of the innovation process16. Investment in education is needed to ensure a skilled, resilient and adaptable workforce, and to nurture a culture of innovation conducive to eco- nomic development. Investment needs to cover all levels of education, starting from early childhood. The work of Nobel laureate James Heckman has highlighted the long-term impact of early educa- tion on cognitive abilities and has found that the economic and social returns of investing in ear- ly childhood and care vastly outweigh the cost17. A highly skilled and educated population, capable of critical thinking and problem-solving, creates an environment where creativity and innovation can
There are wide variations across EU regions in the share of people with tertiary education, reflecting a tendency for them to concentrate in more de- veloped and metropolitan regions. Overall, around 37 % of the population aged 25–64 in more de- veloped regions in the EU had tertiary education as against 25 % in less developed ones. The pro- portion increased in all regions over the 2011– 2021 period, though regional differences have re- mained19. Taking those aged 30–34 only to reflect the most recent developments, in some regions around 70 % or more of people in this age group in 2021 had tertiary education (e.g. in the capi- tal city regions of Denmark, Lithuania or Poland), whereas in other regions, the share was less than 20 % (e.g. Sud-Est in Romania or Sicilia in Italy; Figure 5.1).
16See Biasi et al. (2021) and the discussion in Section 3 of Chapter 6 on education and the risk of falling into a talent development trap.
17Garcia et al. (2020).
18In a review of the literature, Biasi et al. (2021) find that improvements in the accessibility and quality of education have great potential to encourage entrepreneurship and innovation. This happens largely through two channels. First, education helps those who would have been innovators anyway (because of innate traits) to become more successful. Second, and more importantly, education enables individuals who would not have otherwise become innovators to fulfil their potential.
19 European Commission (2023a).
Canarias
Guadeloupe Martinique
Guyane
Mayotte Réunion
209
REGIOgis
Map 5.1 Expenditure on R&D in NUTS 2 regions as a % of GDP, 2021
% of regional GDP
< 0.5
0.5 – 1
1 – 2
2 – 3
3 – 4
> = 4
no data
EU-27 = 2.3
The EU-2020 target is 3 %. DK: 2019.
Source: DG REGIO based on Eurostat data (rd_e_gerdreg).
0
500 km
© EuroGeographics Association for the administrative boundaries
Figure 5.2 Expenditure on R&D in EU Member States as a % of GDP, 2001 and 2021
4.0
2001
2021
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
EU- SE BE AT DE FI DK NL FR SI CZ EE PT HU IT EL PL ES HR LT IE LU SK CY BG LV MT RO
27
Note: The 2001 figure for LU relates to 2000, for MT and HR to 2002. Source: Eurostat [rd_e_gerdtot] and DG REGIO calculations.
210
2.2
Regional R&D expenditure
Spending on R&D in relation to GDP is also concen- trated in more developed regions. Though this Is another widely used indicator of innovation capaci- ty, it is really a measure of input into the innovation process, or the effort made, rather than of output. It is also likely to underestimate innovation activi- ty, especially in sectors outside of manufacturing, where non-technological and non-research-based innovation is common and where expenditure on R&D is hard to define and identify (such as in re- spect of computer software programmes). In 2021, expenditure in the EU amounted to 2.3 % of GDP (Map 5.1) and increased by 0.5 pp over the preced- ing two decades (from 1.8 % of GDP in 2001). In most Member States, expenditure remained well below that in other developed economies, especial- ly Japan or the US (where it was above 3 % of GDP, which has been set as a target for the EU).
There is also no evidence of convergence in spend- ing within the EU over the past 20 years. Indeed, countries with the lowest R&D expenditure in 2001 recorded the smallest increase, resulting in a widening gap. Expenditure in the north-west of the EU (averaging 2.5 % of GDP in 2021) was al- most twice as high as in the east (1.3 %), with the south having only a slightly higher level than the latter (1.5 %).
At the NUTS 2 level, spending was above 3 % of GDP only in more developed regions and above 4 % only in a handful of regions, many of them located in the south of Germany, a centre for advanced manufacturing (Figure 5.2). The highest level of R&D expenditure within countries is in many cases in capital city regions, Belgium, Germany and Italy being notable exceptions.
2.3Regional patent applications
Patent applications are one of the few tangible means of comparing performance in innovation between regions, though they give only a very rough estimate of actual innovation activity. Inno- vations registered with the European Patent Office, the most common indicator, relate predominant- ly to those arising within manufacturing. How- ever, many innovations arising in services, which account for around 75 % of EU gross value add- ed, remain unpatented as they are intangible or non-codifiable (e.g. work organisation or computer programming).
Nevertheless, despite their limitations, as not- ed above, patents provide one of the only tangi- ble means of comparing technological innovation across regions. Over the period 2018–2019,
124 patent applications per million inhabitants were registered at the European Patent Office
Box 5.1 Synergies between Horizon 2020 and Cohesion Policy
Synergies among different EU funds to support inno- vation are important to foster regional development. As indicated in Chapter 9, a substantial amount of EU Cohesion Policy funding goes to supporting R&I through place-based programmes co-managed at the regional level. A large part goes to less devel- oped regions. By contrast, funding from Horizon 2020, the EU programme for supporting R&D, is highly concentrated in the more developed regions1. This reflects the nature of the selection process, which is highly competitive and is aimed at reward- ing excellence2.
Using econometric methods, Peiffer-Smadja et al. (2023) analyse the factors affecting success in respect of Horizon 2020. The results show that critical mass in terms of R&D expenditure, human resources, and research outputs is needed for a region to succeed in obtaining funding. The study finds that regions with low R&D spending could in- crease their success rate by improving institutional quality, though regions with higher levels could also benefit3. The findings highlight the importance of considering a holistic approach that takes account
of several factors at the same time (especially, eco- nomic development, human capabilities and quality of institutions). In the light of the findings, the au- thors suggest that success rates of less developed regions could be improved by supporting and facili- tating collaboration with more advanced regions, in line with their strengths and areas of specialisation, as reflected in their smart specialisation strategies (see Box 5.2).
Recently, significant efforts have been set in place to build stronger synergies between Horizon Europe and the ERDF. Acknowledging some of the legal and practical difficulties of building synergies between Horizon 2020 and the ERDF, the Commission ser- vices in the current multiannual financial framework have resolved some of the legal provisions that hin- dered the creation of synergies in practice and pub- lished practical guidance to implement synergies. In addition, an expert group has been set up that provides analysis and advice on how to overcome persistent difficulties in the implementation of these synergies.
1 Peiffer-Smadja et al. (2023); European Commission (2017); Balland et al. (2019); Protogerou et al. (2010); Enger (2018). Peiffer-Sm- adja et al. (2023) examined the success of regions in participating in Horizon 2020, measured as the number of successful proposals in relation to the total number submitted. The highest success rates (over 18 % of proposals submitted) are in western and northern regions in France, the Netherlands, Austria and Sweden. Interestingly, German regions, with high R&I performance in terms of R&D expenditure and patent applications, have lower (moderate to high) success rates. The lowest success rates (below 10 % of proposals submitted) are in regions in southern and eastern Member States, in Italy, Poland, Hungary, Slovakia and Bulgaria.
2 Horizon 2020 provided financing of EUR 80 billion for R&I in the EU over the 2014–2020 period, most being allocated following an open, competitive process. This resulted in funding being concentrated on a relatively small pool of beneficiaries: see European Com- mission (2017); Balland et al. (2019); Protogerou et al. (2010); Enger (2018).
3 For all regions, a focus on the quality of research outputs, such as scientific publications and patents, rather than on the quantity, appears to be important to be recognised as a partner in international R&I projects, particularly those aimed at tackling societal chal- lenges. For more advanced regions, investing in R&D and in science and technology specialists also seems to increase the chances of participating in Horizon projects.
211
(Map 5.2). Most applications came from regions in the north-western Member States and in north- ern Italy. At the NUTS 3 level, the top-performing regions are, in many cases, those hosting large corporations20. The spatial distribution suggests an innovation divide between regions in the most developed Member States and others.
Metropolitan areas tend to offer an environment that is particularly conducive to the development of new ideas, products and processes. Applications for patents are accordingly much higher there than elsewhere (Figure 5.3). A vast literature explains the reasons for this – the presence of a creative and skilled workforce and specialised clusters of eco- nomic activity, universities and research centres21.
20For instance, the three top-performing regions in the EU are Erlangen in Germany (1 209 patents per inhabitant), home to a major Siemens site, Zuidoost-Noord-Brabant in the Netherlands (973), home to Philips, and Ludwigshafen in Germany, home to BASF (961).
21European Commission and UN-HABITAT (2016).
Canarias
Guadeloupe Martinique
Guyane
Mayotte Réunion
Açores
Madeira
REGIOgis
Map 5.2 Patent applications to the European Patent Office, average 2018–2019
Applications per million inhabitants
< 25
25 – 50
50 – 100
100 – 150
150 – 250
>=250
EU-27 = 125.6
Sources: DG REGIO based on OECD REGPAT database August 2023 and Eurostat population data (nama_10r_3popgdp).
0
500 km
© EuroGeographics Association for the administrative boundaries
Figure 5.3 Patent applications to the European Patent Office by type of region, 2017–2018
10 000
Capital metro region
EU-27 average
Other metro regions
Non-metro regions
National average
Per million people (log scale)
1 000
100
10
1
LU SE FI DK DE NL AT IE BE FR MT SI CY EE ES PT CZ LT LV HU PL EL SK BG HR
RO
0
Source: OECD REGPAT and DG REGIO calculations.
Metro region population:
< 500 000
500 000 – 1 000 000
1 000 000 – 2 500 000
> 2 500 000
Capital metropolitan regions, in most cases, have the highest rates of applications in nearly all Mem- ber States. The only exceptions are Vienna and Lisbon. Only in a very few cases are applications in metropolitan regions below those in others in the same country. It should be noted as well that a larger number of skilled immigrants also tends to increase patents filed, and return migration of those concerned might boost patenting, and inno- vation, in the country of origin22.
2.4The Regional Innovation Scoreboard
The Regional Innovation Scoreboard (RIS) for 2023 highlights the key role played by innovation in re- gional development and a persistent divide in in- novation performance23. The RIS, an extension of the European Innovation Scoreboard (EIS), meas- ures the innovation performance of regions on the basis of a sub-set of indicators included in the EIS. Despite some regional variation within coun- tries, the ranking of regions largely matches that of Member States (Map 5.3), suggesting that indi- cator values at the regional level are affected by
national characteristics or policies (e.g. most R&D support schemes are national). Most regional ‘in- novation leaders’ are in countries also identified as ‘innovation leaders’ or as ‘strong innovators’, and almost all the regional ‘moderate’ and ‘modest’ innovators are in countries classified in the same way. However, there are regional ‘pockets of ex- cellence’ in some ‘moderate innovator’ countries, including capital city regions in Czechia, Lithuania and Spain, as well as País Vasco in the last. Con- versely, there are many regions in ‘strong innova- tion’ countries that lag behind.
There is a close relationship between the lev- el of development of regions and the innova- tion score (Figure 5.4). In less developed regions, an increasing proportion of the population live in ‘emerging innovator’ regions (i.e. the bot- tom category) rather than ‘moderate innovating’ ones – 60 % in 2021, twice as much as in 2016, indicating that the innovation performance of the regions concerned has worsened over time. At the same time, in both southern and eastern re- gions, there was an increase in the share of people
213
22Kerr and Lincoln (2010); Fry (2023).
23The RIS 2023 follows the same methodology as the EIS in the same year to develop a composite indicator of 21 different indicators of regional innovation. Regions are classified into four innovation performance groups according to this: innovation leaders (36 regions), strong innovators (70 regions), moderate innovators (69 regions), and emerging innovators (64 regions). For a list of the 21 indicators used, see Table 4 (page 17) of the RIS methodological report (https://research-and-in-no
vation.ec.europa.eu/system/files/2023-07/ec_rtd_
ris-2023-methodology-report.pdf).
Canarias
Guadeloupe Martinique
Guyane
Mayotte Réunion
Açores
Madeira
REGIOgis
Map 5.3 Regional Innovation Scoreboard, 2023
Emerging innovator - Emerging innovator Emerging innovator + Moderate innovator - Moderate innovator Moderate innovator +
Strong innovator - Strong innovator Strong innovator + Innovation leader - Innovation leader Innovation leader +
Source: European Commission – Regional Innovation Scoreboard 2023 and European Innovation Scoreboard 2023.
0
500 km
© EuroGeographics Association for the administrative boundaries
living in ‘strong innovator’ regions. Nevertheless, innovation leaders have remained largely clustered in the more developed, north-western regions.
In general, the RIS confirms the wide diversity of EU regions in terms of innovation performance, so highlighting the strong regional dimension of
innovation. Because of this, measures supporting innovation, including Cohesion Policy programmes, need to take explicit account of the regional con- text when considering the most useful kind of sup- port to provide. As it is inherently place-based, the smart specialisation approach helps in this regard.
Figure 5.4 Share of EU population by RIS category, level of development and geographic group of Member States, 2016 and 2023
a)Share of EU population by RIS category and level of development, 2016 and 2023
100
Emerging Innovator
Moderate Innovator
Strong Innovator
Innovation leader
% of total in the respective area
80
60
40
20
215
0
b)Share of EU population by RIS category and geographic group of Member States, 2016 and 2023
100
Emerging Innovator
Moderate Innovator
Strong Innovator
Innovation leader
% of total in the respective area
80
60
40
20
0
Note: In cases where the RIS score is only available at NUTS1 level, it is assumed that the same score applies to the constituent NUTS2 regions. Calculations for both years are based on 2021 population data and level of development classification.
Source: Regional Innovation Scoreboard 2023 and DG REGIO calculations.
Box 5.2 Smart specialisation: strengthening industrial and innovation ecosystems
Smart specialisation strategies are part of Cohesion Policy intended to foster regional innovation eco- systems. They do so by building on the ‘partnership approach’ of Cohesion Policy and enabling regions to develop a regional innovation strategy that builds on their assets and strengths. Smart specialisation strategies are structured around three pillars: loca- tion (place-based approach), prioritisation (making strategic choices), and participation (stakeholders’ involvement). Smart specialisation has a strong ‘re- gional development’ objective. Around 85 % of the overall financial allocation for 2014–2020 (about
€40 billion) was concentrated in less developed and transition regions where it is often the main source of innovation support. Periañez-Forte et al. (2021) have carried out case studies to assess the lessons learned during the setting-up of governance struc- tures and have underlined the importance of these for the success of the policy.
In the 2021–2027 programming period, smart
specialisation strategies remain the key require-
ment for Cohesion Policy support for R&I. A total of EUR 34.5 billion is currently programmed for support of R&I investment, in line with 175 smart speciali- sation strategies in EU regions and Member States.
Thematic smart specialisation platforms and part- nerships are key means of bringing together inno- vators with similar or complementary strengths and priorities in areas that are important for strengthen- ing regional ecosystems while addressing EU prior- ities, notably in the context of the digital and green transitions. These include hydrogen, bioeconomy, healthcare and AI. At present, there are 38 partner- ships covering 191 regions in all 27 Member States and nine non-EU countries.
The interregional innovation investment instrument (‘I3’) under Cohesion Policy helps to support existing efforts to strengthen value chains and to link region- al industrial and innovation ecosystems in less de- veloped regions with complementary ones in more developed regions.
216
3.Harnessing the potential
of the digital transition: digital skills, accessibility, and firm take-up of digital technologies
The last decade has seen a rapid increase in the adoption of digital technologies by businesses, people, and governments alike. In the health sec- tor, for instance, digitalisation became a crucial el- ement in the reorganisation of service-provision in the wake of the pandemic, with regional and local health authorities at the forefront of this process in several countries across Europe. More broadly, companies have increased investment in ICT sub- stantially in recent years and this digital transition has greatly accelerated with the COVID-19 pan-
demic24 , with significant national and EU investments put forward to also improve the digital skills of students and teachers. The evidence suggests that digitalisation has increased the productivity of businesses, im- proving their efficiency, and stimulating domestic sales and exports25. While the impact on business- es has been positive, the overall impact on local economies and people, both up to now and in the future, is more difficult to assess. Recent studies indicate that while it has been generally positive for the EU, the effect has varied across regions de- pending on the structure of their economies and skills of the workforce26.
Access to a sufficiently fast internet connection is essential for ensuring that all regions can harness the potential of the digital transition27. The accel- eration of digitalisation in both the private and public sectors across the EU, as a result of the
24European Investment Bank (2021).
25Rossato and Castellani (2020); Cincera et al. (2020); Eduardsen (2018).
26Marques Santos et al. (2023); see Box 5.3.
27Batista e Silva and Dijkstra (2024).
COVID-19 pandemic28, is evident in the improve- ment in broadband connectivity in most regions. The performance of fixed networks has improved in all Member States over the past three years but remains highly variable within them, with Greece, Cyprus and Croatia having the lowest speeds (Fig- ure 5.5 and Figure 5.6). Capital city regions gener- ally have the highest speeds, but with exceptions (France, the Netherlands and Germany).
At the national level, France, Denmark, Spain and Romania have average speeds above 200 Mbps, although several regions in these countries have lower speeds, particularly in France). Over the three years 2020–2023, average speeds increased in all Member States. This is especially so in Cyprus and Greece, with over 70 % of the population being able to access good network speeds in 2023 as against zero in 2020. Speeds also increased signif- icantly in Denmark, Spain and France, with around 80 % of the population being able to access net- work speeds of above 190 Mbps.
ies, but with marked differences between them, those in central and south-east Europe generally having much lower speeds (Map 5.4). In several countries, the biggest increase in speed has been in rural areas (in Estonia, France, Italy and Poland, especially), reflecting the effort made to bridge the digital gap between regions across the EU, though gaps still remain, especially in terms of access to very-high-capacity networks for rural areas29.
At a more detailed level, large variations in network speed are evident between municipalities. (Map 5.5, which shows the average speed in local adminis- trative units – LAUS)30. This is particularly so in Spain, France and Romania, where speeds are part- ly correlated with population density (see Chap- ter 3). On the other hand, speeds are more similar between municipalities in Greece, Bulgaria and Aus- tria, with low average speeds, and in the Nether- lands (with a speed of over 200 Mbps), while in Ire- land, Poland and Italy, the variation in speeds across the country reflects the distribution of urban areas.
Significant differences exist between places within each country. While broadband speeds have gen- erally increased, they have done so more in cit-
Besides access to high-speed broadband, the take-up of digital technologies by EU firms is a precondition for taking advantage of the potential
217
Figure 5.5 Average download speed per Member State and NUTS 2 region calculated for the fixed network, Q1.2023
Capital region
National average
EU-27
Other NUTS 2 regions
300
Average speed fixed network (Mbps)
250
200
150
100
50
0
FR DK ES RO HU LU NL PT PL SE MT FI LT IE BE IT DE LV SK SI BG EE CZ HR AT CY EL
Source: DG REGIO calculations on Batista e Silva, Dijkstra and Sulis, 2024.
28 OECD (2020).
29The data on broadband fixed network speed is available at the EU rural observatory.
30Sulis and Perpina (2022); Melchiorri et al. (forthcoming).
Figure 5.6 Share of population with access to fixed broadband network at different speeds (Mbps) in Member State, 2020 (left panel) and 2023 (right panel)
218
Acces to fixed broadband network–2020
0 to 40 Mbps
40 to 190 Mbps
over 190 Mbps
HR
|
|
EL
|
|
|
|
|
|
SK
|
|
|
|
|
|
CY
|
|
|
|
|
|
CZ
|
|
|
|
|
|
BG
|
|
|
|
|
|
LV
|
|
|
|
|
|
AT
|
|
|
|
|
|
FR
|
|
|
|
|
|
IT
|
|
|
|
|
|
IE
|
|
|
|
|
|
HU
|
|
|
|
|
|
PL
|
|
|
|
|
|
PT
|
|
|
|
|
|
RO
|
|
|
|
|
|
DE
|
|
|
|
|
|
SI ES FI EE BE SE NL MT LU LT DK
Acces to fixed broadband network–2023
0 to 40 Mbps
40 to 190 Mbps
over 190 Mbps
HR EL SK CY CZ BG LV AT FR IT IE HU PL PT RO DE SI ES FI EE BE SE NL MT LU LT DK
0
20
40
60
80
100
% population
0
20
40
60
80
100
% population
Source: DG REGIO calculations on Batista e Silva, Dijkstra and Sulis, 2024.
of the digital transition, which can increase effi- ciency, improve the accessibility of services and help to maintain competitiveness. As part of the digital transition, a goal of the EU is that by 2030, 75 % of businesses in the EU will have taken up three digital technologies, cloud computing, use of big data and AI. In 2021, over 40 % of business- es had adopted cloud computing, while only 15 % were using big data and under 10 % AI (Figure 5.7). The difference may be because of the newness of the latter two and their possibly less general ap-
plicability at the time. For all three technologies, however, the take-up was much greater, on aver- age, in north-western Member States than in other parts of the EU, especially in the eastern countries.
As the digital transition in the EU takes place, dig- ital skills will become increasingly important for labour market participation and inclusion. In 2021, over 60 % of EU enterprises that tried to fill va- cancies for ICT specialists had difficulties. The EU has set the target that, by 2030, at least 80 %
Canarias
Canarias
Guadeloupe Martinique
Guyane
Guadeloupe Martinique
Guyane
Mayotte Réunion
Mayotte Réunion
Açores
Madeira
Açores
REGIOgis
REGIOgis
Map 5.4 Internet fixed network speed in Functional Urban Areas, Q1 2023
Percentage deviation from EU median
Population size
Map 5.5 Average speed for fixed network at municipality level (LAU), 2023
Average speed (Mbps) fixed network
Chapter 5: Regional innovation and the digital transition
< -50
-30- – 50
-5- – 30
-5 – 5
30 – 5
50 – 30
> 50
no data
< 000 100
000 250 – 000 100
000 500 – 000 250
000 000 1 – 000 500
000 000 5 – 000 000 1
>= 000 000 5
Deviation relative to EU median (weighted). Source: JRC.
0
500 km
© EuroGeographics Association for the administrative boundaries
<= 30
30 – 100
100 – 200
> 200
no data
Source: JRC.
0
500 km
© EuroGeographics Association for the administrative boundaries
Figure 5.7 EU enterprise take-up of digital technologies, 2021
North-western
Southern
Eastern
2030 target
EU-27
75
% of enterprises that have taken up the technology
60
45
30
15
0
Cloud computing services
Big data analysis (2020)
Artificial intelligence
Note: All EU enterprises outside the financial sector with 10 or more persons employed are covered (Eurostat code 10_C10_S951_XK). Source: Eurostat [isoc_eb] and DG REGIO calculations.
220
of the adult population should have basic digital skills31. In 2021, this was the case for only 54 % of people aged 16 to 74, well below the target, with major differences between countries, rates ranging from 79 % in Finland and the Netherlands to only 28 % in Romania. Throughout the EU, people living in cities (61 %) are more likely to have at least basic digital skills than those in towns and suburbs (52 %) and rural areas (46 %). While no data on basic digital skills are available at regional level, there are major differences between regions in the extent to which people use the internet on a daily basis, participate in online social networks, use internet banking and take part in e-commerce32. The number of ICT specialists in the EU is estimated to be around 12 million, well below the target of 20 million for 2030 set in the EU’s ‘2030 digital decade’33. Here as well, there are major differences across countries, with Greece and Romania among the countries with the lowest percentage of ICT specialists (respectively 2.5 % and 2.8 % of total employment). Meanwhile, Sweden, Luxembourg
and Finland are the countries with the biggest share of ICT specialists (respectively 8.6 %, 7.7 % and 7.6 % of total employment).
4.Synergies to harness the potential of the digital transition across regions: the role of
cross-border co-operation
Cross-border innovation activity has increased in the EU over time but there is much room for further growth. A useful indicator of regional synergies in R&I is co-patenting. This has increased dramatically in Europe over the past four decades, rising from 1 000 co-patents in 1980 to over 100 000 in 2020. However, most co-patents are filed between firms or organisations located in the same region – around 75 % over the period 1980–2020. Almost 20 % were between organisations in different regions but in the same country and 7 % involved organisa- tions in different European countries (Map 5.6).
31 See ‘digital compass’ of the ‘2030 digital decade’ and European Pillar of Social Rights action plan. Overall digital skills refer to five aspects: information and data literacy skills, communication and collaboration skills, digital content creation skills, safety skills and problem-solving skills, which are covered by the revised digital competence framework (DIGCOMP 2.0). To have at least basic overall digital skills, people need to know how to do at least one activity in each area. See Eurostat:
https://ec.europa.eu/eurostat/web/products-eurostat-news/-/ddn-
20220330-1.
32 In 2022, only 7 % of people aged 16–74 in the EU never used the internet, though with major regional differences. In three regions in Sweden (Sydsverige, Stockholm and Småland med öarna.) only 1 % never use the internet., while in Norte (Portugal), the figure was 18 %, in Calabria (Italy), 19 % and in Kentriki Elláda (Greece), 20 %.
33 European Commission (2023b).
Map 5.6 Inter-regional cooperation in innovation and digital technologies
Inter-countries co-patents, 1979–2020
Number of patents
Actual inter-regional collaborations in digital technologies
Potential inter-regional collaborations in digital technologies
<= 80
|
800 – 1 200
|
|
cross-border
|
|
cross-border
|
|
80 – 200
|
1 200 – 2 000
|
Source: DG Research and Innovation -
Common R&I Stratey and Foresight
|
within country
|
Source: Bachtrögler-Unger
et al. (2023)
|
within country
|
Source: Bachtrögler-Unger
et al. (2023).
|
200 – 450
450 – 800
> 2 000
no data
Service Chief Economist Unit based on OECD-REGPAT data.
NUTS 2 regions
NUTS 2 regions
Chapter 5: Regional innovation and the digital transition
0
1 000 km
© EuroGeographics Association for the administrative boundaries
Box 5.3 Job creation and destruction in the digital age:
assessing heterogeneous effects across Member States
In contrast to the potentially positive effects on the competitiveness of firms, many authors have ar- gued that technological change can be detrimen- tal to labour market conditions. According to Ford (2015) and Acemoglu and Restrepo (2020), for in- stance, automation and robots may replace work- ers and lead to job destruction. On the other hand, according to others, digitalisation may create new job opportunities as new technologies are adopted1.
Changes in the structure of the labour market in- duced by digital technologies have been studied empirically using both micro-economic and mac- ro-economic data2. Findings on the net effect of digitalisation on employment are mixed. A major- ity of studies suggest it may increase high-skilled employment (complementarity effect) and reduce low-skilled employment (substitution effect). The net effect Is likely to depend on the economic charac- teristics of each country, on its knowledge capacity, sectoral composition, and capacity to upskill or reskill the workforce as the structure of activity changes.
As a corollary, regions and countries will tend to be affected differentially by the digital transition.
Marques Santos et al. (2023) have examined whether ICT investment was associated with an increase or decrease in labour demand in Member States between 1995 and 2019. They find an over- all positive effect on total employment over the period, but not in all Member States. This suggests that studies of different countries may yield differ- ent results because of the structural characteris- tics of the economy and that conclusions based on case studies may not hold generally. This suggests that studies of different countries may yield differ- ent results because of the structural characteris- tics of economies and that conclusions based on case studies may not hold generally. At the same time, the findings underline the importance of in- vestigating further the spatial and sectoral impact of digitisation and taking account of the specif- ic economic and employment features of places when formulating policy recommendations.
222
1
Degryse (2016
).
2 For a review, see Marques Santos et al. (2023).
Of the latter, the vast majority involved organi- sations in cross-border regions, notably along the Rhein valley connecting German, Belgian, French and Swiss regions, though also in capital city re- gions with a track record of patenting activity. The importance of physical proximity for co-inno- vation is well established, but the strong national bias in inter-regional collaboration in co-patenting limits the potential to co-operate in the EU Sin- gle Market. One way of overcoming this bias is to strengthen inter-regional knowledge flows and to promote co-operation in innovation between lead- ing and lagging regions, such as through the im- plementation of smart specialisation strategies34 (Section 3). In this way, the untapped potential for cross-border co-operation could be realised (see Box 5.4).
5.
Foreign direct investment (FDI) and access to finance as key drivers of innovation at regional level
FDI is an important means of fostering innovation both directly and indirectly. Direct means are when foreign firms bring new products, technologies or processes into the host economy. In these cases, foreign firms often pay higher wages, have higher levels of productivity and innovate more than do- mestic firms35, as well as opening new direct links to global value chains36. Indirect means are when there are knowledge and technology spill-overs to local firms, or workers move from foreign-owned firms to domestic ones, bringing know-how and new ideas with them.
34Balland and Boschma (2021).
35OECD (forthcoming).
36Comotti, Crescenzi and Iammarino (2020).
Box 5.4 Related variety, complexity and the regional potential for the digital transition and cross-border co-operation
There is significant untapped potential in green and digital technologies. A number of studies have de- veloped a method of identifying the opportunities for regions to diversify, given the capabilities they have accumulated in the past: Balland et al. (2019); Hartmann et al. (2021). They condition which de- velopment paths a region is most likely to follow.
Using a framework based on the notions of ‘relat- edness’ and ‘complexity’, Bachtrögler-Unger et al. (2023) determine whether regions have opportuni- ties to diversify into more complex activities linked to the digital transition as well as the technologies needed for the green transition. The results show that more developed regions are more likely to
Figure 5.8 Potential of more developed EU regions to develop twin transition technologies
Digital transition
Green transition
Low relatedness - High complexity
High relatedness - High complexity
100
90
80
70
60
Virtual Reality and
Augmented Reality
Advanced materials/
Big data
Artificial intelligence
Photonics
Internet of things
Cloud and edge
computing
5G
Autonomous mobility
50
40
30
Bio fertilizers
20
Biocides
nanomaterials
Biofuels
Drones
Solar energy
Battery technology
Electric vehicles
Broadband
Sustainable packaging
10
0
Greenhouse gas capture
Heat pumps
HVAC systems
223
0
10
20
30
40
50
60
70
80
90
100
Low relatedness - Low complexity
Source: Bachtrögler-Unger et al. (2023).
Relatedness density
High relatedness - Low complexity
Figure 5.9 Potential of less developed EU regions to develop twin transition technologies
100
Digital transition
Green transition
Low relatedness - High complexity
High relatedness - High complexity Big data
90
Virtual reality and
80
augmented reality
70
5G
60
Photonics
Internet of things
Artificial intelligence
Cloud and edge computing
Battery technology
50
40
Electric vehicles
30
Additive manufacturing
20
(3D printing)
Autonomous mobility
Drones
Solar energy
HVAC systems
Advanced materials/nanomaterials
Smart farming
Biofuels
Biocides
Bio fertilizers
10
Greenhouse gas
0
Sustainable packaging Heat pumps
0
10
20
30
40
50
60
70
80
90
100
Low relatedness - Low complexity
High relatedness - Low complexity Relatedness density
Source: Bachtrögler-Unger et al. (2023).
224
specialise in digital technologies and benefit from the digital transition, but less developed regions are well placed to develop the technologies and activi- ties relating to the green transition.
For both types of region, there appears to be large untapped potential for cross-border co-operation. Figures 5.8 and 5.9 show the technology opportuni- ties from the twin transition for more and less de- veloped regions, with the relatedness of patents to existing technologies on the horizontal axis and the level of complexity on the vertical axis1. The blue dots represent digital technologies, the green ones green technologies, their size indicating regional comparative advantage in the technology relative to other regions. On average, more developed regions have high potential in the different technologies. Their highest digital potential is in complex technol- ogies (such as 5G), the lowest in low-complex ones. The picture is similar for green technologies, with strong capability in electric vehicles, battery tech- nology and solar energy. Less developed regions have low patent activity in both areas. While, how- ever, their potential for complex digital technologies is limited, they appear to have high potential in a wide range of green technologies, such as biocides, biofertilisers, geothermal energy, biofuels, waste management and recycling.
There is substantial untapped potential for cross-bor- der co-operation across EU regions in developing the value chains needed for the green and digital transitions. Bachtrögler-Unger et al. (2023) exam- ined whether regions are connected to the right set of other regions to develop the next generation technologies, in the sense of the regions that can give them access to the complementary capabilities needed to develop them. The study compared the ideal collaboration network in which complementar- ities across regions are fully exploited with the cur- rent state of collaboration (as indicated by co-inven- tor linkages) in the technological areas concerned. shows the three strongest actual collaborations in digital technologies of each region with others and the three inter-regional linkages that represent the largest untapped potential (based on complemen- tarities). Intra-country linkages are coloured in red, cross-border ones in yellow. The actual inter-region- al collaborations show a clear national bias, while the largest untapped potential is for cross-border collaborations. This applies for both more developed and less developed regions.
1 Bachtrögler-Unger et al. (2023).
An appropriate place-sensitive approach is im- portant for FDI to have positive spill-over effects. According to a study of manufacturing firms in six Member States, productivity spill-overs can be positive, non-existent, or even negative, depend- ing on how close the firms in a given sector are in technology terms37. Embedding FDI can benefit lo- cal communities but requires additional elements to ensure firms ‘stick’ to places38. The public sector and the third sector can play an important role in this by setting the right framework conditions and
generating incentives to co-create value-added
with local firms39.
Co-ordination across places is needed to foster the positive enablers of FDI in terms of efficient insti- tutions, a skilled workforce, an effective research environment and good connectivity. These factors play a key role in shaping regional attractiveness for foreign investors40. However, the choice of FDI location can also be motivated by less desir- able institutional settings, such as lower labour
37Positive spill-overs dominate if domestic firms are using similarly advanced technologies to the foreign firm and operate in the same sector (Fons-Rosen et al., 2018[9]) or in other sectors (Lembcke and Wildnerova, 2020[8]). Negative effects from increased competition dominate if the products of the foreign-owned company are similar to those of domestic ones (Lembcke and Wildnerova, 2020[8]).
38These elements are broadly related to the ecosystem of the firm, including links with other firms and clusters with both suppliers and cus- tomers, complementary firms and even competitors that can attract workers with the right skill set to a region.
39 Bailey and Tomlinson (2018). 40 OECD (2023).
Figure 5.10 Value of regional inward FDI by degree of regional development, NUTS 2, 2019–2022
a)Total value of regional inward FDI deals
b)Total value of regional inward greenfield FDIs
400
Transition
More developed
Less developed
Transition
More developed
Less developed
70
350
60
300
50
250
40
200
30
150
20
100
50
10
0
2019
2020
2021
2022
0
2019
2020
2021
2022
Note: The left panel includes all forms of FDI, mergers and acquisitions as well as greenfield FDI. The right panel features only greenfield FDI. Source: Martinez Cillero et al. (2024) based on Orbis M&A BvD and Orbis Crossborder BvD data.
standards41, lower tax rates or higher tax credits or subsidies42, or laxer environmental standards, es- pecially for highly polluting industries43. This points to the importance of cross-border co-ordination to ensure a level playing field for investment that minimises the risk of beggar-thy-neighbour com- petition (both domestic and foreign), while at the same time strengthening the positive enablers of investment.
Less developed and transition regions have in- creasingly attracted greenfield investment over the past few years44. Regional data on FDI enable two types to be distinguished M&A and greenfield in- vestment45. On average, 53 % of greenfield FDI in the EU over the period 2019–2022 (with an equiv- alent value of EUR 218 billion) went to less devel- oped and transition regions, increasing from 38 % in 2019 to 58 % in 2022, when transition regions
alone accounted for 36 % (Figure 5.10, right panel). Accordingly, greenfield FDI is relatively high in the eastern EU Member States and in almost all regions of Spain and Portugal, but also in Sweden, Finland, Ireland and the Benelux countries46.
By contrast, FDI in the form of M&A goes mainly to more developed regions (Figure 5.10, left pan- el). Capital city regions are major destinations, as in France, Austria, Finland, Spain, Portugal, Poland and Greece, but also regions in northern Italy, north-eastern Spain, southern France, southern and eastern Germany, the North-Rhine-Benelux area, and both sides of the Gulf of Finland. The regions with the highest level of M&A over the pe- riod are Wien (Austria), Eastern and Midland (Ire- land), Limburg and Noord-Holland (Netherlands), Madrid (Spain), Helsinki-Uusimaa (Finland), and Luxembourg.
225
41 Davies and Vadlamannati (2013); Olney (2013).
42 Desai et al. (2005); de Mooij et al. (2003).
43List and Co (2000).
44Gianelle et al. (forthcoming).
45Mergers and acquisitions (M&A) involve the acquisition of at least 10 % of the equity of a company resident in an NUTS 2 region in the EU by a company resident in another country, which may be outside the EU (portfolio investments are excluded). Greenfield investment consists of the construction by a company in another country of new facilities (sales office, manufacturing plants, etc.) or the relocation or extension of existing facilities.
46The regions with the highest levels of greenfield FDI over the period are Észak-Alföld, Közép-Dunántúl, Dél-Alföld and Pest (all in Hungary), Sachsen-Anhalt (Germany), Alentejo (Portugal), Eastern and Midland (Ireland), and Východné Slovenskom (Slovakia).
226
5.1 Access to finance and innovation
Access to finance is essential for fostering innova- tion, but firms in a number of regions find it difficult to obtain bank financing. In comparison with the US, where financial markets are more developed and the risk capital market stronger, the fragment- ed nature of financial markets in the EU poses challenges. This is especially so for less developed regions, which in many cases may lack liquid cap- ital markets and robust financial infrastructure and accordingly have many firms that are credit constrained47. In these cases, targeted support to facilitate access to finance for innovation-related investment can take the form of grants, low-inter- est loans, guarantees, or equity.
The World Bank Enterprise Survey, conducted in 2019, shows large variations between regions in access to finance. In the survey, a firm is considered to be constrained in accessing external finance if either one of two conditions hold: (1) the firm did not apply for a loan for any reason other than they did not need it; or (2) the firm applied for a loan but was rejected. Firms in many regions in east- ern and southern Member States are shown to be constrained in this way (Figure 5.11). The survey also reveals that firms are more constrained in in-
vesting in innovation if it is financed through bank loans than if it is financed through equity. The re- sult is in line with equity financing being gener- ally more suitable the higher the risk associated with the investment, encouraging a collaborative approach to risk-taking. Loans and guarantees, on the other hand, tend to be more suitable when the innovation is less risky, giving firms the financial support needed while offering a structured means for repayment.
Figure 5.11 Share of fully credit constrained firms at EU, national and NUTS 2 level, 2019 (%)
Capital region
National average
Other NUTS 2 regions
EU-27
40
Fully credit constrained (FCC)
30
20
10
0
EL RO IT BG SK PT LT CY EE LV HR LU FR HU IE SE CZ NL MT DE PL AT DK BE SI FI ES
Note: the highest point in EL is 72.8 %. Source: World Bank Enterprise Survey.
47Financial infrastructure in this context refers to the availability and efficiency of financial services, institutions, and the market generally.
References
Acemoglu, D., Restrepo, P. (2020), ‘Robots and jobs: evidence from US labor markets’, Journal of Political Economy, 128(6), pp. 2188–2244.
Bachtrögler-Unger, J., Balland, P. A., Boschma, R. and Schwab, T. (2023), ‘Technological Capabilities and the Twin Transition in Europe. Opportunities for Regional Collaboration and Economic Cohesion,’ WIFO Studies, 70743.
Bailey, D., Pitelis, C. and Tomlinson, P. (2018), ‘A place-based developmental regional industrial strategy
for sustainable capture of co-created value’, Cambridge Journal of Economics, 42(6), pp. 1521-1542.
Balland P.A, Boschma R., Ravet J. (2019), ‘Network dynamics in collaborative research in the EU’, 2003–2017, European Planning Studies, 27(9), pp. 1811–1837.
Balland, P. A. and Boschma, R. (2021), ‘Complementary interregional linkages and Smart Specialisation: an empirical study on European regions’, Regional Studies, 55(6), pp. 1059–1070.
Batista e Silva, F. and Dijkstra L. (eds.) (2024, forthcoming), Challenges and opportunities for territorial cohesion in Europe – contributions to the 9th Cohesion Report. Joint Research Centre Science for Policy report.
Biasi, B., Deming, D. J. and Moser, P. (2021), Education and innovation, 28544, National Bureau of Economic Research.
Boschma R. (2015), ‘Towards an Evolutionary Perspective on Regional Resilience’, Regional Studies,
49(5), pp. 733–751.
Chatterjee, S. and Sampson, T. (2015), ‘Spatial Quality Sorting and Innovation’, The Quarterly Journal of Economics, 130(4), pp. 2007–2055.
Cincera, M., Delanote, J., Mohen, P., Santos, A., Weiss, C. (2020), Intangible Investments and Productivity Performance, Gabinete de Estratégia e Estudos, Paper 145.
Davies, Ronald and Vadlamannati, Krishna Chaitanya (2013), ‘A race to the bottom in labor standards? An empirical investigation’, Journal of Development Economics, 103(C), pp. 1–14.
Degryse, C. (2016), Digitalisation of the Economy and its Impact on Labour Markets, European Trade Union Institute Working Paper.
Delgado, M., Porter, M. E. and Stern, S. (2010), ‘Clusters, Convergence and Economic Performance’,
Research Policy, 39(10), pp. 1291–1304.
de Mooij, R.A., Ederveen, S. (2003), ‘Taxation and Foreign Direct Investment: A Synthesis of Empirical
Research’, International Tax and Public Finance 10, pp. 673–693.
Desai, Mihir A. and Foley, C. Fritz and Hines, James Rodger (2005), The Demand for Tax Haven Operations.
Eduardsen, J. (2018), ‘Internationalisation through digitalisation: the impact of ecommerce usage on internationalisation in small– and medium–sized firms’. In: van Tulder, R., Verbeke, A., Piscitello, L. (eds.), International Business In the Information And Digital Age (Progress in International Business Research), 13, Emerald Publishing Limited, Bingley, pp. 159–186.
Enger, S. G. (2018), ‘Closed clubs: Network centrality and participation in Horizon 2020’, Science and Public Policy, 45(6), pp. 884–896.
European Commission (2017), Executive summary of the interim evaluation of Horizon 2020,
SWD (2017) 222 final.
228
European Commission (2017a), Reflection Paper on Harnessing Globalisation, Publications Office of the European Union, Luxembourg.
European Commission (2020), Science, research and innovation performance of the EU: A fair green and digital Europe, Publications Office of the European Union, Luxembourg.
European Commission (2021a), A long–term vision for the EU’s rural areas – Towards stronger, connected, resilient and prosperous rural areas by 2040, Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions, COM (2021) 345 final.
European Commission (2021b), The 2030 Digital Compass: the European way for the Digital Decade, Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions, COM (2021) 118 final.
European Commission (2022a), A New European Innovation Agenda, SWD(2022) 187 final.
European Commission (2022b), Science, research and innovation performance of the EU 2022: building a sustainable future in uncertain times, Publications Office of the European Union, Luxembourg.
European Commission (2023a), Regional Trends for Growth and Convergence in the European Union,
SWD(2023) 173 final.
European Commission (2023b), 2030 Digital Decade – Report on the state of the Digital Decade,
Publications Office of the European Union, Luxembourg.
European Commission and UN–HABITAT (2016), The State of European Cities 2016: Cities leading the way to a better future, Publications Office of the European Union, Luxembourg.
European Investment Bank (2021), Investment Report 2020/2021: Building a Smart and Green Europe
in the COVID–19 Era.
Fons-Rosen, C. et al. (2018), Foreign investment and domestic productivity: identifying knowledge
spillovers and competition effects, NBER Working Paper 23643. Foray, D. (2009), The Economics of Knowledge, MIT Press.
Ford, M. (2015), The Rise of the Robots: Technology and the Threat of Mass Unemployment,
Oneworld Publications, London.
Fry, C. V. (2023), ‘Bridging the gap: Evidence from the return migration of African scientists’,
Organization Science, 34(1), pp. 404–432.
García Jorge Luis, James J. Heckman, Duncan Ermini Leaf and María José Prados (2020), ‘Quantifying the Life–Cycle Benefits of an Influential Early–Childhood Program’, Journal of Political Economy, 128(7), pp. 2502–2541.
Gianelle, C., Guzzo, F., Barbero, J., Salotti, S. (2023), ‘The governance of regional innovation policy
and its economic implications’, Annals of Regional Science.
Guzzo, F., Gianelle, C. (2021), Assessing Smart Specialisation: Governance, Publications Office of the European Union, Luxembourg.
Hegyi, F. B., Guzzo, F., Perianez–Forte, I., Gianelle C. (2021), The Smart Specialisation Policy Experience: Perspective of National and Regional Authorities, Publications Office of the European Union, Luxembourg.
Kerr, W. R. and William F Lincoln, W. F. (2010), ‘The supply side of innovation: H–1B visa reforms and US
ethnic invention’, Journal of Labor Economics, 28(3), pp. 473–508.
Lembcke, A. and Wildnerova, L. (2020), Does FDI benefit incumbent SMEs? FDI spillovers and competition effects at the local level, OECD Regional Development Working Papers, 2020/02.
List, J. A. and Co, C. Y. (2000), ‘The Effects of Environmental Regulations on Foreign Direct Investment’, Journal of Environmental Economics and Management, Elsevier, 40(1), pp. 1–20.
Marques Santos, A., Barbero, J., Salotti, S., Conte, A. (2023), ‘Job creation and destruction in the digital age: Assessing heterogeneous effects across European Union countries’, Economic Modelling, 126, 106405.
Martinez Cillero, M., Ciani, A., Gianelle, C. (2024), The regional geography of foreign direct investment in the EU: 2019–2022, JRC Science for Policy Brief, JRC136533.
McCann, P. and Ortega–Argilés, R. (2015), ‘Smart Specialisation, Regional Growth and Applications to European Union Cohesion Policy’, Regional Studies, 49(8), pp. 1291–1302.
Melchiorri, M., Sulis, P. et al. (forthcoming), Integration of remote and social sensing data reveals uneven quality of broadband connectivity across world cities, Springer.
Moretti, E. (2010), ‘Local Multipliers’, American Economic Review, 100(2), pp. 373–377.
OECD (2020), Digital Transformation in the Age of COVID–19: Building Resilience and Bridging Divides, Digital Economy Outlook 2020 Supplement, OECD, Paris.
OECD (2023), Rethinking Regional Attractiveness in the New Global Environment, OECD Regional Development Studies, OECD Publishing, Paris.
OECD (forthcoming), Foreign Direct Investment and Global Value Chain Integration in OECD Regions,
OECD Publishing, Paris.
OECD/Eurostat (2018), Oslo Manual 2018: Guidelines for Collecting, Reporting and Using Data on Innovation, 4th Edition, The Measurement of Scientific, Technological and Innovation Activities, OECD Publishing, Paris/Eurostat, Luxembourg.
Olney, W. (2013), ‘A race to the bottom? Employment protection and foreign direct investment’,
Journal of International Economics, 91(2), pp 191–203.
Peiffer–Smadja, O., Mitra, A., Ravet, J., Di Girolamo, V. (2023), ‘The road to success: how regional innovation ecosystems can improve participation’, in The European Framework Programme for Rand I, Publications Office of the European Union, Luxembourg
Perianez–Forte, I., Guzzo, F., Hegyi, F. B., Gianelle, C. (eds.) (2021), Assessing Smart Specialisation: Case studies on Smart Specialisation, Publications Office of the European Union, Luxembourg, 2021.
Porter, M. E. (1998), ‘Clusters and the New Economics of Competition’, Harvard Business Review, 76(6), pp. 77–90.
Protogerou A., Caloghirou Y., Siokas E. (2010), ‘Policy–driven collaborative research networks in Europe’,
Economics of Innovation and New Technology, 19(4), pp. 349–372.
Rodríguez–Pose, A. and Crescenzi, R. (2008), ‘Research and Development, Spillovers, Innovation Systems and the Genesis of Regional Growth in Europe’, Regional Studies, 42(1), pp. 51–67.
Rosenthal, S. S. and Strange, W. C. (2003), ‘Geography, Industrial Organization and Agglomeration’,
The Review of Economics and Statistics, 85(2), pp. 377–393.
Rossato, C., Castellani, P. (2020), The contribution of digitalisation to business longevity from
a competitiveness perspective, TQM Journal, 32(4), pp. 617–645.
Sulis, P., Perpina, C. (2022), ‘Digital connectivity and the urban–rural divide’. In: Proietti, P., Sulis, P., Perpiña Castillo, C. and Lavalle, C. (eds.), New perspectives on territorial disparities, Publications Office of the European Union, Luxembourg.
Taglioni, D. and Winkler, D. (2014), ‘Making Global Value Chains Work for Development, Building Global Value Chains 2014’. In Annual Meetings side event, The World Bank Group, pp. 1–127.
Lembcke, A. and Wildnerova, L. (2020), ‘Does FDI benefit incumbent SMEs? FDI spillovers and competition effects at the local level’, OECD Regional Development Working Papers, 2020/02.
World Economic Forum (2019), The Global Competitiveness Report 2019.
EUROPEAN COMMISSION
Brussels, 27.3.2024
SWD(2024) 79 final
COMMISSION STAFF WORKING DOCUMENT
[…]
Accompanying the document
Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions
on the 9th Cohesion Report
{COM(2024) 149 final}
Figure 6.3 Share of EU population in 2022 by direction and rate of population change by urban-rural typology during 2010–2021
80
60
40
20
0
‑20
‑40
‑60
‑80
Note: Rapid growth/decline is defined as an increase/decline of at least 7.5 per 1 000 a year. Share of population relates to the share
on 1 January 2022.
Source: Eurostat [demo_r_pjangrp3] and DG REGIO calculations.
a year) is also more likely to have been experi‑ enced in rural regions than in others over the pe‑ riod. In remote intermediate regions, the reduction was as much as 37 % over the 12 years.
The relatively large share of rapidly shrinking re‑ gions that are rural and remote is in line with the reduction in population that occurred on average in these regions. Nevertheless, there are also regions with rapid population growth in all the groups, es‑ pecially the two French outermost regions of Guy‑ ane and Mayotte, where the population is project‑ ed to double by 2100.
Eurostat population projections for 20406 indicate an increase in the share of people living in shrink‑ ing regions in all groups by around 18 pp, as com‑ pared to 2020.
1.2The share of the population aged 0–29 relative to 30–59 varies markedly across the EU
In 2022, the EU population aged 0–29 was 139 million, and that aged 30–59 was 183 million. The difference of 44 million people constitutes a
generation gap that is the equivalent of 10 % of the EU’s total population. Inward migration is likely to reduce the difference in the future by adding to those aged 0–29, but is unlikely to eliminate it completely. In light of continued ageing and pro‑ jected levels of fertility, this means that the total population is projected to decline in the coming years and decades, based on the latest Eurostat baseline projections.
The age structure of the population also affects the birth rate7. As the younger age group gets old‑ er over time, the number of women of child‑bear‑ ing age will decline, leading to fewer births even if fertility rates remain unchanged.
The difference between the two age groups ex‑ ists in virtually all EU regions (Map 6.3), though the extent differs. For instance, in many regions in north‑western Spain and eastern Germany as well as in a few regions in Italy and Bulgaria, the population aged 0–29 is 40 % or more smaller than that aged 30–59, implying an increasingly negative natural change in population and a rapid growth in the share of population aged 65 or over compared with other regions.
1Eurostat [proj_19rp3].
2Birth rate refers to the total number of births in a year per 1 000 individuals in a population. The fertility rate refers to the number of live births in a year per 1 000 women of reproductive age in a population.
Canarias
Guadeloupe Martinique
Guyane
Mayotte Réunion
REGIOgis
Map 6.3 Population aged 0-29 relative to population aged 30–59 by NUTS 3, 2022
Percentage
<= 60
60 – 70
70 – 80
80 – 90
90 – 100
> 100
EU-27 = 76.0
Source: Eurostat (demo_r_pjangrp3).
Source: Eurostat [proj_23n].
By contrast, a few regions in France (including some of the outermost ones), Ireland, Sweden, the Netherlands, Finland and Denmark have more people aged 0–29 than aged 30–59, meaning they are likely to experience a slower natural decline in the population or even an increase.
Despite regional variations, there are clear national patterns, with most north‑western Member States, apart from Germany and Austria, having a relatively large share of the population aged 0–29 and south‑ ern Member States a relatively small share. Apart from higher outmigration of young workers, as con‑ cerns young women, in particular, the gap could be linked to lower birth rates because of differences in family policies, which are well developed in France and the northern Member States, and in the availa‑ bility and affordability of early childhood education and care services. Difficult labour market conditions for young people seeking stable employment, as well as difficult economic conditions in general, might also play a role, resulting, for example, in women in Spain and Italy having their first child relatively late in life (see also Chapters 1 and 2).
1.3The older population is growing while other age groups are shrinking
The gradual slowdown in population growth in the EU masks significant differences in the trends for different age groups. Some age groups have start‑ ed shrinking while others have continued to grow (Map 6.4). In particular, the population of working age (those aged 20–64) declined by 2.5 % over the 2014–2021 period, though by more in eastern and southern Member States, with some regions expe‑ riencing reductions of over 10 %8. This decline is expected to continue. At EU level, the working‑age population is projected to fall by 6.5 % by 2040 (Figure 6.4). Some Member States are more affect‑ ed than others. In Latvia, Lithuania and Greece, a reduction of around 20 % is projected. Assuming that the activity rates of people in various education groups (primary, secondary and tertiary) within each population subgroup (young, prime‑age individuals, older people, female, male, mothers) remain con‑ stant, the number of active people is expected to follow a very similar pattern. After rising to a record 205 million in 2022, the number of active people is estimated to decline to 201 million in 2030, 192 million in 2040, and 184 million in 20509.
3For future implications for the size of the labour force in a number of Member States, see European Commission (2023b), Chapter 2.
4Source: DG EMPL calculations, based on Eurostat and Organisation for Economic Co‑operation and Development (OECD) data and EUROPOP2023 population statistics.
Chapter 6: The demographic transition
Map 6.4 Percentage change in population by age group by NUTS 3, 2014–2021
Ages 0–19
Ages 20–64
Ages 65 and over
The reduction in the working‑age population has a significant negative impact on the size of the EU’s labour force and poses a risk to economic growth and fiscal sustainability, especially given the pro‑ jected increase in the population aged 65 and over (see below). Labour market policies can mitigate this decline of Europe’s labour force. In a scenario where the activity of women in the EU converged to the target value in the three top‑performing Member States for this group, an additional 17.3 million women would enter the EU labour market. Under the same assumption for men, an additional
8.8 million men would join the EU workforce.
There was a slightly smaller decline over the 2014–2021 period in the 0–19 age group at EU level (of 1.2 %), though in many southern and eastern regions the reduction was over 10 %. By contrast, there was an increase in several re‑ gions in Sweden, Czechia and the eastern part of Germany, as well as in capital city regions in many other Member States. The projection is for the pop‑ ulation aged 0–19 to decline by over 9 % by 2040, though by more in some eastern Member States (Lithuania, Latvia, Poland, Romania, Croatia and Bulgaria) as well as in Italy and Spain. Large and persistent reductions in this age group tend to im‑ ply a reduction in the need for schools, which can lead children having to travel longer distances to the nearest one as schools are closed down – es‑ pecially in rural areas, where distances are already relatively long10 – posing significant challenges to ensuring fair access (see Section 2).
By contrast, the vast majority of regions in the EU experienced a substantial increase in the popula‑ tion aged 65 and over between 2014 and 2021. This was particularly so in Poland, Slovakia, Ire‑ land and Cyprus, where in most regions the in‑ crease was over 25 %. In Finland, the Netherlands, France, Romania and Portugal, there were also some regions with growth this high. On the other hand, in a number of regions in Bulgaria, Greece, Spain, Lithuania and Latvia, the population of 65 and over declined. The projection is for this age group to increase by 27 % across the EU by 2040, though in Luxembourg, Ireland and Spain by 50 %
or more. This can be expected to lead to increased demand for healthcare and long‑term care and a consequent need for an expansion in capacity and, accordingly, in expenditure. If the domestic working force is shrinking, there may be a need for migrant workers to fill staff shortage gaps in the care sector.
1.4In rural regions the share of older people is higher and the share of the working‑age population lower
While, in the short term, the age structure of the population in the EU as a whole can only be changed by migration from and to the rest of the world, in individual regions it is also affected by movements to and from other parts of the EU. The likelihood of such movements occurring, and their direction, can be expected to depend, among other factors, on people’s ages. Those aged 20–39 may be more likely to move from rural regions to urban ones, while among those aged 40–64 and 65 or over migration from urban regions to rural or interme‑ diate ones may also be expected. These migration patterns would mitigate the ageing of the popula‑ tion in urban regions because of younger people moving in and (possibly) older people moving out; in rural regions they would exacerbate ageing as the reverse occurs.
In the EU as a whole, 21 % of the population was
aged 65 or over in 2022 (Figure 6.5a). This is
2.4 pp more than in 2014 and the projection is for it to continue to increase, reaching 27 % by 204011. This, coupled with a decline in the working‑age population, poses ageing‑related challenges, in‑ cluding increased healthcare and long‑term care needs and so increased pressure on public budgets, social (including inter‑generational) and territorial cohesion, investment, entrepreneurial activity and productivity. The extent of population decline and ageing, and the associated challenges, are likely to vary significantly between urban and rural regions.
In rural regions, the share of the population aged 65 or over tends to be relatively large, especial‑ ly in remote regions, where it exceeded the EU
5OECD (2021).
6Eurostat[proj_23n].
Figure 6.5 Share of different age groups in the total population by urban‑rural typology,
2014 and 2022
a)65 and over
b) 20–64
2022
2014
EU average, 2022
25
Share in total population, %
23
2022
2014
EU average, 2022
Share in total population, %
62
60
21
58
19
56
17
54
15
52
Source: Eurostat [demo_r_pjangrp3].
average share by 3.3 pp in 2022. The share grew more quickly than in other regions over the 2014– 2021 period, and it is expected to continue to do so in the future12. The share of the population of working age13, conversely, is smaller than aver‑ age in rural regions, again especially in remote ones (Figure 6.5b), and declined by more over the 2014–2021 period. Accordingly, rural regions can be expected to face more serious ageing‑related challenges from a shrinking potential workforce and more people aged 65 or over.
Conversely, in urban regions, the share of people of working age tends to be larger than the EU aver‑ age and the share of those aged 65 or over small‑ er (by 1.2 pp). The changes in both also tend to be smaller than in rural and intermediate regions, so that urban regions can be expected to be able to cope better with, or possibly avoid altogether, the challenges indicated above.
It is important to note that the extent of these challenges depends on the proportion of the work‑ ing‑age population that is employed, which in 2022 varied from 83 % in the Netherlands to 65 % in Ita‑ ly. In addition, there is a strong tendency across the EU for employment rates among older age groups to increase14. This is partly driven by increases in the age of retirement, but also by more older people choosing to work because of better health, higher education levels, better working conditions, and less arduous jobs than in the past (see also Section 2).
The employment rate in the EU for those aged 60–64 increased from 35 % to 49 % in the eight years 2014 to 2021, while the rate for those aged 65–74 increased from 8 % to 11 %. These rates vary considerably across the EU, the latter from 28 % in Estonia in 2022, and 19 % in Sweden, to 3 % in Romania, implying there is significant scope for more of those aged 65 or over to be employed in the future.
7Eurostat[proj_19r]. See also the 2024 Ageing Report (European Commission and European Policy Committee, forthcoming).
8Although the age group 20‑64 is referred to here as the population of working age, it should be noted out that the actual age of people in work varies widely across regions. Employment rates differ widely across regions, as do legal retirement age limits, which in some Member States are below age 65. The age of retirement is increasing across the EU, so that a growing proportion of people aged over 64 are in employment. In addition, some of those younger than 20 are also in work, though the proportion is tending to decline.
9See European Commission (2023a), Chapter 2.
Figure 6.6 Population growth in EU settlements, by settlement type and travel time to cities (annual average growth rates), 2011–2021
Annual population growth, %
0.8
0.6
0.4
0.2
0.0
‑0.2
‑0.4
Note: Annual growth rates are computed as compound annual growth rates for the period 2011–2021. Values exclude settlements that did not exist in 2011. First‑rank cities are the largest city in each country. Towns or villages are ‘close to a city’ if they are within a 30‑minute drive (or less) from a city’s boundary, and far from a city otherwise. Towns or villages are close to a large city even if they are also close to a small city.
Source: OECD calculations based on EU GEOSTAT data.
1.Access to high‑quality
services in the face of a shrinking
population and the costs involved
Given the demographic trends noted above, many settlements and regions will experience population decline over the next decade. Already half of the villages and over 40 % of towns in the EU lost pop‑ ulation over the 2011–2021 period. These were mainly places more than 30 minutes travel from cities, whereas towns and villages close to cities experienced on average an increase (Figure 6.6)15.
Places losing population face difficult choices about how to adapt public services to fit their smaller populations and budgets16. While policies need to ensure all citizens have access to essen‑ tial services, outside cities they are required to bal‑ ance accessibility – in terms of availability and the
ease with which services can be reached – against the cost of provision17.
Recent country case studies on population shrink‑ age in Estonia and Latvia show that shrinking places might also need to strategically consider ‘rightsizing’ their built environments to reduce the oversupply and decay of existing housing and oth‑ er infrastructure18 as well as to contain the cost of maintenance of older buildings.
1.1How will demographic change affect school operations and accessibility?
Estimates from a cross‑country study19 show that schools in sparsely populated rural areas tend to be smaller than those in cities and that they already have higher average costs per child20 – around 20 % higher in sparsely populated rural areas and 10 % higher in villages (Figure 6.7a).
10The definition of ‘close to a city’, as applied here to settlements, differs from the one used above in the urban‑rural typology, where it refers to the share of the population in a NUTS 3 region living in proximity to a city.
11Shrinking places may need to find creative solutions for services, involving either providing them virtually or co‑operating with nearby towns
or cities to provide them.
12The European Commission measures access to services and amenities by certain travel modes within fixed travel time intervals: see European Commission (2021), Box 4.2.
13OECD (2022).
14OECD/EC‑JRC (2021).
15The costs per child of small schools are generally higher than for large schools because fixed costs (e.g. for administrative staff and main‑
tenance) are spread across fewer students.
Figure 6.7 Access and cost estimates for specific services by degree of urbanisation, 2021
a)Primary schools
b) Cardiology services
4 500
115
4 300
4 100
3 900
3 700
110
Annual cost per cardiology patient (indexed)
105
3 500
1
2
3
4
Road distance to primary schools (km)
100
0
10
20
30
Distance to cardiology service locations relative
to cities (km)
Source: OECD/EC‑JRC (2021).
As population declines and ageing and other de‑ mographic trends such as urbanisation take hold, the OECD estimates that keeping primary school networks unchanged over the next decade will in‑ crease costs per child by 60 % in villages across the EU by 2035, and double this in sparsely popu‑ lated rural areas. These costs will be even higher in countries where non‑metropolitan areas are los‑ ing population more quickly21. Moreover, children in sparsely populated rural areas already travel much longer distances to school than those in cities.
The geographical accessibility of primary schools and early childhood education and care facilities also has an impact on labour markets, as it influ‑ ences parents’ decisions to work. For parents of young children, and for single parents in particular, the ease and flexibility of access to childcare de‑ termines decisions on taking up employment, as well as the number of hours worked. Analysis of several Member States shows that childcare pro‑ viders are frequently inaccessible by a short walk, but can usually be reached with a short drive. The geographic accessibility of childcare facilities tends to be much higher in urban settings, probably re‑ flecting higher demand and/or population density.
1.2How will demographic change affect healthcare and long-term care services costs and accessibility?
Staff shortages are likely to deepen in long‑term care, which is labour‑intensive but already at a disadvantage in competing for staff with more attractive sectors. The challenge will be particu‑ larly acute in rural areas, characterised by an age‑ ing‑related increase in long‑term care needs and shrinking human resources. Regarding healthcare, work in progress at the OECD has estimated the accessibility of some specialist medical treatment. For cardiology services, a 1 % reduction in the pop‑ ulation served by the average centre is estimated to be associated with over 0.5 % higher costs per patient22. People in sparsely populated rural are‑ as and villages typically travel over 20 km more to access these services than those in cities (Fig‑ ure 6.7b). People in towns also travel an average of 10 km more than those in cities to access them. To address the health needs of ageing populations, the OECD recommends23 that rural and remote places bolster their primary and integrative care systems, which are usually more accessible than specialist centres.
16European Commission (2021), Box 6.1.
17OECD/EC‑JRC (2021).
18OECD (2021).
Accessibility is an important consideration in how public services are distributed and their role in territorial cohesion. Inward migration and internal movements within the EU cannot ensure popula‑ tion growth in all places. Population loss is a demo‑ graphic reality for which many EU regions need to prepare, especially by planning the adaptation of essential service provision to population change24. At the same time, a loss of services can acceler‑ ate depopulation and foster discontent. National and regional governments should, therefore, help to co‑ordinate and fund efforts to limit territorial inequalities in access to services. Shared mobility solutions for rural areas, such as those supported by the Smarta‑NET project25 managed under DG MOVE of the European Commission, can play a role in this.
2.Harnessing talent to address demographic change
The previous section showed that the decline of the working‑age population is widespread, with more than half of people in the EU living in regions where it is occurring. In some regions, it is com‑ bined with additional structural challenges.
Some regions are faced with the combined chal‑ lenges of population ageing, a small and stagnant share of people with tertiary education, and out‑ ward migration of the young and well educated. This puts them at risk of falling into a talent de‑ velopment trap, which interferes with their capac‑ ity to build sustainable, competitive and knowl‑ edge‑based economies.
2.1Many regions in the EU are in a talent development trap26 or at risk of falling into one
Compared with the EU average, some regions have a significantly smaller share of tertiary‑level ed‑ ucated people, with young people (aged 20–24) less likely to be enrolled in tertiary education and more likely to move away to enrol somewhere else. Moreover, while the proportion of people aged 25–64 with tertiary education is growing in the EU at large – because more of those in younger age cohorts have this level of education than in older ones – in these regions it is growing more slow‑ ly than in others.27 The regions, therefore, will be less able to compensate for a declining population of working age by having a better qualified labour force capable of raising labour productivity. If the issue is left unaddressed, it is likely to reduce the regions’ competitiveness and widen the talent gap with other regions28.
Tertiary education can make a significant contri‑ bution to regional dynamism and attractiveness. However, a lack of career prospects, possibly linked to the lack of demand for qualified workers from companies and institutions in those regions, may discourage young people from investing in educa‑ tion and training or lead them to seek opportuni‑ ties elsewhere. Accordingly, it is equally important to create economic opportunities, capitalising on a region’s strengths, to retain and attract talent and to match available skills to current and prospective market needs.
The European Commission29 has formulated a method of identifying regions that are in a talent development trap30 or at risk of falling into one (see Box 6.2). Some 46 regions are identified according to this method as being in a talent development
19In addition to public services such as education, training and hospitals, places with a declining population face challenges in maintaining existing infrastructure that is too big (and too expensive) for the population that remains.
20
https://www.smarta‑net.eu/.
21See Box 6.2 for an explanation of the talent development trap.
22Eurostat [proj_19r].
23Note that, in addition to tertiary education, vocational education and training are also important for a labour force with sufficient relevant skills (see also Chapter 2).
24European Commission (2023a).
25This concept is distinct from that of the development trap discussed in Chapter 1.
Box 6.2 Identifying regions in a talent development trap or at risk of falling into one
The method used to identify regions that are in a talent development trap or at risk of falling into one is applied at the NUTS 2 level.
A region is considered to be in a talent develop- ment trap if:
4.3 pp.
A region is considered to be at risk of falling into a talent development trap if it is not in a talent development trap but:
•the annual average reduction in the popula‑ tion aged 25–64 was greater than 7.5 per 1 000 between 2015 and 2020;
•the share of the population aged 25–64 with tertiary education was below the EU average in 2020; and
•the share of the population aged 25–64 with tertiary education increased by less than the EU average between 2015 and 2020, i.e.
•the annual average net outward migration rate of those aged 15–39 was greater than 2 per 1 000 between 2015 and 2020.
trap (Map 6.5, in red). These regions, which are mostly in Bulgaria, Romania, Hungary, Croatia, the south of Italy, Portugal, eastern Germany and the north‑east and outermost regions of France, have a working‑age population that is increasingly de‑ clining and a small and stagnant number of people with tertiary education. Together, they account for 16 % of the EU population.
A second group of 36 regions is identified as being at risk of falling into a talent development trap because of the significant exodus of people aged 15–39 (Map 6.5, in orange). These are mainly in Latvia, Lithuania, eastern Poland, Slovakia, Greece, inland Spain, the north of Portugal, the northern half of France and Finland and account for 13 % of EU population. Together, around 30 % of people in the EU live in the two groups of regions31.
2.2Which types of regions are in a talent development trap?
Regions in a talent development trap have lower GDP per head than others (Figure 6.8). This might reflect the small share of tertiary‑educated people, which, with the relatively large share of agriculture in GDP, results in lower GDP per person employed and which, in turn, is reflected in lower wages and lower disposable income per head.
Regions at risk of falling into a talent develop‑ ment trap have similarly low levels of GDP per head, wages and disposable household income. In combination with lower employment rates, the low wages and low income relative to other regions are an important driver of outward migration of the population aged 15–39.
The employment rate of the working‑age pop‑ ulation was 7 pp lower in 2020 in regions in, or at risk of falling into, a talent development trap than in other regions. (This is a substantial dif‑ ference, which should be seen in the context of a smaller and declining working‑age population.) The employment rates of the population aged 25–64 with tertiary education were also lower but the difference from other regions was smaller at only 2 pp. The difference in employment rates, therefore, mainly affects people with only basic or secondary education. The unemployment rates for those aged 15–34 were correspondingly higher in trapped regions, and even higher in the regions at risk of falling into a trap. The share of jobs that are skilled was also smaller in both groups than in other regions, adding to the motivation of young people, who tend to be more highly educated than the older generation, to move away.
Over 80 % of the population in regions that are in a talent development trap or at risk of falling into one are living in a predominantly rural or in‑ termediate region as against 50 % of people in other regions (Figure 6.9). Accordingly, people in such regions have a higher probability of being in a trapped or at‑risk region. People in regions at risk are more often in a rural region than those in
26Note that there is considerable overlap in practice between the two categories. Many regions with a shrinking working‑age population and a small proportion of tertiary‑educated people also experience net departure of people aged 15–39. These are classified here as being in the first group, i.e. in a talent development trap.
Canarias
Guadeloupe Martinique
REGIOgis
Map 6.5 Regions in a talent development trap and regions at risk of falling in a talent development trap
Category
Shrinking working-age population and lagging level of tertiary education Net out-migration of people aged 15–39
Other regions
A region is in a talent development trap if it has
(a)a shrinking working-age population,
(b)a below-average and stagnant level of tertiary education and/or
(c)net out-migration of people aged 15–39.
Source: DG REGIO based on Eurostat data (demo_r_d2jan, demo_r_magec, lfst_r_lfsd2pop).
0
500 km
© EuroGeographics Association for the administrative boundaries
Figure 6.8 Productivity and employment indicators in regions in a talent development trap, regions at risk of falling into a talent development trap and other regions, 2020
% 120
In talent development trap
At risk of falling into talent development trap
Other regions
100
80
60
40
20
0
GDP per head (PPS) Compensation per
employee (PPS)
Employment rate
(20‑64)
Employment rate (25‑64 with tert.educ.)
Unemployment rate
(15‑34)
Share of high skilled jobs (%)
Note: Compensation per employee relates to 2019. GDP per head and compensation per employee are expressed in PPS with EU
average=100.
Source: Eurostat [nama_10r_2gdp, nam_10r_2hhinc, nama_10r_2coe, lfst_r_lfsd2pwc, lfst_r_lfe2eedu, lfst_r_lfp2act] and DG REGIO.
a region already in a trap, reflecting the relatively high net outward migration of people aged 15–39.
Regions in a talent development trap or at risk of being so also have a comparatively large share of people working in agriculture – 3–4 times more than in other regions in 2020 – where productivity and growth potential tend to be lower (Figure 6.10).
Over the 2015–2020 period, all regions experi‑ enced a reduction in the share of agriculture, but this was much larger in those in a talent trap or at risk of falling into one (2.5–3 pp) than in oth‑ ers (0.5 pp). The small proportion of people with tertiary education tends to diminish employment prospects further in trapped regions, leading to more outward migration and a consequent further decline in the working‑age population.
Figure 6.9 Urban-rural composition of regions in a talent development trap, regions at risk of falling into a talent development trap,
and other regions, 2020
Figure 6.10 Share of employment in agriculture in regions in a talent development trap, regions at risk of falling into a talent development trap, and other regions,
2010 and 2018
100
Urban
Intermediate
Rural
16
2010
2018
80
12
60
8
40
20
4
0
In talent development trap
At risk of falling into talent development trap
Other regions
0
In talent development trap
At risk of falling into talent development trap
Other regions
Source: Eurostat [demo_r_pjanaggr3] and DG REGIO.
Source: DG REGIO, JRC and Ardeco.
Table 6.4 Quality of government and innovation capacity in talent development‑trapped, at risk of being talent development‑trapped and other regions, 2020 and 2021
|
European Quality of Government Index
|
Regional Innovation Scoreboard
|
Population
with broadband access, %
|
Population
with bb speed
> 100 Mbps. %
|
In talent development trap
|
65
|
60
|
82
|
26
|
At risk of falling into talent development trap
|
85
|
71
|
86
|
40
|
Other regions
|
107
|
115
|
92
|
48
|
Note: Data on broadband access are for 2021. Data on other indicators: 2020.
Source: Eurostat [isoc_r_brod_h], RIS 2021, Ookla for good (TM), European Quality of Governance Index, DG REGIO.
The quality of governance and innovation capaci‑ ty are important enabling conditions for sustained economic development. Less developed regions tend to show a relatively poor performance in these areas (Table 6.4). This also holds for regions that are in a talent development trap and, to a lesser extent, those at risk of falling into one. The Euro‑ pean Quality of Government Index score and the Regional Innovation Scoreboard are both substan‑ tially lower for these regions than for others. More‑ over, the population with access to broadband is also smaller and the speed of internet connections slower.
Transport connections also tend to be poorer in regions that are in a talent development trap, or at risk of being so, than in others. Transport per‑ formance32 by car in trapped regions was 62 % in
2021, meaning that 62 % of the population living within 120 km can be reached within 90 minutes (Table 6.5), and in regions at risk of being trapped, 68 %, both well below the average for other re‑ gions (82 %). For rail connections, the differences are equally large. In trapped and at‑risk regions, only 8 % of the population within 120 km could be reached within 90 minutes by train in 2019, as against 19 % in others.
Poorer transport connections also affect a region’s access to services such as education and health‑ care facilities. Under 80 % of people lived with‑ in a 45‑minute drive of a university in regions in a talent development trap or at risk of falling into one, compared with 93 % in others. A sim‑ ilar difference holds for the share of people liv‑ ing within a 15‑minute walk of a primary school,
Table 6.5 Transport performance and access to services in regions in a talent development trap, regions at risk of falling into a talent development trap, and other regions, 2019 and 2021
|
Road
performance
|
Rail
performance
|
University
< 45 min.
driving, %
|
Primary school
> 15 min.
walking, %
|
Distance to nearest hospital, km
|
In talent development trap
|
62.4
|
7.9
|
78.4
|
56.0
|
11.7
|
At risk of falling into talent development trap
|
67.5
|
8.1
|
79.5
|
58.3
|
10.7
|
Other regions
|
82.2
|
19.1
|
95.9
|
65.7
|
8.6
|
Note: Road performance is for 2021, rail performance for 2019. Source: DG REGIO, based on Eurostat and TomTom data.
27See Box 3.3 for a more detailed description of the transport performance indicator.
Box 6.3 The Talent Booster Mechanism
Helping regions in a talent development trap, or at risk of falling into one, to become more resilient and attractive is a crucial part of the EU’s commitment to leaving nobody and no place behind as develop‑ ment takes place. If traps are left unaddressed, this will widen territorial disparities in the working‑age population and skills as times goes on, so hamper‑ ing the resilience and competitiveness of the EU as a whole.
This is why the Commission has launched the Talent Booster Mechanism to provide support to regions affected by a declining working‑age population to train, retain and attract people with the skills and competences needed to address the impact of the demographic transition. The mechanism consists of eight pillars, as follows.
•A pilot project launched in 2023 to help regions in a talent development trap, selected on the ba‑ sis of an open call, to formulate, consolidate, de‑ velop and implement tailored and comprehen‑ sive strategies, and to identify relevant projects to train, attract and retain skilled workers.
•A new initiative on ‘smart adaptation of re- gions to demographic transition’ was imple‑ mented in 2023 to help regions with high rates of exodus of young people to adapt to the de‑ mographic transition and invest in talent devel‑ opment through tailored place‑based policies. Regions were again selected on the basis of an open call.
•The Technical Support Instrument provides support to Member States to implement reforms at national and regional level to address the de‑ cline in the working‑age population and lack of skills and to respond to local market needs.
•Cohesion Policy programmes and Interregional Innovation Investments are intended to stimu‑ late innovation and high‑skill job opportunities and so help to improve the possibility of retaining and attracting talent in the regions concerned.
•A new call for innovative action is to be launched under the European Urban Initia- tive to test place‑based policy measures, led by shrinking cities, to address the challenge of de‑ veloping, retaining and attracting skilled workers.
•EU initiatives that support the development of talent are to be signposted on a dedicated webpage to provide easier access to informa‑ tion for interested regions on EU policies in areas such as research and innovation, training, edu‑ cation and youth mobility.
•A means will be established for exchange of experiences and dissemination of good prac- tice, and regions will have the possibility of set‑ ting up thematic and regional working groups to address specific employment and territorial challenges.
•The analytical knowledge required to support and facilitate evidence‑based policies on region‑ al development and migration will be further developed.
which was on average around 57 % in trapped and at‑risk‑of‑being‑trapped regions, as against 66 % in others. Equally, the distance to the nearest hos‑ pital was almost 12 km for people living in trapped regions, whereas in others it was under 9 km.
Poor transport connections and access to servic‑ es may simply reflect the more rural and sparsely populated nature of regions in a talent develop‑ ment trap or at risk of falling into one (see also Figure 6.9). Poor connectivity and digital infra‑
structure may also contribute to a less favoura‑ ble socio‑economic environment that causes net outward migration of the young and prevents a region from attracting tertiary‑educated people from outside.
Ensuring that regions in a talent development trap become more resilient and attractive is central to the EU’s commitment to leave nobody and no place behind as development takes place (see Box 6.3). As highlighted in the Communication33, on
28European Commission (2023c).
a demography toolbox, a range of financing instruments are available at the EU level to sup‑ port Member States in managing demographic change. In the partnership agreements 2021– 2027, 26 Member States have identified demog‑ raphy as a major challenge for their territories to be addressed with the support of Cohesion Poli‑ cy funds, such as the European Social Fund Plus. These measures complement other policy tools supporting Member States, including relevant reg‑ ulatory instruments and policy frameworks.
References
Batista e Silva, F., Dijkstra, L. (eds) (2024, forthcoming), Challenges and opportunities for territorial cohesion in Europe – contributions to the 9th Cohesion report, JRC Science for Policy report, Publications Office of the European Union, Luxembourg.
European Commission (2021), Cohesion in Europe towards 2050. Eighth report on economic, social and territorial cohesion, Publications Office of the European Union, Luxembourg, 2021.
European Commission (2023a), Harnessing talent in Europe’s regions, Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions, COM (2023) 32 final.
European Commission (2023b), Employment and Social Developments in Europe Addressing labour shortages and skills gaps in the EU, Publications Office of the European Union, Luxembourg.
European Commission (2023c), Demographic change in Europe: a toolbox for action, Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions, COM (2023) 577 final.
European Commission and Economic Policy Committee (2024, forthcoming), 2024 Ageing Report.
Publications Office of the European Union, Luxembourg.
OECD (2021), Delivering quality education and health care to all: Preparing Regions for Demographic Change, OECD Rural Studies, OECD Publishing, Paris.
OECD/EC–JRC (2021), Access and cost of education and health services: Preparing regions for demographic change, OECD Publishing, Paris.
OECD (2022), Shrinking Smartly in Estonia: Preparing Regions for Demographic Change, OECD Rural Studies, OECD Publishing, Paris.
EUROPEAN COMMISSION
Brussels, 27.3.2024
SWD(2024) 79 final
COMMISSION STAFF WORKING DOCUMENT
Accompanying the document
Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions
on the 9th Cohesion Report
{COM(2024) 149 final}
A recent report from the European Court of Au- ditors concluded that the level of competition for public contracts to deliver works, goods and ser- vices had declined over the past 10 years in the EU Single Market and that the Commission and Member States have not made systematic use of data available to identify the root causes of this21. Insufficient administrative capacity may adversely affect the degree of competition in public procure- ment procedures. Over half of all respondents of a recent EU-wide survey conducted by the Court of Auditors indicated that this could be the case22.
The Single Market Scoreboard uses 12 indicators to monitor how Member States perform each year in this regard. The proportion of single-bidder contracts – those awarded on the basis of a sin- gle tenderer’s offer – is an important indicator of public procurement standards, since it implies an absence of competition in public purchasing. Over the 2011–2021 period, the proportion of public procurement procedures in the EU Single Market where a single bidder was awarded the contract increased significantly, from 23.5 % to 41.8 %. At the same time, the number of bidders per pro- cedure almost halved, from an average of 5.7 to
3.223. In 2021, however, the share of public pro- curement tenders with a single bidder declined slightly, breaking the continuous upward trend in preceding years24.
The proportion of contracts awarded directly with- out any call for tenders being published is also an indicator of public procurement standards and shows a similar tendency. Such a direct procedure means that a public authority does not publish a call for tenders but approaches one or more companies directly, asking them to submit an of- fer, so making the process non-transparent and
potentially reducing the chances of obtaining good value for money.
In 2021, direct procedures accounted for 15.8 % of all procurement procedures in the EU Single Market reported by Member States on the Tender Electric Daily (TED) system, varying from 3.1 % in Greece to 42.3 % in Cyprus.
Data on this are available at regional level and have been monitored by the European Commis- sion since 201725. The Government Transparency Institute database contains details of public ten- ders at regional level published in TED26,27. This section reviews the most recent figures on public procurement contracts awarded following a single offer and those awarded directly without any call for tenders. These are for the period 2021–2022, so they still reflect, to some degree, the effect of the COVID-19 emergency situation, and more re- cent data would be needed to assess the impact of the pandemic.
These data show that single-bidder contracts were most common in regions in the eastern EU, Italy and Spain (Map 7.5). The share was above 70 % in Åland in Finland, Peloponnisos, Dytiki Makedonia and Ionia Nisia in Greece, and Vzhodna Sloveni- ja in Slovenia. By contrast, it was below 10 % in Stockholm, Mellersta Norrland Småland medöar- na and Västsverige in Sweden, Madeira (Portugal), and Malta. On average, single-bidder contracts accounted for a larger proportion of procedures in less developed regions than in others in 2019– 2020 as well as in 2021–2022 (Figure 7.2).
The proportion of regional and local authority contracts awarded directly without a call for ten- ders does not appear to follow a clear geograph- ical pattern, varying from over 30 % in Picardie,
1European Union (2023).
2This number increased to 71 % in the case of respondents working in administrative positions. They highlighted general knowledge con-
straints and shortages of staff qualified to prepare and conduct procedures that would increase competition.
3Source: See footnote 22.
4European Commission (2023b), p. 43.
5Fazekas (2017).
6Fazekas and Czibik (2021).
7The trends at the regional level do not always match those observed by the EU Single Market Scoreboard, as the number of regional contracts as a share of the total (regional, national, and European) varies widely between Member States, the average over the period 2018–2020 ranging from 78 % in Sweden to 4 % in Malta.
Chapter 7: Better governance
Figure 7.2 Single–bidder contracts and contracts awarded without a call for tender, by Cohesion Policy group of regions, 2019–2020 and 2021–2022
%
Less developed
Transition
More developed
45
40
35
30
25
20
15
10
5
0
Single-bidder contracts 2019–2020
Single-bidder contracts 2021–2022
Public contracts with no call 2019–2020
Public contracts with no call 2021–2022
Source: DG REGIO calculations based on e-TED data.
Basse-Normandie and Střední Morava in Czechia to below 3 % in a great many regions, including all of those in Spain, Greece, Denmark and Slovakia as well as in Estonia and Lithuania (Map 7.6).
1.1e-Government as a means of increasing transparency and accountability
Public authorities can increase their efficiency and improve their relationship with the public through e-government – the use of technology to improve and facilitate government services – such as to request birth certificates or submit tax declara- tions. Wider and easier access to public services ultimately increases their transparency and ac- countability, while reducing red tape, fraud and corruption.
In 2021, building on its digital strategy unveiled in 202028, the Commission presented the EU Digital Compass, which set out a vision and set of targets
for 2030 to stimulate digitalisation in the EU29,30. One of the targets involves the digitalisation of public services, the ambition being that all the main public services should be available online by 2030. Digitalisation in public administration ena- bles the streamlined delivery of services to people. Online platforms and digital portals provide con- venient access to these, reducing bureaucratic red tape and long waiting times. In the current 2021– 2027 programming period, over EUR 40 billion of support financed under Cohesion Policy is due to be allocated to investment in digitalisation31.
In 2023, 54 % of EU internet users interacted with public authorities, though with considerable variation between countries. In Finland and Den- mark, the share of internet users having interacted with public authorities was the highest among the Member States, at 92 %. In the Netherlands, the share was 84 %. The lowest rate of internet us- ers having interacted with public authorities was in Romania, at 14 %32.
8European Commission (2020a).
9European Commission (2021a).
10In 2021, 54 % of EU citizens aged 16–74 had at least basic overall digital skills, 26 pp below the 2030 target set in the Digital Compass (Source: Eurostat [isoc_sk_dskl_i21]).
11Source: Cohesion Open Data Platform. See: ‘Cohesion Policy supporting the digital transition 2021–2027’
(https://cohesiondata.ec.europa.eu/stories/s/Cohesion-policy-supporting-the-digital-transition-/vaxt-7rsr).
12Source: Eurostat
(isoc_ciegi_ac)
and
Eurostat (2023)
https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Digital_economy_
and_society_statistics_-_households_and_individuals#Use_of_e-government.
Box 7.3 While the COVID-19 pandemic accelerated the digitalisation of many services, including e-government, the ease of access to them seems to have declined
The 2023 edition of the European Commission sur- vey on the quality of life in European cities asked residents whether the information and services pro- vided by their local public authorities could be eas- ily accessed online. Some 74 %, agreed, 2 pp lower than in 2019, with the figure varying from 86 % in Aalborg in Denmark to 50 % in Palermo in Italy (Figure 7.3).
The COVID-19 pandemic accelerated the pace of digital transformation in the EU. The containment measures put in place meant that people were forced to use the internet to an increasing extent,
boosting digitalisation in the public sector. As a re- sult, Eurostat data show that the proportion of people interacting online with public authorities has steadily increased since 2019, though exist- ing inequalities in digital skills have also widened. The results of the survey show a clear reduction in the proportion of respondents reporting that the in- formation and services provided by their local public administration were easily accessible online in 66 of the 73 cities for which a comparison could be made over the period. The reduction was largest in Zagreb in Croatia (-9 pp), Rostock in Germany (-7 pp) and Miskolc in Hungary (-7 pp).
Figure 7.3 City residents agreeing that information and services of their local public administration are easy to access online, 2019 and 2023
100
95
90
Capital cities Other cities
85
80
Ankara Vilnius
Valletta Nicosia
Budapest Prague
Warsaw
Copenhagen Tallinn
Luxembourg Vienna
Amsterdam
LondoDnublin
75
Bratislava
Helsinki Brussels
Ljubljana
Stockholm
70
Bucharest
65
Berlin
60
Rome
Skopje
55
Riga Lisbon
Sofia
Zagreb Belgrade
Reykjavik
Oslo Podgorica
50
50
60
70
80
90
100
% total agree, 2019
Note: Percentages are based on all respondents (excluding ‘don’t know’/not answered). The dashed line is a 45-degree line (no change between 2019 and 2023). The chart only includes cities for which a time comparison can be made between 2019 and 2023.
Source: European Commission (2023c).
The proportion was smallest in less developed re- gions, averaging 42 % in 202133 as against 69 % in more developed regions and 74 % in transition ones. The proportion was below 20 % in all re- gions in Romania – except for Bucaresti-Ilov, the capital city region – and in several regions in Bul- garia (Map 7.7). Over the period 2013–2021, the proportion increased considerably in eastern EU
regions (except for those in Bulgaria and Romania) and Spain (Map 7.8).
Low usage of e-government services may be linked to a lack of internet access, a lack of e-gov- ernment infrastructure, and/or low levels of digital skills, which is a feature of some regions in the EU. This digital gap particularly affects marginalised communities, such as Roma living in remote segregated settlements. In 2023, some 6 % of the population aged 16–74
13Latest figures available at the time of closing the report.
Canarias
Guadeloupe Martinique
Guyane
Mayotte Réunion
Map 7.7 People interacting with public authorities via the internet in the previous 12 months, 2021
% of people aged 16–74
<= 20
20 – 40
40 – 60
60 – 80
> 80
no data
EU-27 = 58.5
Source: DG REGIO based on Eurostat data (isoc_r_gov_i and isoc_ciegi_ac).
0
500 km
© EuroGeographics Association for the administrative boundaries
Canarias
Guadeloupe Martinique
Guyane
Mayotte Réunion
REGIOgis
Map 7.8 Change in the proportion of people interacting with public authorities via the internet, 2013–2021
Percentage point difference
<= 5
5 – 10
10 – 15
15 – 20
20 – 25
> 25
no data
EU-27 = 17.0
FR: 2014–2021; FR (RUP), SI: 2015–2021.
Source: DG REGIO based on Eurostat data (isoc_r_gov_i and isoc_ciegi_ac).
0
500 km
© EuroGeographics Association for the administrative boundaries
in the EU had never used the internet34, with the proportion of individuals not having used the internet exceeding 10 % in Croatia (14 %), Greece and Portugal (13 % in both), and Bulgaria (12 %). The long-term vision for rural areas’ flagship Ru- ral Digital Futures35 highlights the importance of improving digital connectivity for closing the gap between rural and urban areas and boosting com- petences to make sure everyone benefits from the digital transition.
1.2An efficient business environment is a key asset for regional competitiveness
One of the adverse effects of inefficient institu- tions is a poor regulatory environment that bur- dens firms and adversely affects entrepreneur- ship. Low-quality institutions hamper the creation of new businesses and may lead budding entre- preneurs to seek opportunities abroad or give up altogether.
Over recent years, policy reforms have made the EU more business-friendly36. The Commission, via its Technical Support Instrument, has provided support to Member States for building sustaina- ble and competitive economies, including through reforms to improve the business environment, and strengthening SMEs.
How firms perceive the business environment can be key to whether they grow or feel obstructed from doing so. The sub-national component of the World Bank’s Enterprise Survey37 is a useful means for understanding the business environment across EU regions. The surveys were conducted between 2018 and 2022, in the form of nearly 19 000 in- terviews with top managers and business owners in the private sector. Results are available for a mix of NUTS 1, NUTS 2, and a combination of NUTS 2 or NUTS 3 regions. This section covers three ma- jor aspects of the business environment: access to finance, the extent of corruption, and the burden arising from the administration of tax.
Figure 7.4 Percentage of firms indicating access to finance as a major obstacle to their activity
versus Regional Competitiveness Index 2.0 by GDP per head
Access to finance as a major constraint for regions with GDP
per head >100
40
35
15
10
5
5
0
80
100
120
140
Regional Competitiveness Index, 2022 edition
0
80
90
100
110
120
Regional Competitiveness Index, 2022 edition
Note: GDP per head is the average in 2019–2021 with the EU average=100. Regions are a mix of NUTS 1, NUTS 2 and combined NUTS 2. Source: DG REGIO based on World Bank Business Enterprise Survey at the sub-national level and DG REGIO/JRC.
14In the three months prior to the survey. Source: Eurostat [isoc_r_iuse_i].
15
https://rural-vision.europa.eu/action-plan_en.
16European Commission (2021b).
17A project supported by the European Commission.
Access to external finance plays a critical role in ensuring regional competitiveness, particularly in less developed regions in the EU, since it is linked to business growth and survival (Figure 7.4)38. In 2023, among firms in the EU that judged bank loans to be a relevant source of funding, 7 % faced obstacles in obtaining a loan (5 % of large firms and 9 % of SMEs)39. Across the EU regions covered by the survey, 50 % of firms in Sud-Vest Oltenia in Romania identified access to finance as a major constraint40 on their current activity, 42 % in Attica and 41 % Kentriki Ellada (both in Greece), and 40 % in the Sud region of Italy (Map 7.9, left- hand side).
Corruption can worsen conditions for most busi- nesses, hampering overall regional competitive- ness, particularly in less developed regions. There is therefore a negative correlation between the
proportion of firms reporting corruption to be a major obstacle to their activity and regional com- petitiveness (Figure 7.5).
Corruption imposes a variety of costs on firms, in- cluding both the direct costs of paying bribes and the indirect costs of maintaining relationships with public officials and managing the uncertainty sur- rounding informal and often illegal arrangements, so damaging their incentive to develop and grow. Ultimately, corruption may lead to an inefficient al- location of resources41. Some 34 % of companies in the EU covered by a Eurobarometer survey in 2022 reported that corruption is a problem when doing business, with the largest proportions in Ro- mania (70 %), Greece (75 %) and Cyprus (78 %), and the lowest in Denmark (7 %), Ireland (8 %) and Estonia (9 %). In addition, 79 % agreed that close links between business and politics leads to
Figure 7.5 Percentage of firms indicating corruption as a major obstacle to their activity versus
Regional Competitiveness Index 2.0 by GDP per head
Corruption as a major constraint for regions with GDP
per head > 100
35
30
25
20
15
10
5
0
80
90
100
110
120
Regional Competitiveness Index, 2022 edition
Note: GDP per head is the average in 2019–2021 with the EU average=100. Regions are a mix of NUTS 1, NUTS 2 and combined NUTS 2. Source: DG REGIO based on World Bank Business Enterprise Survey at the sub-national level and DG REGIO/JRC.
18OECD (2024, forthcoming); Mach and Wolken (2012).
19European Central Bank (2023).
20A firm is considered to find an obstacle a major constraint if it responded ‘major obstacle’ or ‘very severe obstacle’ to the question ‘Is access to finance no obstacle, a minor obstacle, a moderate obstacle, a major obstacle, or a very severe obstacle to the current operations of this establishment?’
21Restuccia and Rogerson (2017).
Regional Competitiveness Index, 2022 edition
Note: Regions are a mix of NUTS 1, NUTS 2 and combined NUTS 2 and NUTS 3.
Source: DG REGIO based on World Bank Enterprise Survey at the sub-national level and DG REGIO/JRC.
corruption in their country and 70 % that favourit- ism and corruption hamper business competition42.
In the World Bank business enterprise survey, the largest proportion of firms identifying corruption as a major constraint on their current activity was in the region of Vest in Romania (74 %), followed by the Sud region in Italy (62 %), Centru and Bu- charesti-Ilfov in Romania, and Yugoiztochen in Bul- garia (all 55 %) (Map 7.9, centre).
The burdensome administration of taxes can ham- per regional competitiveness. Indeed, there is a clear tendency for the proportion of firms report- ing that tax administration is an obstacle to their activity to be larger in less competitive regions (Figure 7.6). Of course, this correlation does not imply that causation runs from the former to the latter, but it is consistent with it doing so.
The burden of tax administration includes all costs arising from the obligations that enterprises must fulfil, given the legislation in place. Studies have found that reducing the burden tends to en- courage entrepreneurship and firms to enter the market, irrespective of the corporate tax rate43.
Tax legislation is consequently a major concern of firms, and its simplification can improve the busi- ness environment, enhance competitiveness, and help to stimulate economic growth. In 2020, the European Commission adopted a Tax Action Plan, a set of 25 initiatives, with the aim of reducing the costs for businesses associated with tax collection and unnecessary administrative obligations in the Single Market44.
According to the World Bank Enterprise Survey, over 60 % of firms in Attica, Nisia Aigaiou and Kriti in Greece, Sud in Italy, and the Centro region in Portugal, identified tax administration as being a major concern for their current activity (Map 7.9, right-hand side).
22European Commission (2022), Flash Eurobarometer 507 on business attitudes towards corruption.
23Braunerhjelm and Eklund (2014); Branuerhjelm et al. (2021).
24European Commission (2020b).
Chapter 7: Better governance
Map 7.9 Major constraints identified by firms, 2018–2021
Access to finance
Corruption
Tax administration
Box 7.4 Corruption creates obstacles for nearly 1 in 5 smaller firms in less
developed regions
Corruption represents a greater barrier for smaller firms, especially those operating in less developed regions. Firms with fewer than 100 persons em- ployed are more likely to find corruption a severe obstacle than those with 100 or more, and the dif- ference is widest in the less developed EU regions (Figure 7.7). In these regions, almost 20 % of firms with fewer than 100 persons employed consider corruption to be a severe obstacle to their activity. For firms larger than this, the figure is 11 % in less developed regions (i.e. almost half) and only 5 % in more developed regions.
Part of the problem in regions with higher levels of corruption comes from greater ‘churn’, or the rate of business turnover, among local firms. Corruption in- creases uncertainty, which with the additional costs associated with corruption can increase the share of firms going out of business, leaving room for new entrants that in turn face the same issues. Churn is usually considered to be positive for economic de- velopment, underperforming firms closing and be- ing replaced by new more efficient ones. Corruption seems to distort business dynamics, creating churn without this necessarily leading to more competitive firms being in operation.
Figure 7.7 Percentage of firms in categories of regions that find corruption a severe obstacle to their operations by size class, 2018–2021
Firms with 5 to 19 workers
Firms with 20 to 99 workers
Firms with 100+ workers
30
% of firms that find corruption a severe obstacle
25
20
15
10
5
0
Less developed
Transition
More developed
Note:. Figures cover all EU Member States apart from CY, CZ and MT and refer to the period 2018–2021.
Source: OECD (2024, forthcoming) based on data drawn from the sub-national component of the World Bank Enterprise Survey.
Box 7.5 Small firms in less developed regions are most likely to find access to finance an obstacle
Limited access to finance creates obstacles for firms, particularly smaller ones in less developed regions. Around 9 % of firms with fewer than 20 persons employed in less developed regions report- ed to the Word Bank enterprise survey in 2023 that access to finance was a severe obstacle to their op- erations, more than double the figure in developed regions (4 %). The figure is lower for larger compa- nies in less developed regions (7 %) (Figure 7.8).
Smaller firms have more difficulties in accessing finance, for reasons that are more acute in less de- veloped regions. They usually have limited collateral to pledge against their loans, so banks often charge them higher rates than larger firms, which have more resources and are considered less risky. They
also tend to have less ability to collect information, so they are less aware of the financial products and government programmes that are available.
The difficulties tend to be more severe in less devel- oped regions, where there are fewer banks and so fewer local options for borrowing. Such regions have, on average, only 2 bank branches per 100 square kilometres as against 10 in more developed ones1. This limits choice and competition between banks, which can mean less favourable financing condi- tions for firms, particularly SMEs. The larger dis- tances between firms and banks in less developed regions can also hinder the exchange of information between them and make it harder to find out about suitable financial products.
1 Source: European Observation Network for Territorial Development and Cohesion, database 2021.
Figure 7.8 Percentage of firms in categories of regions that consider access to finance a severe obstacle to their operations by size class, 2018–2021
Firms with 5 to 19 workers
Firms with 20 to 99 workers
Firms with 100+ workers
20
% of firms that find access to finance a severe obstacle
16
12
8
4
0
Less developed
Transition
More developed
Note: Figures cover all EU Member States apart from CY, CZ and MT and refer to the period 2018–2021.
Source: OECD (2024, forthcoming) based on data drawn from the sub-national component of the World Bank Enterprise Survey.
2.The relevance of reforms and the European Semester
Chapters 1 and 2 describe the significant dispar- ities between regions that persist in the EU. In recent years, the European Semester cycle has highlighted disparities that affect economic devel- opment, such as access to education and essential public services, the extent of digitalisation, the lev- el of energy-efficiency, and the state of research and innovation. Disparities are further accentuated in rural areas, where access to basic services gen- erally remains a problem. These often translate into disparities in labour market outcomes (i.e. em- ployment and unemployment rates) and business competitiveness.
The European Semester country reports, in ad- dition to identifying country-wide economic and social issues faced by Member States, have high- lighted the relevance of the regional dimension of the EU’s growth and resilience agenda and the disparities across regions in respect of four dimen- sions of competitive sustainability: safeguarding the environment, productivity, fairness and macro- economic stability.
Tackling these disparities entails tackling the structural factors that cause them. This is relevant for both improving Cohesion Policy delivery and maximising its impact. The sub-national dimension is important for the effectiveness of national re- forms: on the one hand, regional-specific reforms may be required in certain cases, such as servic- es provided primarily at the sub-national level; on the other, the adoption of national reforms at the sub-national level may require specific measures to take proper account of regional features.
In the first place, several types of reforms can have a strong territorial dimension and require adapta- tion to the regional and local context. In the case of wide reforms intended to improve economic performance in a structural way, such as sectoral liberalisation or labour market reforms, these can have very diverse effects across regions, especial- ly on employment and wealth45. Adapting these reforms to the specific subnational contexts, in
particular in the most exposed areas, may require the definition of dedicated timelines and action plans for the implementation, possibly including ancillary measures at the subnational level.
Secondly, in areas where regional and local au- thorities are in the front line of providing services to businesses and citizens, national reforms can have differing effects depending on the local con- texts and the capabilities of local authorities. In these areas, ranging from education, healthcare, and social services to local transport, country-wide reforms that shift responsibility more to the local level need to take account of local differences in the demand for the services and in the capacity of the authorities concerned to deliver them.
Thirdly, sub-national authorities are in some in- stances best suited to addressing land use and territorial planning issues. As a place-based policy, the implementation and effectiveness of Cohe- sion Policy programmes are highly dependent on targeted territorial delivery. Reforms that help to better target Cohesion Policy funds would increase impact and mitigate adverse spill-over effects, or magnify beneficial ones, across regional borders.
As described in Section 2 above, effective and effi- cient public administration is an essential element in economic development, for both national and sub-national authorities. The administrative capac- ity to design regional development programmes, to allocate funding to projects in line with EU reg- ulations, and to account for the funding spent is a major determinant of effective policy delivery. The level of administrative capacity varies markedly across the EU, and many authorities, especially sub-national ones, are significantly limited in this respect (Box 7.6).
Public procurement procedures are a notable ex- ample. In a survey of municipalities conducted by the Organisation for Economic Co-operation and Development (OECD), smaller ones identified the simplification of such procedures as one of the main reforms needed to improve operational ca- pacity. Another OECD survey, this time with the Committee of the Regions, found that ‘lengthy
25See for instance: Kovak (2013).
Figure 7.9 Challenges in the strategic planning and implementation of infrastructure investment in municipalities in the EU
Major challenge
Somewhat of a challenge
Excessive administrative procedures and red tape
Lenghty procurement procedures
Local needs are different from those given priority at central level
Lack of long-term strategy at central level
Co-financing requirements for central government/EU are too high
Lack of coordination across sectors
Lack of political will to work across different levels of government
Lack of incentive to cooperate across jurisdictions Lack of joint investment strategy with neighbouring SNGs
Multiple contact points (absence of a one-stop shop)
Lack of (Ex post) impact evaluations Ex ante analyses not adequately take into account the full life-cycle
of an investment
Monitoring not used as a tool for planning and decision making
Insufficient involvement of civil society in the choice of projects
Ex ante analyses/appraisals not consistently used in decision making
Lack of long-term/strategic planning capacity
Lack of adequate own expertise to design projects
No relevant up-to-date data available at local level
Source: OECD-CoR survey [OECD-CoR (2016)]. Results of the survey on regional and local obstacles to investments.
procurement procedures’ were the second most frequently identified challenge, with over 50 % of respondents regarding them as a ‘major chal- lenge’ (see Figure 7.9). Reforms to strengthen sub-national capacity as regards public procure- ment could include a mixture of decentralisation measures, the mutualisation of procurement, and digitalisation (i.e. e-procurement46).
Access to finance is at the core of the capacity of sub-national authorities to deliver services and carry out investment. This, along with effective multilevel governance, is a key part of the re- forms. The importance of a sound fiscal frame- work for multilevel governance is recognised in the EU Directive on this47. As indicated in Chapter 8, sub-national authorities are responsible, on aver- age, for the execution of a third of total govern- ment expenditure (current plus capital) in the EU.
26Allain-Dupré et al. (2017).
27European Union (2011). The Directive envisages that ‘Member States shall establish appropriate mechanisms of coordination across sub- sectors of general government to provide for comprehensive and consistent coverage of all subsectors of general government in fiscal planning, country-specific numerical fiscal rules, and in the preparation of budgetary forecasts and setting-up of multiannual planning as laid down, in particular, in the multiannual budgetary framework’.
Box 7.6 The evolution of the organisational model of Managing Authorities between 2000 and 2020
The introduction of general provisions on the Struc- tural Funds for the 2000–2006 period marked a sig- nificant milestone by formally recognising the role of managing authorities (MAs) for the first time. The regulation mandated that MAs are accountable for the effective and accurate management and imple- mentation of funds. This shift positioned MAs at the forefront of the management of EU funds for Cohe- sion Policy.
An ongoing study1 covering the period from 2000 to 2020 investigates the significant transformations within MAs responsible for interventions financed by the European Regional Development Fund across Member States, excluding transnational coopera- tion. The study looks at aspects such as staff com- position, internal processes and organisation, lead- ership dynamics, and management of relations with partners. Furthermore, the study considers external factors that might affect the organisation of MAs, including EU regulations, national and institutional frameworks, and socio-economic factors, aiming to explain organisational changes and project the potential challenges for the implementation of pro- grammes in the 2021–2027 programming period and the preparation for future periods.
Preliminary findings reveal that the introduction of a unified EU-level regulatory framework and shared responsibilities led to a diverse range of organ- isational models among MAs in different Member States. Initially, the size of these authorities varied significantly, as did their internal organisational structures, which ranged from entities with bespoke
processes to those integrating or sharing processes with encompassing organisations or other authori- ties within their respective countries.
Over time, changes reflected the evolution of the EU regulatory framework from one programming period to another. For instance, shifts in policy ob- jectives and implementation tools (such as finan- cial instruments and integrated territorial delivery mechanisms) had some effect on the organisational structure, the number and specialisation of structur- al units and the delegation of tasks and processes. Other organisational changes followed new nation- al policies and legislation, including changes in the overall governance of regional and Cohesion Policy at national level. External audits also triggered or- ganisational changes within MAs, especially revi- sions of internal processes and procedures.
Increased programme budgets led to expanded au- thority sizes. Yet recruiting and retaining skilled staff, developing soft and managerial skills, and achieving gender balance remained challenging. The analysis revealed the importance of consistent leadership as a driver for change, though MA leaders primarily fo- cused on financial achievements and the effective functioning of management and control systems rather than on the achievement of policy objectives. Managing relations with stakeholders has seen little evolution and was mainly focused on running the activities of the monitoring committee, suggesting a lack of emphasis on broader trust-building and con- flict management initiatives.
1 PPMI Group and University of Strathclyde (2024, forthcoming).
There are considerable variations, however, be- tween Member States, reflecting differences in the institutional setting. Nevertheless, in all cases, even in the most decentralised countries, enhanc- ing inter-governmental co-operation and a sound fiscal framework is essential to avoid coordination failures, the emergence of ‘unfunded mandates’ and, ultimately, inadequate policy implementa- tion. Addressing the nexus between the different
institutional levels in the design and implementa- tion of reforms is a key aspect in the definition of an effective governance structure.
The multiannual programming of Cohesion Pol- icy has been a major driver for the integration of public investment in medium-term budgetary frameworks and public financial management structures. Integrated strategic planning and
methods of project appraisal and selection that guide budget allocation effectively and use asset registers as input are key to carrying out public investment efficiently. While wide-ranging reforms to systems for managing public investment have been implemented in several Member States, room for improvement is evident in many others. In this regard, the success of decentralisation de- pends to a large extent on effective vertical and horizontal co-ordination across layers of govern- ment to avoid duplication and to ensure policies are consistent. Among EU Member States, there is evidence that difficulty in absorbing funding for in- vestment can be a sign of poorly co-ordinated fis- cal policy as well as inadequate administrative ca- pacity at sub-national level48. Capacity constraints and co-ordination deficiencies also hinder the use of diverse methods of financing by sub-national governments.
To strengthen economic, social and territorial co- hesion in the European Union, the Commission provides to Member States and regions support through the Technical Support Instrument. Support measures cover several reform areas, including: improving the quality of governance and public services; strengthening productivity, innovation and the green transition; and harnessing talent and employment opportunities. The tailor-made support measures help regions define and imple- ment appropriate processes and methodologies to address the development challenges in an inte- grated manner, taking into account good practices and lessons from other regions. In addition, the TSI also aims to incentivise peer learning and promote intra Member State and cross-border regional co-operation, and complements existing Commis- sion initiatives – Harnessing Talent in Europe’s Re- gions, the New European Innovation Agenda, the Just Transition Platform, the Smart Specialisation Platform, and others.
Reflecting on the structural issues inhibiting con- vergence across regions identified in recent Euro- pean Semester country reports and annexes is a
precondition for tackling the underlying factors49. This includes pointing to the spatially targeted re- forms that could be instrumental in this respect, and providing, where relevant, guidance to Mem- ber States on where to focus investment for the effective use of funding. This is particularly rel- evant for the 2024 Semester, in which Country Specific Recommendations provide guidance to Member States on allocating the flexibility amount included in budgets for the 2021–2027 program- ming period50.
28OECD (2020).
29The 2019 Country Reports included in Annex D a set of regional factors, as well as investment guidance for the 2021–2027 programming period.
30Article 18.1.a of the Common Provision Regulation (Regulation (EU) 2021/1060 of the European Parliament and of the Council of 24 June 2021).
References
Acemoglu, D. and Robinson, J. (2010), ‘The Role of Institutions in Growth and Development’, Review of Economics and Institutions, 1(2).
Allain–Dupré, D., Hulbert, C. and Vincent, M. (2017), ‘Subnational Infrastructure Investment in OECD Countries: Trends and Key Governance Levers’, OECD Regional Development Working Papers, 2017(05), OECD Publishing, Paris.
Barbero, J., Christensen, M., Conte, A., Lecca, P., Rodríguez–Pose, A. and Salotti, S. (2023), ‘Improving Government Quality in the Regions of the EU and its System–Wide Benefits for Cohesion Policy’, Journal of Common Market Studies, 61, pp. 38–57.
Barbero, J., Mandras, G., Rodríguez–Crespo, E. and Rodríguez–Pose, A. (2021), Quality of government and regional trade in the EU, Territorial Development Insights Series, JRC124475, European Commission.
Botero, J.C. and Ponce, A. (2011), Measuring the Rule of Law Available at SSRN.
Braunerhjelm, P. and Eklund, J.E. (2014), Taxes, tax administrative burdens and new firm formation, Kyklos, 67, pp. 1–11.
Braunerhjelm, P., Eklund, J. and Thulin, P. (2021), ‘Taxes, the tax administrative burden and the entrepreneurial life cycle’, Small Business Economics, 56, pp. 681–694.
Brezzi, M., González, S., Nguyen, D. and Prats, M. (2021), ‘An updated OECD framework on drivers
of trust in public institutions to meet current and future challenges’, OECD Working Papers on Public Governance, 48, OECD Publishing, Paris.
Charron, N., Dijkstra, L. and Lapuente, V. (2014), ‘Regional governance matters: Quality of government within European Union Member States’, Regional Studies, 48(1), pp. 68–90.
European Central Bank (2023), Survey on the Access to Finance of Enterprises (SAFE).
European Commission (2020a), Shaping Europe’s digital future, Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions, COM(2020) 67.
European Commission (2020b), Action plan for fair and simple taxation supporting the recovery strategy, Communication from the Commission to the European Parliament and the Council, COM(2020) 312 final.
European Commission (2021a), 2030 Digital Compass: the European way for the Digital Decade, Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions, COM(2021) 118 final.
European Commission (2021b), Cohesion in Europe towards 2050 – Eighth report on economic, social and territorial cohesion, Publications Office of the European Union, Luxembourg.
European Commission (2023a), 2023 Rule of Law Report, Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions, COM (2023) 800 final.
European Commission (2023b), 2023 Annual Single Market Report: Single Market at 30’, SWD (2023)
26 final.
European Commission (2023c), Report on the quality of life in European cities, 2023, Publications Office
of the European Union, Luxembourg.
European Commission and High Representative of the Union for Foreign Affairs and Security Policy (2023), Joint Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions: the fight against corruption, JOIN(2023) 12 final.
European Union (2011), Council Directive 2011/85/EU of 8 November 2011 on requirements for budgetary frameworks of the Member States.
European Union (2023), Special report of the Court of Auditors on: Public procurement in the EU, Less competition for contracts awarded for works, goods and services in the 10 years up to 2021, Publication Office of the European Union, Luxembourg.
Fazekas, M. (2017), Assessing quality of government at the regional level using public procurement data, DG Regional and Urban Policy, Working Paper 12/2017.
Fazekas, M. and Czibik, A. (2021), ‘Measuring regional quality of government: the public spending quality index based on government contracting data’, Regional Studies, 55(8), pp. 1459–1472.
Gianelle, C., Guzzo, F., Barbero, J. and Salotti, S. (2023), ‘The governance of regional innovation policy and its economic implications’, Annals of Regional Science.
Gründler, K. and Potrafke, N. (2019), Corruption and economic growth: New empirical evidence, European Journal of Political Economy, 60.
Ketterer, T.D. and Rodríguez–Pose, A. (2018), ‘Institutions vs. ‘first–nature’ geography: What drives
economic growth in Europe’s regions?’, Papers in Regional Science, 97, S25–S62.
Kovacic, M. and Caperna, G. (2022), Joint Research Centre Statistical Analysis of the World Justice Project Rule of Law Index 2021, JRC 131884.
Kovak, B.K. (2013), ‘Regional Effects of Trade Reform: What Is the Correct Measure of Liberalization?’,
American Economic Review, 103(5), pp. 1960–1976.
Mach, T. and Wolken, J.D. (2012), Examining the Impact of Credit Access on Small Firm Survivability (24 February 2012), FEDS Working Paper 2012–10.
OECD (2020), Strengthening Governance of EU Funds under Cohesion Policy: Administrative Capacity Building Roadmaps, OECD Multi–level Governance Studies, OECD Publishing, Paris.
OECD (2022), Building Trust to Reinforce Democracy: Main Findings from the 2021 OECD Survey on Drivers of Trust in Public Institutions, Building trust in Public Institutions, OECDD Publishing, Paris
OECD (2024, forthcoming), Boosting Business in Regions: Easing Administrative Burdens and Financing Constraints, OECD Publishing, Paris.
PPMI Group and University of Strathclyde (2024, forthcoming), Organisational Models of Programme Authorities Managing EU Cohesion Policy 2000–2020, REGIO/2022/OP/0011.
Restuccia, D. and Rogerson, R. (2017), ‘The Causes and Costs of Misallocation’, Journal of Economic Perspectives, 31(3), pp. 151–174.
Rodríguez–Pose, A. and Garcilazo, E. (2015), Quality of Government and the Returns of Investment: Examining the Impact of Cohesion Expenditure in European Regions, Regional Studies, 49(8).
EUROPEAN COMMISSION
Brussels, 27.3.2024
SWD(2024) 79 final
COMMISSION STAFF WORKING DOCUMENT
Accompanying the document
Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions
on the 9th Cohesion Report
{COM(2024) 149 final}
PUBLIC FINANCES, NATIONAL POLICIES AND COHESION
•The degree of decentralisation of both national public expenditure and Cohesion Policy programmes is generally lower in less developed countries, where there is scope for greater involvement of sub-national governments.
•Preliminary evidence shows that nationally funded investment for territorial cohesion in less developed countries represents in most cases only a small fraction of the funding provided under Cohesion Policy. There is therefore ample scope for increasing the efforts of the Member States concerned to strengthen cohesion as well as for improving the co-ordination with Cohesion Policy.
•Sub-national governments are responsible for carrying out a large share of public expenditure, though with significant differences across the EU.
•Sub-national governments are responsible for the majority of public investment in the EU. This is less the case in less developed countries, but the difference with more developed countries diminished significantly between 2004 and 2022 as public investment became more decentralised in the former. Since all governments decentralise certain public services and investment, a sound fiscal framework, as well as intergovernmental fiscal cooperation, is essential to improve the delivery of public services.
•Cohesion Policy multiannual programming has been a key driver of public investment integration in medium-term budgetary frameworks and public financial management structures. If managed well, decentralised investment, can improve the efficiency and effectiveness of public services to citizens and firms. Effective multilevel governance, in turn, relies on vertical and horizontal co-ordination across government's layers.
•Preliminary evidence from the OECD for several Member States shows considerable heterogeneity in the mix of funding sources at the regional and local levels. Transfers from other levels of government are the most important source of revenue. Countries where there is heavy reliance on one or only a few revenue sources are less resilient to shocks.
Chapter 8
Public finances, national policies and cohesion
Figure 8.1 Share of Cohesion Policy support implemented through regional programmes and share of sub-national public expenditure, 2014–2020
Less developed countries
More developed countries
EU fuiniig aoriugo rcgiiinl prigrnmmce (%)
l20
1.Introduction
This chapter reviews national policies for territorial cohesion and sub-national public finances. It be- gins by examining preliminary evidence on the extent of nationally funded policies for territorial cohesion in a number of Member States using the data collected through ad hoc studies. It moves on to examine sub-national trends in public expendi- ture, revenue and investment over time and across Member States (Section 3). It then considers the
and resilient national economy. Improving the eco- nomic performance of all regions also increases the opportunities for co-operation and can create a dynamic environment in which innovation and knowledge are shared more widely, improving the competitiveness of the whole country.
These are compelling reasons why Member States should apply the ‘do no harm to cohesion’ princi- ple to their national policies in all areas, meaning that national, regional and local authorities should
l00 80
60
40
20
0
-20
Sub-national public expenditure (%)
composition of regional and municipal public ex- penditure and revenue in a number of EU Member States on the basis of data collected by the Or-
be aware of the asymmetric territorial impact that any policy measure might have and take account of this in the policy-making process (the Treaty
Source: DG REGIO calculations based on Eurostat gov_l0a_main and Cohesion Open Data.
Ganisation for Economic Co-operation and Devel- opment (OECD) with the support of the European Commission (Section 4).
In order to bring aut broad differences, the chap- ter divides the EU Member States into two groups according to their gross national income (GNI) per head, which is taken as a proxy for their level of development. The lƼ countries with GNI per head below 90 % of the EU average – the threshold for eligibility for the Cohesion Fund – are included in the less developed group (i.e. Bulgaria, Czechia, Estonia, Greece, Croatia, Cyprus, Latvia, Lithuania, Hungary, Malta, Poland, Portugal, Romania, Slove- nia and Slovakia), the remaining l2 in the more developed group.
2.National policies addressing territorial disparities
National policies to tackle regional disparities have a key role in strengthening territorial cohesion in the EU, especially contributing to reducing with- in-country disparities. Reducing internal territori- al disparities is essential for optimising economic efficiency and improving competitiveness, and it needs to be a priority in Member States. By secur- ing balanced development between regions, Mem- ber States can exploit the unique strengths and
on the Functioning of the EU, it should be noted, explicitly calls on Member States to contribute to strengthening the economic, social and territorial cohesion of the EU through their economic policies (Articles l74 and l7Ƽ)).
Where disparities exist within countries, these should be addressed in a complementary man- ner by national policies and EU funding. Where EU-funded interventions are planned and imple- mented, there may be a need for further support from national resources. This may be the case, for example, where the demand for a certain type of assistance exceeds the expectations of pro- grammes or where unforeseen circumstances arise that require an immediate response. In areas not covered by EU funding, national policies represent the only level of support for sub-national govern- ments to spend on policies aimed at strengthen- ing socio-economic performance, recovering from immediate crises, addressing long-term deficien- cies and building resilience to future shocks and a rapidly changing environment.
National policies and Cohesion Policy should be mutually reinforcing, leading to a more compre- hensive and effective approach to regional devel- opment. By actively tackling regional disparities, Member States align their national strategies with
Figure 8.l shows the share of EU Cohesion Policy support implemented through regional programmes in 20l4–2020 (y-axis) in relation to sub-national public expenditure as a share of total government spending in the same period (x-axis), the size of the bubbles representing the amount of EU Cohesion Policy funding. There is a positive relationship be- tween the two, implying that the degree of decen- tralisation of Cohesion Policy funding is positively correlated with that of national funding, or, in other words, that EU policy and national policy go broad- ly in the same direction. Figure 8.l also shows that larger Member States and federal countries tend to be more regionalised in general (upper right- hand corner of the graph), while smaller Member States tend to be less regionalised in terms of general government expenditure and be dominat- ed by national Cohesion Policy programmes. Re- markably, less developed countries are clustered in the lower left-hand corner of the graph; i.e. they are in general less regionalised, which gives ample scope for a greater involvement of sub-national governments in the design and implementation of both national public expenditure programmes and Cohesion Policy programmes (Box 8.l).
A more in-depth examination of the measures taken by countries to tackle territorial dispari- ties is limited by the fact that available evidence on national policies is scarce and unsystematic, and, where it exists, is mainly limited to specific, time-limited case studies. To fill this knowledge gap, the European Commission has promoted a series of studies starting in 20l9 to analyse poli- cies for tackling territorial disparities that are fully funded by national resources.
One such study defined national policies for cohe- sion to encompass all policy initiatives and meas- ures with the direct objective of reducing territorial disparities, together with those without such an objective but with a significant potential to achieve this. It covered ll Member Statesl. All of these have national policies for cohesion, as defined, in place, with a range of policy instruments targeting different aspects of development, the most com- mon being direct support for business develop- ment and innovation, transport infrastructure pro- jects, and tax incentive schemes to support trade and improve the business environment.
Assets of each, contributing to a more diversified
overarching EU objectives.
L European Commission (20l9). The study was based on a combined analysis of statistical data, case studies, and stakeholder interviews.
It covered ll Member States, namely Bulgaria, Croatia, Czechia, Hungary, Italy, Poland, Portugal, Romania, Slovakia, Slovenia and Spain.
growth objective in terms of sustainable growth. This is based on the recognition of the uneven ter- ritorial impact of climate change measures and the impact on already structurally weak regions.
The place-based approach to regional policy is now well established and widespread, often in the form of integrated development strategies tailored to the specific needs of places. It should be noted that the EU ‘smart specialisation’ approach has helped to disseminate and mainstream this approach among regional authorities in the EU and beyond. Closely linked to this are visible efforts to increase coher- ence between regional and sectoral policies, for ex- ample by giving a territorial dimension to sectoral policies. Again, smart specialisation is an early ex- ample of the regionalisation of an otherwise typical sectoral policy.
The study found an increasing focus on vulnera- ble or marginalised regions. In several cases, this reflects a renewed political concern with the eco- nomic and social difficulties faced by rural areas,
often in remote parts of countries, where there is a perception of neglect in favour of a policy focus on cities. This focus is also linked to the objective of improving regional resilience, as a consequence of the territorial vulnerabilities revealed by the impact of the COVID-l9 pandemic and the need for regions to be more resilient to shocks. This renewed focus is also part of a wider policy objective of using re- gional policy interventions to improve quality of life and access to public services where these are under pressure or linked to demographic decline.
Finally, governance and institutional reform and ca- pacity-building at regional and local level remain high on the regional policy agenda across Europe. In some cases, this involves the redefinition of exist- ing administrative boundaries or units, for example through mergers and rationalisation of municipali- ties or increased co-operation between regional and local authorities. Notably, and in line with the global trend observed in the OECD/UCLG report, the decen- tralisation process in some countries is asymmetric.
Reducing territorial disparities is often pursued as part of growth and industrial policy, especially in Member States where all or most of the regions are less developed according to the EU Cohesion Policy classification. In these cases, territorial cohesion is often an integral part of a country’s broader effort to reduce economic disparities with more developed parts of the EU.
Nationally funded policies complement EU Cohe- sion Policy in two main ways. Either they provide additional funding in national priority areas where Cohesion Policy funding is considered insufficient,
or they support activities that are not eligible for EU funding. In practice, in budgetary terms, na- tional policies for cohesion, as defined, appear to account for a very small fraction of EU Cohesion Policy funding. Of the Member States covered by the study, only Italy and Romania have a signifi- cant budget for territorial cohesion, of much the same size as Cohesion Policy in the case of Italy and just over a third of this in Romania. In the re- maining countries, national funding ranges from just under 3 % of Cohesion Policy funding, here including national co-financing, in Croatia, to just under 9 % in Spain.
As regards the regions targeted, there is evidence of different approaches and mixed experience. According to the findings of the study, some countries (e.g. Czechia and Croatia) active- ly support the more prosperous regions, includ- ing capital city regions, considering them to be the driving centres of economic growth that can help reduce the country’s development gap with the more advanced parts of the EU. Other Member States – Italy, Romania and Spain, especially, as indicated above – are more active in supporting less developed regions to reduce disparities. The first approach is more common in countries that devote very limited national resources to this type of policy, while the second approach, targeting less developed regions, is more common in countries that invest more.
The vast majority of national policy measures for cohesion in the countries covered are designed by central government (90 %), some are co-designed with the regions, while only 3 % of the initiatives examined are designed at regional level. Imple- mentation is the responsibility of central govern- ment in 70 % of cases and only l6 % of measures
2European Commission (forthcoming).
Are implemented by regional authorities, the rest being implemented by local authorities. Countries where sub-national authorities carry aut only a small share of public expenditure tend to have a more centralised governance of national policies for cohesion (as in Bulgaria, Croatia, Hungary, Por- tugal, Romania and Slovenia).
Further evidence is obtained by restricting the scope of the analysis to investment programmes or initiatives fully financed from national resources in the fields of economic development (including
e.g. investment in innovation, ICT, and SME com- petitiveness), transport (including all forms of mo- bility), energy, environment, health and education, thus excluding non-investment measures, and by focusing only on policies that either have a specific territorial/spatial focus or are explicitly aimed at reducing territorial disparities and strengthening territorial cohesion, thus excluding measures with- aut direct cohesion objectives2.
Preliminary results for seven Member States (Cro- atia, Czechia, Estonia, Lithuania, Poland, Romania
and Slovenia) show that, for the period 20lƼ– 202l, 36 investment initiatives were planned with a budget of EUR 7.9 billion. This represents only Ƽ.4 % of the combined European Regional Devel- opment Fund and Cohesion Fund allocations (in- cluding national co-financing), for these countries for the 20l4–2020 programming period. There are, however, big differences between the coun- tries, especially between Romania, where national investment for cohesion amounted to around 30 % of Cohesion Policy funding for investment, and the other six countries, where the figure ranged from
3.8 % in Slovenia and l.7 % in Czechia to only
0.7 % in Poland and under 0.Ƽ % in Croatia, Esto- nia and Lithuania.
The implemented budget of national investment policies for cohesion as of the end of 2023 is overall equal to 76 % of the planned budget for the seven countries surveyed, with a maximum
Evidence is available with a breakdown by cate- gories of beneficiary of national investment pol- icies for cohesion, where again a single measure may address more than one category of benefi- ciary. The policies identified cover a wide range of different beneficiaries. In particular, it can eve l- ed that the majority of measures (67 %) are tar- geted at municipalities, followed by SMEs (39 %), public organisations (2Ƽ %), non-profit organisa- tions (2Ƽ %), start-ups (22 %), scale-ups (ll %), large enterprises (l7 %), industrial parks and oth- er types of parks or innovation zones (ll %) and multinationals (8 %).
Some 86 % of the investment measures are de- signed by central government, ll % by region- al authorities and only 3 % by local authorities. The latter two, however, have more importance in the implementation of investment, being respon- sible for implementing l9 % and 2Ƽ % of meas-
in Czechia at l07 %, and a l00 % execution in Croatia, Estonia, and Lithuania, while Slovenia, Po- land and Romania implemented 87 %, 84 % and 73 % respectively. If we compare the implemented budget with total public expenditure (taking into account the sum of central, state and local gov- ernment) over the same period 20lƼ–202l, we aut that, in the seven countries surveyed, national policies for cohesion account for a total of 0.2 % of public expenditure, a tiny fraction. Again, there are huge differences between countries: in Roma- nia, this figure is over l %, in Czechia it is almost
ures, respectively. Overall, in these seven coun- tries, therefore, national investment policies for cohesion appear to be predominantly centralised in terms of design, but both regional and local au- thorities have a significant role in implementation.
3.Sub-national public finances
and investment
3.1The national context: public finances on the way to a gradual improvement aŁer the COVID-19 crisis and the
The Eighth Cohesion Report described the sig- nificant improvement in the public finances of EU Member States in the years following the Great Recession of 2008–2009 and the sovereign debt crisis of 20ll. While there was fiscal consolidation to reduce budget deficits in the period after 20ll, which was supported by economic recovery from 20lƼ to 20l9, trends were abruptly reversed in 2020 with the outbreak of the COVID-l9 pandem- ic and the restrictive measures taken to contain it,
along with the financial support provided to aut- guard businesses and jobs. In 202l, the EU deficit started to decline, as a result of a reduction in expenditure on pandemic-related emergency measures, combined with a recovery of GDP from the collapse the year before. The decline continued in 2022, despite government spending on energy support measures in response to the en- ergy crisis triggered by the war in Ukraine.
0.6 %, while in the other five countries it is less
than 0.l %.
While recognising that a national investment pol- icy for cohesion may cover different policy ar- eas, it can be seen that Ƽ0 % of the measures include the area ‘business & enterprise’, while areas such as ‘connectivity’, ‘human capital’ and ‘living standards’ are each included in around a third of the measures; l7 % of the measures in- clude ‘climate change & environment’, while 6 % include ‘research & innovation’. In terms of poli- cy instruments, the vast majority of the measures identified (94 %) mainly use grants and transfers, although some also offer interest rate subsidies (l4 %), tax breaks (8 %) or loan guarantees (3 %), sometimes used in combination.
Energy crisis
In order to fully understand the situation and evolution of sub-national public finances in EU Member States, it is important to set aut the macro-economic context in which they operate. Far from having a uniform impact across countries, macro-economic factors often have strong asym- metric effects that constrain the potential room for manoeuvre of sub-national finances. This is particularly true in the recent crises triggered by the COVID-l9 pandemic and the Russian war of aggression against Ukraine. The section provides an overview of the markedly heterogeneous situ- ation of national public finances across the EU in recent years.
Government deficits in the EU was 0.Ƽ % of GDP in 20l9, the same as before the Great Recession. In 2020, the deficit rose sharply to 6.7 % of GDP, as a result of the exceptional fiscal measures tak- en by Member States in response to the economic downturn caused by the restrictions imposed to contain the pandemic, the automatic stabilisers triggered by it, and the fall in GDP. As economies recovered, the deficit fell to 3.3 % of GDP in 2022. The phasing-aut of energy support measures is expected to reduce the deficit further.
A similar counter-cyclical pattern is evident for government debt. Consolidated government debt in the EU rose to 87 % of GDP in 20l3 and grad- ually declined to 78 % in 20l9. It rose sharply to 90 % of GDP in 2020, before falling back to 84 % in 2022, with a further decline forecast for 2023–202Ƽ.
The government financial balances in the different EU Member States in 2020 and 2022 give an in- dication of the changes in public finances induced by the pandemic and the energy crisis (Figure 8.3). In 2020, only Denmark had a budget surplus, while 2Ƽ Member States had a deficit of over 3 % of GDP, with Spain having the largest at l0 % of GDP, followed by Greece, Malta, Italy and Romania, all above 9 %. In 2022, the deficit in the EU was re- duced by half, with six Member States recording a
surplus, while l2 still had a deficit of 3 % or more, with Italy having the largest at 8 % of GDP, fol- lowed by Romania and Hungary at 6 %.
The effect of the post-pandemic recovery is also evident in public debt levels, which decreased in 2022 relative to GDP as compared with 2020 in 23 Member States (Figure 8.4), with reductions of over 30 pp in Greece, over 20 pp in Cyprus and Portugal, and over l0 pp in Croatia, Ireland, Ita- ly, Denmark and Sweden. In 2022, public debt remained above l00 % of GDP in six countries (Greece, Italy, Portugal, France, Spain and Belgium), being largest, on average, in the southern Member States (l30 % of GDP) and smallest in the eastern ones (Ƽ0 %).
3.2Sub-national governments carry aut a large share of public expenditure, but with marked differences across the EU
This sub-section examines government expendi- ture and revenue at regional and local eve land the changes that have occurred in recent years, in- cluding in response to the COVID-l9 pandemic and the energy crisis of 2022. Around a third of total government expenditure in the EU-27 is carried aut by regional and local authorities, highlighting their importance in the delivery of public servic- es, and their fundamental role in the functioning
of the public sector. However, there is substantial variation across Member States in the formal ex- tent of decentralisation of government expendi- ture and revenue (Box 8.2).
It is important to note that the figures for govern- ment expenditure and investment carried aut at the sub-national eve land the revenues collected at this level indicate the amounts that are channelled through the authorities concerned. While these may be responsible for managing expenditure or
collecting revenue, they may have limited auton- omy over the underlying policy and the decisions on investment or taxes. Section 4 below sets aut an exploratory examination of the composition of revenue and expenditure, which might shed some light on the actual decision-making powers of re- gional and municipal authorities.
In 2022, sub-national expenditure and revenue in the EU were both l7 % of GDP, or around a third of total government spending, and slightly
Figure 8.5 Total and sub-national government expenditure and revenue in the EU, 2004–2022
Expenditure – Total
Revenue – Total
Expenditure – Sub-national
Revenue – Sub-national
55
Ƽ0 4Ƽ
40
35
30
2Ƽ
20
lƼ l0
2004 200Ƽ 2006 2007 2008 2009 20l0 20ll 20l2 20l3 20l4 20lƼ 20l6 20l7 20l8 20l9 2020 202l 2022
expenditure in total government spending, reflect- ing in part differences in the institutional make- up (Figure 8.6). The share is largest in federal countries (Austria, Belgium and Germany) and in those where government is highly decentralised (Denmark, Spain, Sweden and Finland). In Den- mark, around two thirds of public expenditure in 2022 was carried aut by sub-national authorities; in Spain, Sweden, Germany and Belgium, around half; and in Finland, over 40 %. By contrast, in Cy- prus and Malta, reflecting their size, sub-national authorities were responsible for under Ƽ % of pub- lic expenditure, and in Greece, Ireland and Luxem- bourg, only around l0 % or less.
Although the share of expenditure carried aut by sub-national authorities in the EU has been stable
Source: Eurostat gov_l0a_main.
More of revenue (Figure 8.Ƽ). The share of GDP has been very stable over time – in 2004, it was just over l6 %. In the same way as the total, sub-national government expenditure varies coun- ter-cyclically with GDP, tending to increase as a share when GDP falls and to fall when it increas- es. The share increased sharply in 2020, jumping by l.7 pp as a consequence of the pandemic and the measures taken in response to it, and falling back in the following two years as GDP recov- ered. In 2022, it was l.l pp lower than in 2020,
though 0.6 pp higher than before the pandemic. Sub-national revenue also increased in 2020, by
l.2 pp to almost l8 % of GDP, and in 2022 it was still 0.4 pp higher than before the pandemic, partly because of increased transfers from central gov- ernments to combat the pandemic and to recover from the recession caused by the restrictive meas- ures taken.
There are significant differences between Member States in the share of sub-national government
over time, this is the result of differing develop- ments across Member States. Between 20l0 and 2022, the share increased in eight Member States and declined in lƼ. More specifically, it increased by around 8 pp in Belgium, by over 3 pp in Den- mark and Germany, and by 2 pp in Sweden and Ireland, while it fell by over l pp in ll countries, declining by 6 pp in Italy and l3 pp in Hungary.
Overall, government expenditure tends to be significantly less decentralised in less devel- oped Member States than in more developed ones, with sub-national spending accounting for l8 % of expenditure in the former in 2022 and 36 % in the latter. Over the period 20l0–2022, expendi-
(Figure 8.7). Over the period 2004–202l, there was a marked and almost continuous increase in the decentralisation of spending on general public services (by 8.2 pp, equivalent to an increase of almost 20 %) and education (by 4.l pp, or 7 %). Sub-national expenditure in other areas, on the other hand, fell, in economic affairs (by 8.Ƽ pp, or l7 %), health (by 3.4 pp or 9 %) and environmen- tal protection (by 4.9 pp, or 6 %).
Figure 8.6 Sub-national government expenditure in EU Member States, 2010, 2014, 2018
and 2022
20l0
20l4
20l8
2022
80
% of total government expenditure
70
60
Ƽ0 40
30
20
DN ES BE SE DE FI AT PL NL CZ HR IT LV LT EE RO FR SI SN BG PT HU LU IE EL CY MT
Less developed
More developed
EU-27
l0 0
ture became less decentralised in less developed countries, with the sub-national share falling by
l.6 pp, while it increased by 0.Ƽ pp in the more developed ones.
Sub-national government expenditure tends to be concentrated in certain policy areas (see Box 8.3 for a description of the breakdown by function).
In 202l, sub-national authorities were responsible for eve lan 82 % of public expenditure on envi- ronmental protection3 and 66 % of education ex- penditure, as well as almost Ƽ0 % of spending on general public services, 4l % of spending on eco- nomic affairs4, and over a third of that on health
Social protection was the largest area of sub-na- tional government expenditure in the EU in 202l, accounting for 3.6 % of GDP, followed by education at 3.2 %, general public services at 3 %, health at
2.7 % and economic affairs at 2.6 %, while ex- penditure on environmental protection was just
0.7 % of GDP (Figure 8.8).
Again, there is considerable variation between Member States. Overall, the expenditure carried aut by sub-national authorities relative to GDP in less developed countries was only just over half of that in more developed ones (l0 % as against l9 %). Spending in all areas was lower in the former,
Source: Eurostat gov_l0a_main.
3The COFOG category ‘environmental protection’ includes waste and wastewater management activities.
4The COFOG category ‘economic aRairs’ includes transport and communication services, which represent a large share of expenditure.
Figure 8.7 Sub-national government expenditure in selected policy areas in the EU, 2004, 2010,
2016, 2019 and 2021
2004
20l0
20l6
20l9
202l
% of General government expenditure in the respective areas
l00 90
80
70
60
Ƽ0 40
30
20
l0 0
and Greece, and again zero in Cyprus and Malta. Health expenditure was just under 9 % of GDP in Denmark, around 7 % in Italy, Sweden and Spain, and around 6 % in Finland and Austria, but well below l % in ll countries.
3.3Sub-national governments undertake the majority of public investment
Sub-national authorities have a major responsi- bility for public investment, more than for public expenditure as a whole. Over half of public in- vestment in the EU is carried aut by sub-national governments – over the period 2004–2022, their
suggesting potential scope for further regionalisation in less developed countries. While, however, public investment as a share of GDP has tended to vary pro-cyclically in the two groups, declining dur- ing economic downturns and increasing during up- turns, the variation has been more pronounced in less developed countries than in more developed ones (Figure 8.l0).
In 2022, public investment carried aut by sub-na- tional governments was particularly high in rela- tion to GDP in Finland and Sweden (2.3–2.4 %). It was also over 2 % in Slovenia, Romania, Czechia, Belgium and France, but below l % in Ireland, Cy- prus and Malta. In general, countries with relative- ly low sub-national public investment also have low total public expenditure at sub-national level
General public services
Economic affairs
Environmental
protection
Source: Eurostat gov_l0a_exp.
Health
Education
expenditure on investment accounted for between
Ƽ4 % and Ƽ8 % of the total carried aut by govern- ment (Figure 8.9). Regional and local authorities, therefore, have a key role in providing the infra- structure to support development. At the same
(Figure 8.ll).
There has been no uniform pattern of change in sub-national public expenditure in relation to GDP over the past decade or so. In l4 Member States,
especially on social protection (2.Ƽ pp lower), gen- eral public services (2.l pp lower), health (l.4 pp lower), education and economic affairs (l pp lower in both).
The differences between countries are even more marked. Sub-national expenditure on social protection was almost l8 % of GDP in Denmark, around 6 % or over in Belgium, Sweden, Germa- ny and Finland, but only around l % or less in
l7 Member States and zero in Malta and Cyprus. Expenditure on general public services at sub-na- tional level was above Ƽ % of GDP in Spain and Germany, over 4 % in Belgium and Finland, but below l % in l2 Member States. Expenditure on education at this level was 7 % of GDP in Belgium, around Ƽ % in Sweden and Germany, and around 4 % in Spain, the Netherlands, Czechia, Croatia, Latvia, Finland and Estonia, but below l % in Italy, Hungary, Portugal, Luxembourg, Romania, Ireland
time, the sub-national share of public investment is smaller in less developed countries than more developed ones – 42 % of total investment in 2022 as against Ƽ9 % – although the difference declined by over ll pp between 2004 and 2022.
As a share of GDP, total public investment in the less developed countries has been consistently higher than in the more developed ones over the last two decades (Figure 8.10), also due to the key role of Cohesion Policy support in the former. At the sub-national level, public investment as a share of GDP was of a similar magnitude in both more developed and less developed countries over the period 2004–2022,
it was higher in 2022 than in 20l3, most notably in Luxembourg, Croatia and Greece (0.Ƽ pp higher), while in l2 Member States it was lower, notably in Bulgaria and Latvia.
Cohesion policy multiannual programming has been a key driver of public investment integration in medium-term budgetary frameworks and pub- lic financial management structures. Integrated strategic planning and appraisal/selection models
Figure 8.8 Sub-national government expenditure in selected policy areas, by EU Member States, 2021
Figure 8.9 Sub-national public investment in the EU and in more developed and less developed Member States, 2004–2022
General public services
Economic affairs
Environmental protection
Health
Education
Social protection
Others
35
30
2Ƽ
20
lƼ l0 5
DN BE ES DE SE FI AT IT PL NL HR CZ FR LV EE LT SI RO BG SN PT HU LU EL IE CY MT
Less developed
More developed
EU-27
0
Less developed
More developed
EU-27
70
65
% of total government investment
60
55
Ƽ0 4Ƽ
40
35
30
2Ƽ
20
lƼ l0 5
0
2004 200Ƽ 2006 2007 2008 2009 20l0 20ll 20l2 20l3 20l4 20lƼ 20l6 20l7 20l8 20l9 2020 202l 2022
Source: Eurostat gov_l0a_exp.
Source: Eurostat gov_l0a_main.
Figure 8.10 Sub-national and total public investment in more developed and less developed Member States, 2004–2022
Less developed – Sub-national
More developed – Sub-national
Less developed – Total
More developed – Total
Ƽ.Ƽ
Ƽ.0
4.Ƽ
4.0
3.Ƽ
3.0
2.Ƽ
2.0
l.Ƽ
l.0
2004 200Ƽ 2006 2007 2008 2009 20l0 20ll 20l2 20l3 20l4 20lƼ 20l6 20l7 20l8 20l9 2020 202l 2022
Source: Eurostat gov_l0a_main.
4.New evidence on regional
and local finances
Sub-national public finances are examined in more detail below in order to better understand the role of sub-national governments in the institutional ar- chitecture of Member States, and ultimately to as- sess their degree of autonomy over decision-mak- ing. This is based on an initial, and still preliminary, dataset showing the relationship between current and capital expenditure and between different rev- enue sources for the regional and municipal lev- els of government in several EU Member States, developed by the OECD in collaboration with the Directorate-General for Regional and Urban Policy (DG REGIO)6.
4.1A comparative overview of current and capital expenditure
It should be noted that regional capital expenditure includes the contribution from EU funding, which is particularly important in regions with more respon- sibility for investment programmes and for regional development more generally and less responsibility for service-provision (Box 8.4).
Current expenditure exceeded capital spending in the regions of almost all countries in 2020, im- plying that a major proportion of regional govern- ment revenue was spent on personnel costs and purchases of goods and services.
Capital expenditure amounted to only just over l8 % of the total on average in the countries covered. This varied, however, from over 20 % in Czechia, Ro- mania, Poland, France and Greece to under l0 % in the Netherlands, Sweden, Denmark, Italy, Austria, Belgium and Spain, with Germany and Croatia in between. The share of investment in total regional
that effectively guide budget allocation eve la asset registers as input are key for the delivery of public investment. A recent paper discusses a num- ber of good practices across the public investment lifecycle, drawing on recent survey evidence from all EU Member States commissioned by DG ECFIN5.
A higher level of scrutiny. Similarly, EU financed investments tend to follow stricter rules through- aut the project cycle than nationally financed ones. However, evidence also points to wide-ranging re- forms of public investment management systems in several Member States, while room for improve-
Figure 8.l2 compares current and capital expend- iture for 2020 of regional governments in the l4 EU Member States included in the regional gov- ernment finance and investment database (REGOFI).
Expenditure was largest in Greece, where regions are mainly responsible for regional planning and development, much of which is financed by fund- ing under EU Cohesion Policy. Regions in Poland,
Overall, it finds that more sizeable projects tradi- tionally in the transportation sector are subject to
ment is evident across many Member States.
Figure 8.12 Breakdown of regional government expenditure in selected EU Member States, 2020
20
0
NL
SE
DN
IT
AT
BE
ES
DE
HR
CZ
RO
PL
FR
EL
0.0
Source: OECD, MUNIFI and REGOFI Databases 2024.
FI SE SI RO CZ BE FR PL ES DE LU LV HR EE NL IT DN AT HU SN EL PT LT BG IE CY MT
Less developed More developed
EU-27
Source: Eurostat gov_l0a_main.
5Belu Manesco (2022).
6The dataset consists of two databases, REGOFI and MUNIFI (municipal nscal data), which currently cover 2l EU Member States at the municipal eve land l4 at the regional level. They were built using a standardised methodology in collaboration with the national statistical institutes of most of the countries covered to facilitate in-depth comparison of the revenue, expenditure and investment pronles of regions and municipali- ties across countries. REGOFI covers regions denned at NUTS 2 level (nomenclature of territorial units for statistics) in all the EU Member States surveyed, except Belgium and Germany, where regions are denned at NUTS l level. The two databases cover only the regional and municipal levels and do not include other territorial units that fall between the two, such as Belgian provinces, French departments or Italian metropolitan cities, the public nnances of which are included in Eurostat’s sub-national government statistics. See: OECD (2024).
which devoted around a third of their expenditure to investment, are also large recipients of Cohesion Policy funding and tend to play a relatively limited role in the provision of public services (for the 2014–2020 period, Cohesion Policy funding corresponded to around 13 % of public investment in the EU as a whole and to 51 % in the less developed Member States, see Chapter 9, section 8). Similarly, in France, where the regions are responsible for their economic development, non-urban transport and spatial planning, capital expenditure accounted for 37 % of total regional public expenditure in 2020. When the share of capital expenditure is higher, the margins for adjusting the level and allocation of current expenditure in response to emerging exceptional circumstances may be limited, and public expenditure management should therefore be particularly careful. On the other hand, the share of capital spending in total government expenditure at regional level was smallest in the Netherlands, Denmark and Swe- den, where regional authorities have large respon- sibility for public services, such as healthcare, and administrative tasks. Regions in these countries also accounted for a smaller share of sub-national investment than local authorities. Figure 8.l3 shows personnel costs as a share of total government expenditure at regional level for the l4 EU Member States covered. Personnel costs accounted for a particularly large share in Swe- den, Denmark and Spain (over 40 %), but less than l0 % in the
Netherlands, Czechia, Croatia and Italy (only 3 % in the last). Figure 8.l4 shows that, in all the 2l Member States for which municipal data are included in the database, current spending was the largest com- ponent of total government expenditure at this level in 2020. Capital expenditure accounted for just under l9 % of total municipal expenditure, on average, much the same as for regional govern- ment, although the set of countries covered is dif- ferent and a comparison not meaningful. Again, there is substantial variation between coun- tries. Capital expenditure in municipalities was only around l0 % or less of total spending in the Netherlands, Denmark, Austria, Sweden and Fin- land, but over 20 % in Latvia, Lithuania, France and Portugal and over 30 % in Ireland, Romania, Slovenia and Croatia, in the last 4l %. In the last three countries, municipalities have the main re- sponsibility for urban development, transport and housing. On the other hand, the small share of capital expenditure, and the correspondingly large share of current spending, in the first group of countries reflects their major role in the provision of education and social services (and social pro- tection in Denmark). Figure 8.lƼ shows personnel costs in 2020 as a share of total expenditure at municipal level for the Member States covered. These accounted for over Ƽ0 % of the total in Belgium and Sweden and over 40 % in Lithuania, Estonia, Latvia and France, while they accounted for under 20 % in Croatia, Austria, the Netherlands and Malta, and under l0 % in Slovenia and Czechia.
Figure 8.13 Regional personnel cost as a share of total regional expenditure in selected EU Member States, 2020
Ƽ0
% of total regional government expenditure
40
30
20
l0
0
SE
DN
ES
AT
DE
BE
RO
EL
PL
FR
HR
CZ
NL
IT
Figure 8.14 Breakdown of municipal expenditure in selected EU Member States, 2020
Box 8.5 Building resilience: the need for diversified revenue sources
l00
% of total municipal expenditure
80
60
40
20
0
Capital expenditure
Current expenditure
NL
DN AT
SE
FI
ES
BE
PL
CZ
EE
HU
IT
MT
LV
LT
FR
PT
IE
RO
SI
HR
In an era of unprecedented challenges and crises, the ability of sub-national governments to respond effectively depends on their capacity to adapt both the level and the composition of expenditure to changing circumstances. This requires access to fi- nancing, to taxation or borrowing. Where borrowing is constrained (usually by central government re- strictions) – because, for example, of tight monetary conditions, as in the aftermath of the COVID-l9 and energy crises – the key factor in ensuring financ- ing at sub-national level is the diversity of revenue sources available.
Diversified revenue sources give sub-national gov- ernments operational flexibility, while overdepend- ence on a main single source increases vulnerabili-
sources, sub-national governments can better with- stand shocks. A balanced mix of sources, such as revenue from assets, user fees, grants, and taxes contributes to fiscal resilience, acting as a buffer and giving financial stability when one source is ad- versely affected.
The importance of cultivating flexibility in revenue sources for sub-national governments cannot be overstated. The ability to weather crises, respond skilfully to unforeseen challenges and promote long-term sustainability depends on the diversifica- tion of revenue streams. By adopting a multi-fac- eted approach to revenue generation, sub-national governments can strengthen their fiscal resilience and ensure the well-being of their constituents in
Source: OECD, MUNIFI and REGOFI Databases 2024.
ty, especially during crises. By diversifying revenue
the face of an ever changing world.
Figure 8.15 Personnel cost as a share of total municipal expenditure in selected EU Member States, 2020
70
% of total municipal expenditure
60
Ƽ0 40
in deciding and managing their finances. Regional governments have different degrees of control over tax rates and provisions, especially with regard to shared taxation, i.e. national taxes where a specified proportion of the revenue raised is allocated to re- gional or other sub-national authorities7.
In general, the main source of regional govern- ment revenue in 2020 was grants and subsidies,
i.e. transfers from central government and the EU, accounting on average for half of the total reve- nue (see Box 8.6 on the challenges of managing transfers between different levels of government).
30
Figure 8.16 Breakdown of regional government revenue in selected EU Member States, 2020
20
Tax revenue
Grants and subsidies
User charges and fees
Income from assets
Other revenue
l0
l00
% of total regional government revenue
0
BE
SE
LT
EE
LV
FR
DN
ES
PT
FI
HU
PL
RO
IE
IT
MT
NL
AT
HR
CZ
SI
80
Source: OECD, MUNIFI and REGOFI Databases 2024.
4.2Municipal and regional revenue sources
This section examines the revenue sources used to finance regional and municipal government ex- penditure. Relying on a single or only a few revenue sources as opposed to having a more diverse mix has important implications for the sustainability and resilience of public finances at sub-national level. Other things being equal, reliance on a few sourc- es generally means less resilience to shocks and
changing socio-economic conditions. Resilience can, therefore, be improved by diversification of revenue sources, but effective institutions and mechanisms need to be in place to achieve this (see Box 8.Ƽ).
Figure 8.l6 shows the breakdown of regional reve- nue sources for l4 EU Member States in 2020. It is important to note that a larger share of revenue from taxes as compared with, for example, transfers from central government does not automatical- ly mean a higher degree of autonomy for regions
60
40
20
0
FR
HR
SE
DE
CZ
RO
ES
BE
IT
PL
NL
AT
DN
EL
Source: OECD, MUNIFI and REGOFI Databases 2024.
7In Germany, for example, tax revenue is the main source of revenue for the Länder, but they have little influence over it, as most comes from shared taxation (from personal and corporate income tax and value added tax).
Figure 8.17 Breakdown of municipal revenue in selected EU Member States, 2020
Tax revenue
Grants and subsidies
User charges and fees
Income from assets
Other revenue
l00
% of total municipal revenue
80
60
40
20
0
SI
DN
HR
FR
LV
EE
FI
SE
RO
ES
LT
IT
PL
BE
HU
PT
CZ
AT
NL
IE
MT
Source: ECD, MUNIFI and REGOFI Databases 2024.
This revenue source was the only one present in all l4 countries covered, ranging from 94 % in Greece, over 70 % in Denmark and Italy and over Ƽ0 % in Belgium, Spain and Romania to under 30 % in Aus- tria, France, Croatia and the Netherlands.
The second major source of revenue at regional level is taxes, including both shared and own-im- posed, which, on average, accounted for a third of total regional government revenue in 2020. It is notable that regions in both Denmark and Greece had no revenue from taxes, reflecting their lack of tax-raising power. Much the same was the case in Austria, where taxes accounted for under Ƽ % of revenue. By contrast, in Sweden and Germany over ƼƼ % of regional government revenue came from taxes and over 6Ƽ % in France and Croatia.
User charges and fees and asset-based revenue made up a much smaller share of government rev- enue at regional level, averaging just under 4 % and just over 6 %, respectively. However, in Swe- den and Denmark, user charges and fees account- ed for over l0 % of revenue, and in the Nether- lands, asset-based revenue for over half.
Funding sources at regional level are most diverse in Poland, the Netherlands, Austria and Sweden, while they are most concentrated in Greece, Den- mark, Italy, France and Croatia.
Contrary to the situation at regional level, trans- fers and taxes were of a similar weight in fund- ing municipal governments in 2020 (Figure 8.l7), each accounting for around 40 %. However, differ- ences between Member States are again consider- able. The most diverse mixes of funding sources at this level were in Poland, Austria and Portugal, fol- lowed by Finland, Sweden, Italy, Belgium and Hun- gary, while they were most concentrated in Malta, Ireland, Czechia and Slovenia.
References
Anam, C., Plaček, M. (2023), ‘The Development of Fiscal Decentralization and Its Impact on Economic Growth’, The Palgrave Handbook of Global Social Problems, Palgrave Macmillan, Cham, pp. l–28.
Aray, H., Pedauga, L. (2024), ‘The relationship between decentralization and economic growth across regimes’, Annals of Regional Science, 72, pp. l–2Ƽ.
Bachtler, J., Downes, R. (2023), Rethinking Regional Transformation: The State of Regional Policy in Europe, EoRPA Report 23/l, EoRPA Regional Policy Research Consortium, European Policies Research Centre, University of Strathclyde, Glasgow and EPRC, Delft.
Belu Manescu, C. (2022), ‘New Evidence on the Quality of Public Investment Management in the EU’,
European Economy Discussion Papers, l77, Publications Office of the European Union, Luxembourg.
Bergvall, D., Charbit, C., Nraan, D.–J., Merk, O. 2006), ‘Intergovernmental Transfers and Decentralised Public Spending’, OECD Working Papers on Fiscal Federalism, 3, OECD Publishing, Paris.
Buser, W. (20ll), ‘The impact of nscal decentralization on economics performance in high–income OECD nations: an institutional approach’, Public Choice, l49, pp. 3l–48.
Canavire–Bacarreza, G., Martinez–Vazquez, J., Bauyrzhan Y. (2020), ‘Identifying and disentangling the impact of nscal decentralization on economic growth’, World Development, l27, l04742.
European Commission (2023), ‘European Economic Forecasts – Autumn 2023’, European Economy Institutional Paper No 258, Directorate–General for Economic and Financial ARairs, Brussels.
European Commission (forthcoming), Study on National Investment Policies and Cohesion – Final Report, Contract No 202lCEl60AT009 (Technopolis Group, CSIL – Centre for Industrial Studies and OPIX consulting), Publications Office of the European Union, Luxembourg.
European Commission (20l9), Study on National Policies and Cohesion – Final Report, Contract
No 2017CE16BAT125 (Prognos AG, Politecnico di Milano and Technopolis Group), Publications Office of the European Union, Luxembourg.
Eurostat (20l9), Manual on sources and methods for the compilation of COFOG statistics – 2019 edition, Publications Office of the European Union, Luxembourg.
Filippetti, A., Sacchi, A. (20l6), ‘Decentralization and economic growth reconsidered: The role of regional authority’, Environment and Planning C: Government and Policy, 34(8), pp. l793–l824.
Iimi, A. (200Ƽ), ‘Decentralization and economic growth revisited: an empirical note’, Journal of Urban Economics, Ƽ7(3), pp. 449–46l.
Lago, M. E., Lago–Peñas, S., Martinez–Vazquez, J. (2022), ‘On the ERects of Intergovernmental Grants: A Survey’, Working Paper 22–03, Georgia State University Andrew Young School of Policy Studies, International Center for Public Policy, Atlanta.
Oates, W. E. (l999), ‘An Essay on Fiscal Federalism’, Journal of Economic Literature, 37(3), pp. ll20– ll49.
OECD (2024), Going Granular with Regional and Municipal Fiscal Data in OECD and EU Countries, OECD Publishing, Paris.
OECD/UCLG (2022), 2022 Synthesis Report of the World Observatory on Subnational Government Finance and Investment – Highlights, OECD Publishing, Paris.
Spahn, P. B. (20l2), Conditioning Intergovernmental Transfers and Modes of Interagency Cooperation for Greater Effectiveness of Multilevel Government in OECD Countries, Paper prepared for the OECD workshop on ERective Public Investment at Sub–National Level in Times of Fiscal Constraints: Meeting the Co–ordination and Capacity Challenges, OECD Publishing, Paris.
Treisman, D. (2006), ‘Fiscal Decentralization, Governance and Economic Performance: A Reconsideration’, Economics and Politics, l8(2), pp. 2l9–23Ƽ.
EUROPEAN COMMISSION
Brussels, 27.3.2024
SWD(2024) 79 final
COMMISSION STAFF WORKING DOCUMENT
Accompanying the document
Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions
on the 9th Cohesion Report
{COM(2024) 149 final}
THE IMPACT OF COHESION
9
•Macro-economic model simulations indicate that the 2014–2020 and 2021–2027 programmes of Cohesion Policy investment will have increased EU GDP by almost 1 % by 2030, at the end of the implementation period.
•The same model indicates that all EU regions – including the most developed
•ones, benefit from the investment financed under Cohesion Policy
•This shows that Cohesion has delivered on its mission to promote convergence and harmonious development, as well as contributed to support EU competi- tiveness and investment to help create a greener, more connected and socially integrated Europe. It also helped finance the response in EU Member States to the COVID-19 pandemic.
•A great many studies and evaluations have shown that Cohesion Policy has had a significant impact on the socio-economic development of EU regions, especially in the less developed ones. The increase is particularly large in less developed regions; in several less developed regions GDP is expected to be 10 to 13 % higher by 2030 than it would have been without Cohesion Policy. Cohesion Policy therefore contributes to reducing regional disparities, both at EU level and within Member States.
•The conditions imposed on the receipt of Cohesion Policy funding starting from the 2014–2020 period, along with the technical assistance provided, have helped to improve institutional capacity across the EU, the overall investment environ- ment, and the ability of Member States to make the best use of EU support. They have also helped speed up reforms, by raising political awareness of their need and reinforcing the commitment of governments to them.
.
Chapter 9
The impact of Cohesion Policy
Box 9.1 Thematic priorities
In the 2014–2020 programming period, the invest- ment financed under Cohesion Policy was aimed at supporting 11 broad priorities or Thematic Objec- tives, as follows.
2.Greener Europe (including a low-carbon econo- my, climate action, protecting the environment, and clean urban transport - corresponds to the 2014-2020 thematic objectives 4, 5 and 6).
1.Introduction
The sustainable development of all regions in the EU is important for its prosperity economic, social and territorial cohesion. Cohesion Policy has con- tributed substantial funding to support Member States and regions to overcome obstacles to their socio-economic development and reduce territorial disparities across the EU. Cohesion Policy is firmly place-based, which means that most programmes are adapted to the specific needs of individual re- gions, so providing tailored responses to develop- ment challenges to the local context.
This chapter reviews the features of Cohesion pol- icy and the evidence relating to its impact. It high- lights the place-based nature of the policy and summarises some of the main achievements of the 2014–2020 programming period. It also ex- amines the 2021–2027 programmes and the way that they support the political priorities of the EU. It ends by assessing the impact of the 2014–2020 and 2021–2027 programmes on GDP across the EU, and on less developed regions in particular.
2.Achievements and evaluation of the 2014–2020 programme
Under the EU budget’s 2014–2020 Multiannual Fi- nancial Framework, Cohesion Policy was the EU’s main means of funding investment in economic and social development across the EU. As of De- cember 2023, EUR 405 billion of support1 had been committed under the 2014–2020 programmes, which, with national (public and private) co-financ- ing, is estimated to have resulted in EUR 551 bil- lion of investment. The support came from three funds: the European Regional Development Fund (ERDF), the Cohesion Fund (CF) and the Europe- an Social Fund (ESF), supplemented by the Youth Employment Initiative (YEI). Financing from these was aimed at 11 Thematic Objectives, 10 of which
12014–2020 figures include Interreg (UK, and REACT-EU).
2European Commission (2024).
for the 2021–2027 period were transformed into five Policy Objectives (see Box 9.1 and Figure 9.1). To enable comparisons to be made between the two periods, these 10 Thematic Objectives, and the expenditure under them, have been mapped for the analysis here to the five Policy Objectives.
The ERDF financed projects under all 11 Thematic Objectives listed in Box 9.1, but predominantly those under the first seven. Four Objectives (the first four in the box) – ‘Strengthening research, technological development and innovation (RTDI)’, ‘Enhancing ac- cess to, and the use and quality of, ICT’, ‘Enhancing the competitiveness of SMEs’ and ‘Supporting the shift towards a low-carbon economy’ – accounted for between 50 % and 80 % of total ERDF expendi- ture in Member States, the share varying according to the level of development. A larger share went on these four Objectives in the more developed coun- tries and regions, and a larger share on the other three in the less developed ones, particularly on environmental and transport infrastructure, under Thematic Objectives 6 and 7, which was the focus of the CF. Although the ERDF also financed investment under Thematic Objectives 8–11 (on employment, social inclusion, education and training, and institu- tional capacity), current expenditure, as opposed to capital expenditure, was financed by the ESF.
The following sections review the progress made up to the end of 2022 in spending the funding allocated for the 2014–2020 period, the output and results so far achieved, and the findings from evaluations car- ried out up to now by Member States. A more detailed presentation of the implementation of 2014–2020 programmes is contained in the Commission’s 2023 annual summary of implementation reports, while more details of national evaluation findings are set out in the Commission’s annual summary2. The ex post evaluation of the 2014–2020 programmes is being carried out at present and will be published between end–2024 and mid–2025 (see Box 9.2).
During this period, the Union faced several crises which required exceptional measures to support Member States and regions. This implied adjusting the policy objectives to changing priorities and, in a some cases, targets are likely to underachieved and in other case overachieved compared to the original programmes.
1.Strengthening RTDI.
2.Enhancing access to, and the use and quality of, ICT.
3.Enhancing the competitiveness of SMEs.
4.Supporting the shift towards a low-carbon
economy.
5.Promoting climate change adaptation, risk pre- vention and management.
6.Preserving and protecting the environment and
promoting resource-efficiency.
7.Promoting sustainable transport and removing bottlenecks in key network infrastructures.
8.Promoting sustainable and high-quality em- ployment and supporting labour mobility.
9.Promoting social inclusion, and combating pov- erty and discrimination.
10.Investing in education, training and vocational training for skills and lifelong learning.
11.Enhancing the institutional capacity of public
authorities and efficient public administration.
In the 2021–2027 programming period, the first 10 Thematic Objectives have been replaced by five Policy Objectives, as follows.
1.Smarter Europe (including RTDI, digital econ- omy, and SME competitiveness - corresponds to the 2014-2020 thematic objectives 1, 2
and 3).
2.1Policy Objective: Smarter Europe
The Smarter Europe Policy objective aims to con- tribute to a more competitive and smarter Europe by promoting innovative and smart economic transformation and regional ICT connectivity.”
In 2014–2020, Cohesion Policy provided ERDF support of EUR 96 billion (24 % of total Cohesion
3.More connected Europe – the trans-European transport network (TENT-T) and other trans- port priorities (corresponds to the 2014-2020 thematic objective 7).
4.Social Europe (employment and labour market measures, social inclusion, and human capital).
5.Europe closer to citizens.
For the sake of consistency and to facilitate compar- ison between the two programming periods, in this chapter the 11 Thematic Objectives are mapped, approximately, to the new Policy Objectives as listed above.
Following the COVID-19 crisis, in 2021–2022 an additional Objective of ‘Fostering crisis repair and resilience’ was introduced, financed from REACT-EU with a budget of EUR 50 billion as part of the Next- GenerationEU (NGEU) recovery package.
For the 2014–2020 period, the present chapter sets out figures for the EU shares of planned invest- ments, the amounts allocated to the projects select- ed for funding, and expenditure on the five Policy Objectives. The financing and indicator data go up to the end of 2022 (the latest date for which data are available). It should be noted that the amount allocated to projects selected for funding can ex- ceed the EU funding available since it is often the case that more projects are selected than can be financed so as to ensure that all the funding avail- able is ultimately spent, given a belief that not all projects selected will actually come to fruition.
Policy funding) to enhance RTDI, ICT infrastruc- ture and services, and SME competitiveness. Up to the end of 2022, estimated expenditure on these amounted to around 94 % of the total allocated to them.
The common indicators give an indication of the outputs across the EU from this investment and how they relate to the targets set.
Figure 9.1 EU Cohesion Policy budget (2014–2020) approximated to 2021–2027 Policy Objectives
Smarter Europe
Box 9.2 Progress in the Commission’s ex post evaluation of 2014–2020 programming
Greener Europe
More connected Europe
Social Europe
Other (REACT-EU, Outermost, Technical assistance)
Smarter Europe
96 669.8
90 807.4
94 %
Greener Europe
69 060.8
55 332.8
80 %
assistance)
Notes: The funding allocated to the 11 Thematic Objectives (and multithematic priorities) for 2014–2020 is mapped to the 4 main Policy Objectives for 2021–2027 (see Box 9.1). Data as at 31 December 2022 (which are not final values as spending is ongoing; formal closure of programmes will occur only in 2025).
Source: DG REGIO calculations based on Cohesion Open Data.
The Commission launched its ex post evaluation of 2014–2020 ERDF and CF programmes with a view to completing it in 2025. The evaluation is com- posed of: four cross-cutting work packages – on Interreg, Integrated Territorial Investment (ITI), the response to the COVID-19 pandemic, and the mac- ro-economic effects of Cohesion Policy; seven work packages covering all the 2014–2020 Thematic Ob- jectives; and a work package for creating a database of projects to be used in the evaluation. A synthesis report will summarise the results of the evaluation.
The thematic work packages adopt a theory-based approach to evaluating the effects of the invest- ments financed. For each Thematic Objective, the theory of change – or logic – underlying the policy instruments used to pursue the policy aims is first spelled out, identifying the various steps by which each instrument is assumed to achieve these aims and the links between them, as well as the condi-
The final reports of the work packages will be pub- lished in the second half of 2024, providing as- sessments of how the various programmes have performed over the period, which will be used to prepare proposals for the next period. They will also assess the contribution of Cohesion Policy to the pursuit of its ultimate goals. The final synthesis re- port is scheduled to be published in spring 2025. The Commission’s conclusion on the evaluation, in the form of a staff working document, will then be finalised later in 2025.
The Commission is in parallel carrying out an ex post evaluation of the ESF and YEI for the 2014–2020 period. It will assess the performance of the pro- grammes financed in the same way as for the ERDF and CF – i.e. in terms of their effectiveness, effi- ciency, relevance, EU added-value, and coherence with policy measures financed in other ways. It will consider the pursuit of all ESF priorities, including
•Over 2.36 million enterprises had received sup- port by the end of 2022 (109 % of the target).
•Nearly 370 000 jobs were directly created as a result of the expenditure (98 % of target).
•228 000 new enterprises were created (101 % of target).
•84 000 enterprises developed new-to-market or new-to-firm-products/services (102 % of target).
•7.88 million additional households had access to broadband (66 % of target). The final achieve- ment will be closer to the target if the projects already selected for funding are completed.
Funding for research and innovation went most- ly to increasing collaboration between compa- nies, particularly SMEs, and universities and oth- er research centres. The evaluations carried out in Member States have identified positive results from the support provided, such as in Romania, where support for research and development (R&D) and innovation increased the capacity of SMEs to develop new products and processes and improve worker competences; in Wallonia, where between 2014 and 2018 support helped increase
the survival rate of companies; and in Slovakia, where start-up SMEs had a significantly higher growth of value-added and employment over the period than those not supported.
Cohesion Policy funding has also helped to boost digitalisation and the development of ICT servic- es. In Corsica, it has enabled the development of new ways of learning adapted to students’ per- sonal needs, which have increased their motiva- tion and helped to reduce social and territorial di- visions. Equally, in Lithuania, it has increased the availability of e-services, with estimated savings of EUR 1.89 billion, mostly from people not having to travel to physical locations.
2.2Policy Objective: Greener Europe
The Greener Europe Policy Objectives contributes to a greener, low-carbon transitioning towards a net zero carbon economy and resilient Europe by promoting clean and fair energy transition, green and blue investment, the circular economy, climate change mitigation and adaptation, risk prevention and management, and sustainable urban mobility.
Cohesion Policy provided EUR 69 billion from the ERDF and CF for investment in the Greener Europe
tions that need to prevail for this to be successful. The evaluation then assesses how far the various steps in the theory of change can be observed in practice and how far the aims have actually been achieved, based on the evidence available or that can be collected. In the process, the performance of the programmes implemented by means of the poli- cy instruments will be judged in terms of their effec- tiveness, efficiency, relevance (in terms of meeting the needs identified), coherence (both internally and with other policy measures) and the EU added-val- ue they have generated. The work packages are be- ing carried out by independent contractors and the Commission is supported by experts who critically assess the reports that the contractors produce and the soundness of their findings.
Objective in 2014–2020. This funding targeted in- creases in: energy-efficiency and renewable ener- gy; improvements in environmental infrastructure; the development of the circular economy; miti- gation of, and adaptation to, climate change; risk prevention; biodiversity; and clean urban transport (Box 9.3). The amount allocated represented 17 % of the total funding available under Cohesion Policy for the period. By the end of 2022 the expenditure amounted to around 80 % of the total EU allocation
funding initiatives in response to the COVID-19 pan- demic and the effects of Russia’s war of aggression against Ukraine – i.e. the Coronavirus Response In- vestment Initiative (CRII), Coronavirus Response In- vestment Initiative Plus (CRII+), REACT-EU, and Co- hesion’s Action for Refugees in Europe.
The evaluation is based on a range of data sources to reach its conclusions, including monitoring sys- tems, national statistical offices, surveys, targeted interviews and public consultation, as well as case studies and focus groups.
The findings of the ESF evaluation will be published before the end of 2024.
and projects already selected by Member States, if they are completed, will absorb the amount avail- able. The common indicators reported by the end of 2022 show significant achievements, including:
•17.3 million people benefiting from the flood protection measures supported (83 % of target);
•3.4 million hectares of habitats conserved (76 % of target);
•Nearly 6 000 megawatts of renewable energy capacity created (69 % of target);
•9.1 million people given access to completed wastewater treatment systems (45 % of target);
•6.9 million people given access to an improved water supply (50 % of target); and
•257 kilometres (km) of new or improved met- ro or tram lines completed in various EU cities (47 % of target).
The final achievements (by end–2023) will only be reported in the Final reports in 2025–2026. Those reports are likely to reports achievements approaching the targets set, as the great majority of projects selected for funding are expected to be completed.
The substantial funding allocated to increasing energy-efficiency and renewable energy sources has helped further the shift towards a low-carbon and less polluting economy. In Poland, for example, heating systems using high-efficiency cogenera-
tion were modernised in 34 % of district heating systems, while in the Opolskie region low-emission transport projects have helped to expand the use of public transport, to extend the cycle path net- work and to increase the attraction of walking and cycling in urban areas.
At the same time, support for investment in envi- ronmental infrastructure in Hungary, for instance, has helped reduce the number of water supply are- as not complying with the Drinking Water Directive to only 4 % of the total and led to a substantial ex- pansion of wastewater treatment. In the Auvergne and Rhône-Alpes regions in France, ERDF-financed investment has helped to improve energy-effi- ciency in public buildings and social housing, so reducing greenhouse gas emissions, while under the Czechia-Poland Interreg programme joint risk management measures have increased the ca- pacity of the authorities concerned to tackle crises and emergency situations.
2.3Policy Objective: More connected Europe
The Connected Europe Policy Objective contributes to a more connected Europe by enhancing mobil- ity, in particular on the Transport Trans European Network.
Nearly EUR 63 billion from the ERDF and CF was allocated to the Connected Europe Objective in 2014–2020 to improve rail and road networks and other strategic transport and energy infra- structure. This represents 16 % of total Cohe- sion Policy funding for the period. By the end of 2022, projects selected suggest that an estimated EUR 57.4 billion, 91 % of the total allocated, was spent on the pursuit of this Objective. The invest- ment was mainly in the less developed Member States (those receiving support from the CF) and in less developed and transition regions elsewhere.
According to the common indicator, the achieve- ments by the end of 2022 include:
•3 560 km of new roads being constructed by the end of 2020 (99 % of target), mostly on the TEN-T network, with another 8 400 km of road being renovated (76 % of target); and
•2 100 km of rail being reconstructed (47 % of target) again mostly on the TEN-T network.
As regards the latter, while the funding set aside for selected projects suggests that the target for the rail might be achieved, these are complex projects which often experience some difficulty in being completed within the set deadline.
Support under Cohesion Policy in the 2014–2020 period, as in earlier years, has led to tangible im- provements in transport links both between coun- tries and within them. In Warmińsko-Mazurskie in Poland, for example, co-financed investment has had a significant impact on increasing the ease of movement in the region. It has led to improve- ments in road safety and reductions in CO2 emis- sions through facilitating the use of railways and public transport.
In Czechia, projects have helped to save an esti- mated 1 hour 25 minutes on average per person in travel time a year in the five urban agglomera- tions. They have also helped to increase the num- ber of passengers using public transport and their safety. Similarly, in Bulgaria, connectivity to the TEN-T has been improved significantly, while trav- el time has been reduced at the same time as the adverse effects of transport on the environment have been mitigated.
2.4Policy Objective: Social Europe
The Social Inclusion Policy Objective contributes to a more social and inclusive Europe implementing the European Pillar of Social Rights.
Cohesion Policy funding of nearly EUR 115 billion, mainly from the ESF and YEI but also from the ERDF (for infrastructure and equipment), was al- located to the ‘Social Europe’ Objective targeting support for employment and labour market inte- gration, education and training, and social inclu- sion. Funding represents 28 % of the overall Cohe- sion Policy budget for 2014–2020. By the end of 2020, estimated expenditure was around 87 % of the amount available.
The common indicators covering all EU Member States in respect of the ESF (including the YEI in the 20 Member States where it is applied) show that up to the end of 2022:
•there had been 64.5 million participants in the measures supported, including nearly 22.2 mil- lion who were unemployed and nearly 25 mil- lion who were inactive (in the sense of not ac- tively seeking employment);
•7.4 million participants in EU-funded schemes had found a job and 10.2 million had obtained a qualification;
•up to 2 030 000 firms had been supported un- der the ESF; and
•46 % of participants had a low level of educa- tion (only up to compulsory schooling or less), and 14 % were migrants, had a foreign back- ground, or were from ethnic minorities.
ERDF common indicators on support for invest- ment in social infrastructure, which was mainly in less developed and transition regions in eastern and southern Member States, show that:
•63 million people had benefited from improved health service facilities (72 % of target) up to the end of 2022; and
•nearly 24.6 million children and young people had benefited from the childcare facilities and education infrastructure that had been built (132 % of target).
The ESF and ERDF combined over the period to support social inclusion across the EU, the former through funding measures to increase employa- bility and for job-search, education at all levels, healthcare, long-term care and community ser- vices of various kinds, and the ERDF by financing investment in the infrastructure and equipment involved. In Portugal, for example, measures un- der the YEI increased the probability of being in employment three years after participation by up to a third depending on the measure, while in Lazio, the ‘Torno subito’ work experience scheme raised the probability by 11 percentage points (pp) 18 months afterwards. In Slovakia, the employ- ment rate of people with disabilities was increased by 20 pp by subsidies to employers to take them on, while in Marche, traineeships for disadvan- taged people helped to increase their employment rate six months later by 6–8 pp more than those not receiving training.
In Poland, ESF support helped to improve the qual- ity of medical training; in Portugal, to increase the standard of vocational education; and in Slovakia, to reduce early school-leaving among the Roma community.
The results of an updated3 meta-analysis4 of the available ESF and YEI counterfactual impact eval- uations carried out in the 27 Member States and the UK showed that participants in ESF/YEI meas- ures had, on average over the 2014–2020 period,
wards than comparable non-participants, amount- ing to 6–8 pp (depending on the method used).
2.5Policy Objective: a Europe closer to citizens
The Europe Closer to the Citizen Policy Objectives contributes to bring Europe closer to citizens by fostering the sustainable and integrated develop- ment of all types of territories and local initiatives.
Unlike the other 2021–2027 Policy Objectives, ‘a Europe closer to citizens’ has no direct equiva- lent under the Thematic Objective categorisation used for 2014–2020. Nevertheless, it is evident that this Policy Objective includes investments in community-led local development (CLLD), support for ITI and other territorial measures relating to urban regeneration, which were funded under mul- tiple Thematic Objectives in 2014–2020. Support of EUR 32 billion from the ERDF, ESF and CF was allocated for integrated approaches to local and territorial development for the period, around 8 % of the overall Cohesion Policy budget. At the end of 2022, expenditure under the projects selected for funding was around 65 % of the amount al- located. The level of expenditure relative to the amount allocated is lower than for the other Policy Objectives, reflecting the fact that much of the in- vestment involved mobilisation of local communi- ties and/or the formulation of development plans involving different sectors or aspects, which tend to need more time to be carried out.
The common indicators show that achievements by end–2022 include:
•27.75 million people benefiting from integrated urban strategies (71 % of target);
•20 million square metres of open space being created or rehabilitated through the investment undertaken (63 % of target); and
•1.7 million square metres of buildings being constructed or renovated in urban areas (78 %
The final achievements by the end of 2023 are expected to be close to the targets, given the large number of projects selected for funding that are likely to be completed.
Cohesion Policy funding for local development took the form especially of helping to redevelop degraded areas. In Puglia, for example, financing was directed to the renewal of urban infrastruc- ture, refurbishing abandoned buildings, and im- proving cultural sites. This was accompanied by strengthening public services, so increasing the quality of life for residents and attracting both businesses and people to move in and encourag- ing those already there to stay. In Toscana, urban regeneration measures in towns and small cities in the region led to the extension of green areas and of cycle paths as well as to improvements in public safety.
Support also went into CLLD and ITI to ensure both the involvement of residents in the redevel- opment of their local area and the coherence of the projects undertaken. In Středočeský, in Czechia, for example, CLLD projects took place in almost 100 smaller municipalities, leading to the renewal of local roads and infrastructure, especially school buildings. At the same time, ITI projects were used to improve public transport and road connections to reduce the isolation of rural areas farthest from large cities.
3.Response to the COVID-19 pandemic and to Russia’s war of aggression against Ukraine
In response to the COVID-19 pandemic, the EU reacted in two main phases. The initial response was to provide much needed financial support by reorienting the existing 2014–2020 programmes through the CRII and CRII+. These allowed Mem- ber States to support the healthcare response to COVID-19, provide working capital for SMEs, and assist vulnerable groups. Around EUR 23 billion
of EU funding was mobilised under CRII for these measures. The rationale for repurposing Cohesion Policy funding in this way was to avoid long-term socio-economic consequences in Member States that could exacerbate existing disparities. It was, in particular, to support more vulnerable, and more affected, regions, that had limited capacity to sup- port the economy, health services, and vulnerable workers and households.
The second phase of the Cohesion Policy response was the adoption of the NGEU recovery package, for the EU to emerge more resilient from the cri- sis and to support its digital and green transition. NGEU included the REACT-EU with funding of EUR 50.6 billion programmed through the ERDF, ESF and Fund for European Aid to the Most De- prived (FEAD)5. In parallel, the core of NGEU was the Recovery and Resilience Facility (RRF) delivered through the Recovery and Resilience Programs (RRPs) (see Box 9.4).
Member States reported using Cohesion Policy support for COVID-19-specific measures up to the end of 2022 in the following ways6:
•to purchase EUR 3.7 billion of personal protec- tive equipment;
•to procure around 12 500 ventilators;
•to procure nearly 97 million vaccination doses and to vaccinate 49 million people; and
•to provide financial and other support to over 920 000 enterprises.
According to the preliminary evaluation of the support provided by the ESF and FEAD under CRII and CRII+7, the two initiatives represented an ef- ficient way of using funding that remained to re- spond to the COVID-19 pandemic and for integrat- ing the funding into national strategies for tackling the crisis.
a higher likelihood of being in employment after-
of target).
5For more details on the use of REACT-EU see this Cohesion Open Data story:
https://cohesiondata.ec.europa.eu/stories/s/26d9-dqzy
.
6An overview of the reported outputs from COVID-19-related measures under CRII/CRII+ and REACT-EU are presented on this dashboard:
https://cohesiondata.ec.europa.eu/stories/s/c63b-b6in.
3Joint Research Centre (JRC), Competence Centre on Microeconomic Evaluation calculations.
4European Commission (2022).
7Preliminary evaluation of the support provided by ESF and FEAD under the CRII and CRII+, SWD(2023) 249 final, European Commission,
Brussels, 2023.
In the aftermath of Russia’s war of aggression against Ukraine, the EU put forward the three initi- atives for Cohesion’s Action for Refugees in Europe (CARE/CARE+ and FAST-CARE) to provide emer- gency shelter and basic social support to people fleeing the war. This resulted in the reallocation of EUR 1.7 billion and increased liquidity of EUR
13.6 billion, targeting primarily the Member States bordering Ukraine and with greatest influx of refu- gees. To support SMEs and vulnerable households affected by the high energy prices and finance short-time work schemes to keep people in jobs, the Supporting Affordable Energy Initiative (SAFE), reallocated around EUR 4 billion.
4.Institutional capacity and the role of reforms
As shown in Chapter 7, the quality of institutions, in terms of technical capacity but also transparen- cy, accountability, rule of law, and effective gov- ernance structures, is essential for the creation of a healthy business environment and for economic and social development. The quality of managing authorities, and of government more generally, has proven to be an important determinant of the performance of Cohesion Policy, in terms of the capacity to absorb the funding, the effectiveness and efficiency of the investment financed, and the impact on socio-economic development. The past two decades have seen increased scientific evidence on the effect of institutional and admin- istrative factors, particularly the quality and ca- pacity of public administration, in accounting for
asymmetries in the performance of Cohesion Poli- cy across EU regions. There is a general consensus in the literature that the ability of national, region- al and local authorities to design robust strategies, allocate resources effectively, and administer EU funding efficiently is a major contributor to the overall effectiveness of the policy8.
Both the European Commission and Member States have given increased attention to the re- form of public administration and administrative capacity-building to assist national and sub-na- tional bodies improve their management of the European Structural and Investment Funds. This has led, on one side, to the Commission imposing certain ex ante conditions on Member States for the receipt of funding, starting from the 2014– 2020 programming period. On the other side, the Commission has supported the strengthening of the administrative capacity of regional authorities in Member States through a dedicated budget.
Ex ante conditionalities were introduced in the 2014–2020 programming period. Member States were required to comply with a series of conditions in relation to regulation compliance, governance and administrative capacity before the program- ming period started, with the aim of ensuring that the investments funded were effective. These con- ditionalities were both ‘horizontal’ (relating to pub- lic procurement, State aid, anti-discrimination, gen- der equality, disability, environmental legislation and statistical systems); and thematic, setting out sector-specific conditions. These gave an incentive for Member States to implement structural chang- es and policy reforms, including those linked to relevant country-specific recommendations (CSRs) made as part of the European Semester process.
Ex ante conditionalities were also aimed at im- proving the targeting of public investment through better and more strategic policy frameworks, prior- itisation of projects, and ensuring complementarity with other sources of funding. They were, in addi- tion, expected to contribute to improving the insti- tutional and administrative capacity of public insti- tutions and to stimulate co-ordination within public administrations and with relevant stakeholders.
8Bachtler et al. (2016).
In case of the non-fulfilment of ex ante condition- alities, Member States were required to include in their programmes and partnership agreements ac- tion plans setting out how they intended to fulfil them. The evidence is that the majority of these plans were put in place to meet general condi- tions in respect of public procurement and com- pliance with State aid regulations. As regards pub- lic procurement, the fulfilment of conditionalities entailed:
•adoption of national strategies and the estab- lishment of legislation in several Member States (including Bulgaria, Hungary, Italy, Romania and Slovakia);
•establishment of an adequate control system (as in Bulgaria and Romania);
•introduction of e-procurement
(e.g.in
Hungary, Italy and Latvia);
•simplification of procedures and increased effi- ciency (e.g. in Italy and Slovenia);
•creation of a specific advisory unit and consul- tation groups for identifying key issues and pro- posing improvements (e.g. in Slovenia);
•development of guidelines (e.g. Romania, Italy and Slovenia); and
•training and capacity-building (as in Bulgaria, Greece, Croatia, Hungary, Italy, Malta, Romania, Slovenia and Slovakia).
Romania developed a comprehensive action plan, while six Member States reported action plans on State aid. These included the adoption of legisla- tion, the setting-up of a central State aid electronic register and database, the publication of a list of aid recipients on the website, and the implementa- tion of dedicated training programmes.
As regards thematic ex ante conditionalities, sev- eral Member States designed and implemented action plans in respect of smart specialisation, digitisation and digitalisation, energy, healthcare,
education and institutional capacity. Many of the plans adopted involved both national and regional authorities and, though in varying degrees of de- tail, the evidence shows that in many cases they were instrumental in improving the effectiveness and efficiency of programmes.
For some environmental areas such as air qual- ity, ex ante conditionalities were not desirable or possible. However, in cases where air pollution ex- ceeded EU limits, it proved useful to have concrete references to air quality plans, which were man- datory in such situations, in the text of partnership agreements and Operational Programmes.
ming period. There are fewer enabling conditions than ex ante conditionalities, and they benefit from a simplified procedure for reporting on their fulfil- ment. Unlike in the case of ex ante conditionalities, the regulation sets the fulfilment of enabling con- ditions as a prerequisite for the disbursement of funds: if enabling conditions are not fulfilled at the time of submission of a payment application to the Commission for the specific objective concerned, the related expenditure will not be reimbursed from the Union budget until the Commission as- sesses those enabling conditions as fulfilled. Ena- bling conditions have to remain fulfilled during the whole programming period.
Figure 9.2 Planned, decided and spent amounts by field of intervention (EUR billion)
Spent
Decided
Planned
122- Evaluation and studies
123- Information and communication
096 - Institutional capacity of public administrations (ERDF)
121 - Preparation, implementation, monitoring and inspection
In addition, ex ante conditionality required partner- ship agreements to address the CSRs relevant to Cohesion Policy made by the Council as part of the European Semester.
In the case of the horizontal enabling conditions in cross-cutting areas, all Member States have ful- filled those relating to public procurement, State aid, and the UN Convention on the Rights of Per- sons with Disabilities; all but one, have fulfilled the
Source: DG REGIO based on Cohesion Open Data.
0
2
4
6
8
10
12
Overall, the introduction of ex ante conditional- ity has improved the investment environment in the EU and the targeting of EU and other public funding. It has also accelerated the transposition and implementation of EU legislation and helped speed up reforms, reinforcing the commitment of governments to them and raising political aware- ness about them. In addition, by requiring public authorities to formulate development strategies, it has improved institutional capacity across the EU.
The 2021–2027 programming period has seen the introduction of enabling conditions under which investments are supported by Cohesion Policy fund- ing. As in the case of ex ante conditionalities, they are either horizontal (e.g. compliance with the EU Charter of Fundamental Rights, public procurement and State aid rules) or thematic (e.g. governance of smart specialisation strategies to build local in- novation ecosystems, compliance with 2020 binding national renewable energy targets, the planning of investments in environmental and transport infra- structure, the establishment of strategic policy frameworks for active labour market measures in the light of the employment guidelines, and for social inclusion, poverty reduction, and Roma inclusion). They are rules establishing preconditions for funding, which have to be complied with throughout the program-
condition on the Charter of Fundamental Rights.
As regards thematic conditions, i.e. those linked to specific Thematic Objectives and investment prior- ities, such as the existence of appropriate strate- gies/plans/frameworks in the policy areas covered by Cohesion Policy9, two thirds were fulfilled at the time of adoption of programmes and 90 % were fulfilled as of first of March 2024.
In addition to establishing conditions for funding, financing under Cohesion Policy has also gone to strengthening the administrative capacity to imple- ment the policy. This has entailed making availa- ble to Member States a set of tools for building administrative capacity, such as guidance on how to develop roadmaps for this, a means for peer exchange, communities of good practice, and ac- tivities (including training) focused on key strategic issues, such as public procurement, State aid, Integ- rity Pacts, and prevention of fraud and corruption.
In the 2014–2020 programming period, support for administrative capacity was used by Member States on activities for strategic capacity-building, scaling up existing practices, introducing innova- tions, and improving management of human re-
sources. Overall, over EUR 13.5 billion of EU fund- ing was allocated to such activities (Figure 9.2, which distinguishes between planned, decided and already spent amounts)10.
Preliminary evidence from administrative capaci- ty-building activities carried out in the 2014–2020 period shows that ERDF-financed investments have had a positive impact on public authorities, beneficiaries and stakeholders. Pilot case studies carried out in Romania, Greece, Spain and Italy provide a first indication of the effectiveness of these investments. In Romania, a digital regis- ter of properties and land was created to facili- tate interaction between property owners and the authorities. In Spain, the governance of ERDF-fi- nanced projects in specific areas was digitalised. In Greece the emphasis has been on administra- tive and organisational reform, e-government and public sector management, while in Italy there is a commitment to bridging the digital divide and optimising administrative procedures using ERDF financing for digitalising governance.
The ESF provided support under the institution- al capacity-building objective (TO11) for some
840 000 participants for lifelong learning and training and 3 000 projects targeting national, re- gional or local authorities or public services. For example, with ESF support, the National Customs Agency in Bulgaria implemented a series of pro- jects to simplify and rationalise legislative proce- dures and improve the efficiency of customs oper- ations, including by establishing a fully electronic working environment.
The ex post evaluation now underway will shed fur- ther light on how Cohesion Policy funding contrib- uted to the implementation of reforms in Member States and on whether programme strategies, ex ante conditionalities and horizontal principles have led, directly or indirectly, to CSRs being taken up.
5.Cohesion Policy funding 2021–2027
Cohesion Policy funding for the 2021–2027 pe- riod amounts to a third of the EU’s long-term budget under the Multiannual Financial Frame- work. The EUR 378 billion11 of support is expected to result in EUR 542 billion of investment once na- tional (public and private) co-financing is included.
9These include smart specialisation, broadband, energy-efficiency, responding to climate change, prevention and alleviation of risks and disas- ters, water supply and wastewater treatment, waste management, transport, labour market policies, education, social inclusion, alleviation of poverty, support for Roma and other minorities, and improving health and social services.
10Based on data from the system for fund management in the EU at 31 December 2022 for the following fields of intervention: ‘institutional capacity of public administrations and public services related to implementation of the ERDF or actions supporting ESF institutional capacity initiatives’; ‘preparation, implementation, monitoring and inspection’; ‘evaluation and studies’; and ‘information and communication’.
112021–2027 figures cover shared management, including Interreg programming, and funds managed directly and indirectly by the Commission.
Table 9.1 EU Cohesion Policy allocations under shared management by Policy Objective (2021–2027)
Goal / Policy objective
EU planned amount
Total planned amount
% of total EU planned
PO1 Smarter Europe
73 830
114 692
19.6 %
PO2 Greener Europe
93 356
128 930
24.8 %
PO3 More connected Europe
40 474
53 504
10.8 %
PO4 Social Europe
112 351
167 079
29.9 %
PO5 Europe closer to citizens
19 554
26 907
5.2 %
Just Transition Fund specific objective
18 049
25 363
4.8 %
Technical assistance
9 267
13 436
2.5 %
Goal: Investment in jobs and growth
366 882
529 911
97.6 %
Goal: Territorial co-operation (Interreg)
9 041
12 032
2.4 %
ducing regional disparities. The rationale for policy intervention is to provide more direct development support to those areas that need it the most but have less capacity to fund the investment required themselves. Some support is also provided to re- gions with higher level of GDP. Importantly, national co-financing is required for all types of regions, al- though at much lower rates for less developed ones.
Aid intensity (i.e. the amount of support per inhab- itant per year) is a useful indicator to show how
Cohesion Policy funding provides more support to less developed regions, in line with aim of the pol- icy to reduce regional disparities. The direct alloca- tion of funding, however, does not fully reflect the overall impact of the policy. To grasp the benefits it brings fully, the allocation of funding needs to be considered in conjunction with taking account of the effects of interventions on the EU econo- mies, including not only the local and immediate impact of programmes but also the many spill- over effects that they generate. Several studies
Note: The table covers the budget delivered through shared management programming and excludes initiatives managed directly and indirectly by the Commission.
Source: DG REGIO calculations based on shared management programmes adopted and Cohesion Open Data.
Table 9.2 Cohesion Policy aid intensity, GDP per head, and Cohesion Policy funding, in Member States, average 2014–2020
The less developed regions are the main benefi- ciaries, 70 % of the ERDF and ESF+ being allocat- ed to them. In addition, the CF provides support to 15 Member States12, and is targeted at investment in environmental infrastructure and trans-Euro- pean networks. Moreover, a new facility, the Just Transition Fund, has been set up to address the impact of the transition towards climate neutrality.
These funds are invested in the pursuit of two high-level Cohesion Policy goals, jobs and growth (national and regional programming) and European territorial co-operation (Interreg). These two goals, as indicated above, are pursued, in turn, predominantly through the five Policy Objectives, indicated earlier, which are aimed at creating a more competitive, smarter, greener, more connected, and more social and inclusive Europe, closer to citizens (Table 9.1)13.
6.Cohesion Policy as a placed-based policy
Cohesion Policy is the main EU instrument for sup- porting regional development. The policy follows a place-based approach to pursuing EU-wide overar- ching policy priorities. Such an approach is essen- tial for tailoring policy interventions to local char- acteristics, preferences and circumstances, which
tend to differ very significantly across space and time within the EU and Member States, as high- lighted in previous chapters.
A first indication of the place-based nature of the policy is reflected in the way funding under Cohe- sion Policy is allocated14, which is based on catego- rising regions in terms of their level of development, as indicated by their GDP per head. The ‘less devel- oped’ category includes regions with GDP per head below 75 % of the EU average (PPS); the ’transition’ category includes those with GDP per head between 75 % and 90 % of the EU average for the 2014– 2020 period and of between 75 % and 100 % for the 2021–2027 period; and the ‘more developed’ category includes all the other regions. Several ad- ditional indicators are then used to fine-tune the allocation according to the situation of individual regions, specifically, to reflect socio-economic, en- vironmental, and demographic challenges – overall unemployment, youth unemployment, low levels of education, greenhouse gas emissions, and outward migration. The allocation for each Member State is the sum of allocations for its eligible regions.
As indicated above, most funding under Cohesion Policy goes to the less developed regions and Mem- ber States, in line with the policy’s mandate of re-
12The CF is available to those Member States with gross national income per head below 90 % of the EU average. The 15 Member States eligible in 2021–2027 are Bulgaria, Czechia, Cyprus, Estonia, Greece, Croatia, Hungary, Lithuania, Latvia, Malta, Poland, Portugal, Romania, Slovenia and Slovakia.
13For a more complete summary of the Objectives and contents of the programmes adopted, see European Commission (2023).
14Regulation 2021/1060 (Annex XXVI) of the European Parliament and of the Council.
United Kingdom
25.90
31 347.50
0.07 %
*Average 2014–2020, except for the EU-28 and UK for which the figures correspond to average 2014–2019. Note: Aid intensity is defined as the amount of funding per inhabitant per year.
Source: Eurostat, DG REGIO.
emphasises15 that the programmes implemented in the main beneficiary regions also benefit more
Figure 9.3 Aid intensity in categories of regions, 2014–2020
Box 9.5 Research into the regional impact of Cohesion Policy
developed regions. Indeed, for some of them, these indirect spill-over effects can be larger than the di- rect effects of funding, in large part because of the goods and services that more developed regions export to less developed ones. These effects are examined in detail in Section 8 below.
Table 9.2 shows the aid intensity (funding per head) implied by the investments financed by the ERDF, ESF and CF for the 2014–2020 period, the average level of GDP per head over the period and Cohesion Policy funding in relation to GDP.
350
Average EUR per inhabitant per year
300
250
200
150
100
50
0
Less developed
Transition
More developed
A 2013 study1 used a regression discontinuity design on a dataset covering the 1994–2006 period to find a substantial positive impact of Cohesion Policy on regional economic growth. Two other studies2 also used a regression discontinuity approach to test for the impact of Cohesion Policy on Objective 1 regions (i.e. the least developed ones, receiving the most support) using a dataset including programmes from 1989 to 2013. They find a positive effect on GDP growth, every 1 EUR spent on Objective 1 trans- fers leading to EUR 1.20 of additional GDP.
A 2020 study3 used a spatial regression discontinui- ty approach on a database covering the 2000–2013
for instance, the meta-analysis of the ESF counter- factual impact evaluations carried out by Member States)5.
Model simulations constitute another strand of research to assess the impact of Cohesion Poli- cy. While this used to be conducted mostly at the national level6, sub-national models have become more developed in recent years. For instance, a 2017 study7 found a positive effect of smart spe- cialisation strategies on regions, though the extent differed between them. A 2020 study8 applied a dy- namic spatial computable general equilibrium mod- el to NUTS 2 regions in Poland, Estonia, Lithuania
As is evident, aid intensity is highest in the less developed Member States, amounting to EUR 404 per inhabitant per year in Estonia and EUR 381 in Slovakia. Funding represents a substantial injec- tion into all the less developed economies, reach- ing 2.7 % of GDP in Croatia, 2.6 % in Hungary, and
2.4 % in Poland, Slovakia and Lithuania.
Reflecting its mandate to reduce the extent of re- gional disparities across the EU, support, as noted above, goes predominantly to the regions with the greatest development needs and smallest financial means for meeting these. Aid intensity, therefore,
Source: Eurostat and DG REGIO.
averaged EUR 297 per inhabitant per year over the 2014–2020 period in the less developed regions, much more than the EUR 127 in the transition re- gions and well over 5 times more than the EUR 55 in more developed ones (Figure 9.3).
In general, there is a clear inverse relationship between aid intensity at regional level and GDP per head, reflecting the relative concentration of funding on the less developed regions (Figure 9.4).
period to find that Cohesion Policy has a positive impact on growth, though the scale varies across re- gions. A 2019 study4 found a positive effect of the policy in about 40 % of Objective 1 regions, depend- ing on their human capital endowment and quality of institutions.
For the evaluation of the 2007–2013 period, the Commission also relied on these kinds of approach, with counterfactual analysis based on propensity score matching (PSM), which attempts to match re- gions receiving support with those not receiving it in terms of their relevant characteristics, and a regres- sion discontinuity design. These pieces of analysis also point to a positive and statistically significant impact of EU funding on the growth of the regions supported. For instance, the analysis using PSM esti-
and Latvia and found that Cohesion Policy invest- ments have resulted in substantial welfare gains. The JRC of the Commission, in collaboration with DG REGIO, has developed the ‘RHOMOLO’ model, which is regularly used to assess the impact of Cohesion Policy9 and to address more specific issues such as the international spill-over effects of the policy10.
In general, model-based simulations indicate a size- able and long-lasting impact of the policy on the performance of EU regions, particularly on the main beneficiaries. However, this rests on a number of assumptions, some of which can legitimately be considered as optimistic. For instance, it is generally assumed that funding is spent efficiently on all pro- jects, which clearly is not necessarily the case. Model simulations, therefore, should be taken as estimates
Figure 9.4 Aid intensity in relation to GDP per head, NUTS 2 regions, averages 2014–2020
1 000
900
EUR per inhabitant per year
800
700
600
500
400
300
mates that funding raised the growth rate of the re- gions supported by 0.5 to 0.7 pp on average. Coun- terfactual impact evaluations have also been used by Member States to analyse their programmes (see
1Pellegrini et al. (2013).
2Becker et al. (2013, 2018).
3Crescenzi and Giua (2020).
4Di Caro and Fratesi (2019).
5European Commission (2022).
more of the potential impact of the policy than of the actual impact, and interpreted in close conjunc- tion with counterfactual impact evaluations and empirical estimates of macro-economic multipliers.
200
100
0
0
10 000
20 000
30 000
40 000
50 000
60 000
70 000
80 000
90 000
GDP per head, EUR at PPS
6See for instance: Bradley et al. (2003); Bayar (2007); Allard et al. (2008); Varga and in ’t Veld (2011a and 2011b); or Monfort et al. (2017).
7Varga (2017).
8Korzhenevych and Bröcker (2020).
9See for instance Di Comite et al. (2018) or Crucitti et al. (2023b).
10Crucitti et al. (2023a).
Source: Eurostat and DG REGIO.
15See for instance Crucitti et al. (2023).
Aid intensity is particularly high in less developed regions located in Member States with low GDP per head. Accordingly, it is highest in eastern and southern Europe, where it reaches levels above
€400 per inhabitant per year in most regions of Slovakia, Hungary and Estonia. It is also high- er in outermost regions that benefit from a top- up linked to their specificities. It is much lower in north-west Europe.
7.Place-based policies and economic performance
This section reviews the latest empirical econom- ic literature on the impact of Cohesion Policy on EU regions, bringing together studies using a va- riety of methods and with different geographical and temporal coverage, to provide an overall view of the issue, the availability of larger, and more reliable, complete and detailed data-sets (part- ly as a result of stricter performance monitoring requirements introduced in the 2007–2013 and 2014–2020 programming periods), together with progress made in analytical methods, has led to improvements in the way the effectiveness of the policy is assessed. In particular, there has been a more thorough application of econometric tech- niques to micro-level data and more sophisticated approaches to identifying the counterfactual situ- ation, i.e. what would have happened without Co- hesion Policy-financed investment16.
In methodological terms, these studies have moved largely away from trying to assess the impact of Cohesion Policy on growth at the macro-economic level, at which it is especially difficult to isolate the effect of the policy from the many other fac- tors that can affect outcomes, to focus on the micro-level impact of funding. By and large, this strand of research tends to find that Cohesion Pol- icy has a positive impact on beneficiary regions and, through spill-over effects, on Member States in general (see Box 9.5).
Simulations of macro-economic models are an- other means of investigating the effects of Cohe- sion Policy and, in recent years, regional versions of these have been developed. These have shown positive effects of smart specialisation strate- gies on regions and of EU-funded investment on welfare. They have also shown that the effect is sizeable and long-lasting, especially on the less developed regions receiving the largest amount of support. It should be noted, however, that the mod- els concerned rest on a number of assumptions, not least that the investment funded is effective in achieving its immddediate objectives, which may not necessarily hold in reality.
Overall, the large majority of the research stud- ies, from the financial crisis onwards, find an over- all positive effect of Cohesion Policy on regional development17. They suggest, moreover, that the place-based focus of the policy and its redistribu- tive effect have not come at the expense of overall economic growth in the EU and that the positive impact is not confined to the less developed re- gions but has occurred in more developed ones as well.
8.The macro-economic impact of Cohesion Policy
8.1How to assess the impact of the policy
According to the Treaty establishing the European Community, the objective of Cohesion Policy is to: ‘promote economic and social progress as well as a high level of employment, and to achieve bal- anced and sustainable development’ (Article 2) and ‘… reduce the disparities between the levels of development of the different regions and the back- wardness of the least favoured regions or islands, including rural areas’ (Article 174).
Cohesion Policy is aimed at promoting conver- gence and an harmonious development, fostering sustainable growth and improving the well-being of people living in the EU. It is the EU’s main long- term instrument to achieve these objectives, with the main instruments, the ERDF, the ESF and the CF, achieving its objectives through channels such as increasing R&D, supporting companies, and public investment in education, transport, telecom- munications, or public infrastructure.
The impact of Cohesion Policy entails a combina- tion of direct and indirect effects. For instance, out- put and employment may increase in SMEs receiv- ing support. At the same time, the SMEs concerned may also increase their demand for intermediate inputs and hence boost activity in firms that are not the direct beneficiaries of the support. The policy may generate significant spatial spill-over effects and externalities outside the economies benefiting from the programmes. In particular, the increase in local demand stemming from the programmes implemented in less developed regions is likely in some degree to be met by imports from more de- veloped regions, which therefore end up indirectly benefiting, in some cases to a considerable extent.
At the same time, economic performance is affect- ed by a wide range of other developments that coincide with the investment financed under Cohe- sion Policy, including other policy action or changes in the business cycle. The specific impact of the policy can, therefore, not be identified simply by looking at the data in the national and regional ac- counts. In order to identify the impact that can be attributed to the policy, the world as it is needs to be compared with what it would have been with- out the policy, which obviously cannot be observed in reality.
Macro-economic models enable these issues to be addressed in a consistent way. Firstly, models can be used to simulate developments without the pol- icy and so provide a counterfactual base against which the impact of the policy can be assessed.
and indirect effects. Thirdly, models can account for spill-over effects and externalities and so en- able the full impact of the policy to be assessed. Fourthly, models help to trace back the effects of policy interventions and to shed light on the chan- nels through which the policy produces its impact on the economy.
Over the past few decades Cohesion Policy has been the second most important line in the EU budget, accounting for around a third of the Multiannual Financial Framework. Between 1990 and 2024, the funding allocated increased over 10-fold in relation to EU GDP, from 0.03 %, on average, for the 1989–1994 programming peri- od to 0.3 % for the 2014–2020 period, and 0.4 % if REACT-EU is included. This increase reflects the need to accompany the deepening and widening of EU integration, the strengthening of the Single Market and successive rounds of enlargement, which have meant addressing the needs of a growing number of less developed regions. For the 2014–2020 period, EUR 356 billion was allocated to Cohesion Policy (EUR 405 billion with REACT-EU) and for 2021–2027, EUR 376 billion (less than in the previous period, reflecting the exit of the UK). While, as indicated above, this funding is allocat- ed to all regions across the EU, it goes predomi- nantly to the less developed regions and Member States, in some of them representing close to 3 % of GDP. For the 2014–2020 period, Cohesion Poli- cy funding corresponded to around 13 % of public investment in the EU as a whole and to 51 % in the Member States eligible for the CF.
As Figure 9.5 shows, spending tends to be concen- trated at the end of implementation periods18, but is not discontinued between programming periods. Indeed, the objective of the policy to reduce the development gap between EU regions is a long- term one, which is maintained throughout the EU budget cycle. The overlapping of funding between programming periods means that there is no in- terruption to the support provided. Accordingly, in the analysis below programming periods are not
16More specifically, increasingly in the last decade, studies have applied techniques such as difference-in-difference or regression discontinu- ity design to quantifying the impact of Cohesion Policy, attempting, for example, to estimate the effect of the interventions by comparing similar regions just above and below the threshold for eligibility for funding see e.g. Crescenzi and Giua (2016). The studies rely in the main on identifying a counterfactual situation, in which beneficiaries of the support are compared with a control group in a quasi-experimental
Secondly, models enable both the short- and long- term effects of the policy to be simulated, taking explicit account of the interaction between direct
considered in isolation but as continuous sources of support.
framework.
17McCann (2023).
18The N+3 rule allows funds to be used up to three years after they have been committed, which implies that the programmes are actually
implemented over a period of 10 years rather than seven.
Figure 9.5 Cohesion Policy funding 1989 to 2030
Total
Period 1989–1993
Period 1994–1999
Period 2000–2006
Period 2007–2013
Period 2014–2020
Box 9.6 Model description
The model is calibrated on a set of fully integrated
en into account in the model through regional trade
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
Period 2014–2020 (REACT-EU)
Period 2021–2027 (expected)
EU regional social accounting matrices (SAMs) for all the EU NUTS 2 regions and for the year 20171, which is taken as the baseline state of the econo- my. The SAMs include all the standard information of input-output tables on the production and use of goods and services, as well as information on the secondary distribution of income, detailing the roles of labour and households.
The model economies are disaggregated into 10 sectors (based on the NACE rev. 2 industry clas- sification)2. Firms are assumed to maximise profits and produce goods and services according to a con- stant elasticity of substitution production function3.
flows and the relatively high elasticity of substitu- tion between domestic and imported goods and ser- vices4. (This is set to 4, based on empirical estimates using European data5.) The presence of significant inter-regional spill-overs is an important feature of the model. This borrows from economic geography by incorporating a notion of spatial equilibrium cor- responding to a balance between agglomeration forces (pushing economic activity to concentrate in particular places) and dispersion forces (pushing economic activity to be less concentrated).
RHOMOLO is used for scenario analysis, in the sense that shocks mimicking the effects of policies are
Note: Figures relate to EU payments except for 2021–2027, where they are planned amounts. The timing of payments for 2021–2027 is estimated from that for 2014–2020, net of REACT-EU funding.
Source: DG REGIO.
The other agents in the model are households and a government that collects taxes and spends money on public goods and transfers. Capital and labour are used as factors of production (public capital enters
introduced to disturb the initial assumed steady state calibrated with the SAMs, resulting in different values for the endogenous variables of the model, such as GDP, employment, imports and exports, and
8.2Model and results
The impact of the policy is assessed using the European Commission’s spatial computable gen- eral equilibrium model, RHOMOLO19. In this type of model, policy interventions – disbursements of funding for specific purposes – are modelled as shocks to an economic system, generating, on the
as final consummers, and governments that im- pose taxes and borrow to finance their expenditure (see Box 9.6 for a description of the model).
In the present analysis, Cohesion Policy expenditure is regrouped into six fields of intervention. In order to simulate the impact of the policy, each field of intervention is assumed to generate a set of mod-
the production function as an unpaid factor). Trade in goods and services – within and between regions – is assumed to be costly, with transport costs in- creasing with distance. The estimate of transport costs is based on a transport model (see below). Re- gional economies are typically more open than na- tional ones, due to their smaller size, and this is tak-
1Thissen et al. (2019).
prices. The model is solved in a recursively dynamic process, where a sequence of static equilibria linked to one another through the law of motion of state variables. This implies that economic agents are not forward-looking and their decisions are solely based on current and past information.
basis of a set of assumptions, responses that are reflected in changes in macro-economic variables, such as GDP, employment, investment, and house- hold consumption.
The economic foundations of the model lie in the literature on general equilibrium models20. The model itself is featured in numerous articles contributing to this literature21, and it is regularly used for policy impact assessment purposes. The model covers all EU NUTS 2 regions and divides the economies in these into 10 (NACE22) produc- tion sectors. It incorporates input-output matrices
el ‘shocks’, which are intended to capture the eco- nomic transmission mechanisms through which the expenditure concerned is most likely to have effects. Specifically, one or more model shocks are used to simulate the spending categories relating to the six fields of interventions. The shocks can be broadly separated into demand-side shocks, with temporary effects, and supply-side shocks, with more permanent structural effects on the econo- my. The shocks – i.e. the demand and supply-side effects – assumed to be associated with expendi- ture in the six fields of intervention are as follows.
2The 10 (NACE) sectors are: agriculture, forestry and fishing (A); mining and quarrying, electricity, gas, steam, and air conditioning,
water supply, sewerage, waste management and remediation activities (B, D, and E); manufacturing (C); construction (F); wholesale and retail trade, repair of motor vehicles and motorcycles, transportation and storage, accommodation and food service activities (G-I); information and communication (J); financial and insurance activities, and real estate activities (K-L); professional, scientific and technical activities, and administrative and support service activities (M-N); public administration and defence, and compulsory social security, education, human health and social work activities (O-Q); and arts, entertainment and recreation, other service activities, activities of the households as employers, undifferentiated goods- and services- producing activities of households for own use, and activities of extraterritorial organisations and bodies (R-U).
3Constant elasticity of substitution is a class of production functions frequently used in applied economics. It describes the rela- tionship between production and production factors in the technological production process. It accounts for various substitution possibilities across inputs and determines demand for the various types of factors of production.
4This elasticity specifies the degree of substitution in demand between similar products produced in different countries.
5See: Németh et al. (2011); and Olekseyuk and Schürenberg-Frhosch (2016).
to represent the flow of raw materials and goods and services between these sectors and their dis- tribution to final users. It also incorporates capital and labour as factors of production, households
•Transport infrastructure (TRNSP) – Invest- ments in transport infrastructure are assumed to generate both demand- and supply-side ef- fects. Demand-side effects are produced by the
temporary increases in government consump- tion, i.e. in the purchase of goods and services required to build the infrastructure concerned. On the supply side, the investments are as- sumed to reduce transport costs, so reducing
mates obtained from the fully fledged transport cost model23 used to assess the investments in transport infrastructure financed under Cohe- sion Policy for the 2014–2020 period.
19
https://joint-research-centre.ec.europa.eu/tedam/rhomolo-model_en.
See also: Christou et al. (2024).
20For the full mathematical description of the model, see: Lecca et al. (2018).
21See, among others: Lecca et al., 2020; and Di Pietro et al. (2021).
22Nomenclature statistique des activités économiques (statistical classification of economic activities).
the prices of goods and stimulating trade flows. The induced reduction is based on the esti-
23Persyn et al. (2022 and 2023).
•Other public infrastructure (INFR) – Investment in non-transport infrastructure, such as electric-
ity networks, water treatment plants and waste management facilities, are modelled as public investments when associated with industrial processes, and otherwise as government con- sumption. In the latter case, only temporary de- mand-side effects are produced. Public invest- ments not only trigger an increase in demand, but also have supply-side effects, since they increase the stock of public capital used to pro- duce goods and services. (The output elasticity of public capital, i.e. the goods and services it produces, is set to 0.1, in line with the existing literature24). A congestion parameter of public capital, set to 0.5 (equivalent to a medium level of congestion25) captures the fact that, to some extent, the use of public infrastructure by a user prevents other users from using it as well.
•Research and technological development (RTD) – Subsidies to R&D are modelled as in- creases in private investments as a result of a
The main assumption is that an additional year of training leads to an increase in productivity, which is set at 7 % based on the literature28. The cost of education per pupil or student is used to calculate the amount of training implied by Cohesion Policy funding going to investment in human capital, with country-specific efficiency adjustments based on PISA scores29. On the other hand, interventions aimed at promoting the socio-economic integration of marginalised communities, participation in the labour market, or the modernisation of labour market institu- tions, are assumed to generate an increase in aggregate labour supply. In this case, a higher cost per trainee is assumed, and it is further as- sumed that it takes two to three years of train- ing to integrate a worker into the labour force.
•Aid to private sector (AIS) – Aid to the private sector is modelled as an increase in private in- vestment via a reduction in the risk premium, as
the typically longer life of public infrastructure31). This implies that, in the absence of further invest- ment, the structural effects from Cohesion Policy gradually vanish and the economy is assumed eventually to return to its initial steady state32.
The model simulations take into account the fact that Cohesion Policy is financed by the pro rata con- tribution of Member States to the EU budget, which is assumed to be proportional to their share of EU GDP. Member State contributions to the funding of Cohesion Policy are assumed to be financed by a lump-sum tax that reduces household disposable in- come, so adversely affecting economic performance and partly offsetting the positive impact of the pro- grammes33. This implies that a larger share of Mem- ber State contributions to Cohesion Policy comes from the more developed parts of the EU, while the bulk of the interventions take place in the less devel-
oped parts. The next section presents the results of the analysis based on the assumed effects of the different kinds of intervention described above.
8.3Impact at EU level
The impact of the policy is estimated by comparing the results of the model under a scenario exclud- ing Cohesion Policy interventions (the ‘baseline’ scenario) with a scenario including these. The dif- ference between the two scenarios for a given variable, such as GDP, indicates the impact of the policy, which is expressed as the percentage differ- ence from the baseline34.
The results of the simulation suggest that Cohe- sion Policy interventions are likely to have a pos- itive and significant impact on the EU’s economy (Figure 9.6)35. The impact of Cohesion Policy builds
reduction in the risk premium, which increase the stock of private capital26. Moreover, these investments are assumed to increase total fac- tor productivity (TFP) according to an elasticity that depends on the importance of spending on R&D in the region relative to GDP, and which is based on the literature27.
•Human capital (HC) – Investments in human capital are assumed to increase demand via government current expenditure. They are also assumed to have two alternative supply-side ef- fects, depending on the nature of the interven- tions. The spending categories associated with human capital development, such as training to improve the skills of the workforce and simi- lar active labour market policies, are assumed to generate an increase in labour productivity.
in the case of RTD investment, but without any impact on TFP.
•Technical assistance (TA) – Technical assistance is modelled as a demand-side shock increasing public current expenditure with no supply-side effects.
It is further assumed that a fixed interest rate of 4 % applies across regions30, and that all long-run supply-side effects diminish over time. Specifical- ly, increases in labour productivity and TFP, and reductions in transport costs, are assumed to di- minish at a rate of 5 % a year. In addition, stocks of private and public capital are assumed to have a depreciation rate of 15 % and 5 %, respectively (a higher rate for private than public capital is a common assumption in the literature and reflects
Figure 9.6 Impact of Cohesion Policy programmes 2014–2020 and 2021–2027 on EU GDP, 2014–2043
1
% difference from baseline
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
0
Note that if the 2007–2013 programmes had been included in the analysis, their impact would have been visible in the initial years of the graph and the cumulative impact would have been larger. Similarly, starting from 2030, the effects of post-2027 programmes would be expected to progressively kick in.
Source: RHOMOLO simulations (GDP impact) and DG REGIO (Cohesion Policy data).
24See: Ramey (2020). Note that 0.1 is slightly below the average of 0.12 found by the meta-study by Bom and Lightart (2014).
25Alonso-Carrera et al. (2009). A value of zero would make public capital a pure public good (i.e. one for which one person’s use has no effect
on its availability to others).
26In the production function, the capital-labour elasticity of substitution is 0.4, in line with, among others: Chirinko (2008) and Leon-Ledesma et al. (2010).
27See: Kancs and Siliverstovs (2016).
28De la Fuente and Ciccone (2003); and Canton et al. (2018).
29Programme for international student assessment, which measures 15-year-old students’ reading, mathematics, and science literacy in
different countries.
30Following Smets and Wouters (2003).
31See: Bom (2017).
32Various pieces of sensitivity analysis (not reported here) have been conducted to check the robustness of the results for the values selected for some of the key parameters.
33This means that, in the model, the EU regions are not constrained to run a balanced budget and can have deficits or surpluses. The EU budget is constrained to be balanced, as the amount of spending incurred by regions that is financed from Cohesion Policy is repaid through an equal amount of lump-sum transfers from households.
34The baseline is established on the basis of assuming that observed trends in key variables continue, which is common practice in modelling exercises. The results, which correspond to the difference between the baseline and the ‘with-policy’ scenario, are largely independent of the baseline assumptions.
35The UK is excluded when reporting results because of its exit from the EU. The aggregate effects are also reported net of the UK. Including the UK in the analysis does not alter the substance of the results.
up over time, especially when the two program- ming periods overlap between 2021 and 2023. The impact is the greatest in 2030, when GDP in the EU is estimated to be 0.9 % higher as a result of the combination of the 2014–2020 and 2021–2027 interventions36. The cumulative impact of these programmes is particularly significant in less de- veloped Member States and especially in Croatia (an increase of 8 % in GDP), Poland and Slovakia (an increase of 6 %) and Lithuania (a 5 % increase).
In the short run, a substantial part of the impact stems from the increase in demand, which is as- sumed to be partly crowded out through increases in wages and prices. In the medium and long run, productivity-enhancing effects of Cohesion Policy investment as well as increases in the stock of public and private capital materialise, so boosting both current and future GDP as production capaci- ty is increased. The policy-induced increases in po- tential output leave room for increases in GDP free of inflationary pressures from 2031 onwards. The interventions therefore continue to stimulate eco- nomic activity long after the interventions come to an end, as would be expected from a policy aimed at strengthening EU regional economies.
The policy yields a positive return at EU level. The cumulative multiplier, i.e. the ratio of cumula- tive changes in GDP to the amount of expenditure, is estimated at 1.29 in 2030 and 2.97 in 2043. This means that 30 years after the start of the programmes, for each 1 EUR invested under Cohe- sion Policy, EU GDP is increased by almost EUR 3, which is equivalent to an annual rate of return of around 4 %.
These results are consistent with the literature on the impact and the effectiveness of public policies and spending. The vast majority of the studies con- cerned rely on econometrics and provide estimates of impact multipliers, i.e. the ratio of the change in GDP to a change in government spending in the periods directly following the one in which the spending takes place. Most of them, however, do not go beyond a time horizon of more than four
years, whereas model-based analysis can inves- tigate the long-term, lasting effects. Most studies, therefore, provide estimates of cumulative multi- pliers calculated at a given, relatively short, time after the policy shock, which can be considered to be a short-run estimate of the multiplier, while models can also estimate the long-run multiplier over an infinite time horizon37 (see Box 9.7 for a review of recent studies).
8.4Impact at regional level
Cohesion Policy is a place-based policy aimed at fostering convergence, with both the amount and composition of expenditure it finances differing between regions according to their characteris- tics, notably their level of development and their economic and social circumstances. As a con- sequence, the impact on GDP is heterogeneous across regions. Maps 9.1 and 9.2 show the effect of Cohesion Policy on GDP in EU regions in 2023 – the last year for which the two programming pe- riods overlap – as the percentage difference from the baseline. The impact increases over time in all regions up to 2030. In both 2023 and 2030, the largest increases occur in less developed regions, such as those in Bulgaria, Greece, Hungary, Por- tugal, Poland and Slovakia. The increase is par- ticularly large in Voreio Aigaio in Greece (12.7 % in 2030), the Portuguese Açores (12.0 %), and Swietokrzyskie (117 %) and Warminsko-Mazur- skie (103 %) in Poland. There are also significant differences between regions in the same country. For example, in Poland the increase in GDP ranges from 3.8 % to 11.7 %, and in Hungary from 2.2 % to 8.0 %.
In the more developed regions, the short-run im- pact of the Policy is smaller and more difficult to estimate38. However, in the medium to long run, the differences in the impact on GDP between re- gions diminishes and it is positive in all regions. This is partly because of the strong positive spatial spill-over effects generated by the policy, which stem mostly from the fact that the main bene- ficiaries are often small, open economies with
36The long-term cumulative impact on GDP is positive for both the EU as a whole and for all Member States.
37See, for instance: Tesfaselassie (2013); or Ilzetzki et al. (2011).
38As noted above, it is assumed that regions finance the policy proportionally to their share of EU GDP.
narrow industrial bases and limited R&D capacity. Many goods or services needed for the implemen- tation of Cohesion Policy programmes are, there- fore, not produced domestically and so need to be imported, to a large extent, from more developed regions39.
8.5Impact on regional disparities
Cohesion Policy helps to reduce regional dispari- ties significantly. The coefficient of variation, which measures the extent of regional disparities in GDP per head, is estimated to decline by around 3 % 10 years after the beginning of the 2021–2027 programming period (Figure 9.7). It increases after that as the supply-side effects of the interventions diminish. The same pattern is observed in other
measures of dispersion such as the ratio of the 80th to the 20th percentile of the distribution of regional GDP per head (the top 20 % and bottom 20 % of regions in these terms). However they are meas- ured, regional disparities are estimated to be much lower than without Cohesion Policy for many years to come even if the policy were to come to an end.
Cohesion Policy also helps to increase internal convergence and reduce regional disparities within Member States. The extent of regional disparities (again as measured by the coefficient of variation) is estimated to decline in all Member States as a result of policy interventions (Figure 9.8). In Hun- gary, it is reduced by 2.5 pp compared with a situa- tion without Cohesion Policy, and by around 2.0 pp in Portugal and Poland.
Figure 9.8 Impact of Cohesion Policy programmes 2014–2020 and 2021–2027 on the coefficient of variation, GDP per head in 2030, NUTS 2 regions
HU PT
PL
EL
BG SK RO CZ
LT
SI
IT
ES
DE BE
IE
FR
HR
AT
SE
NL DK
FI
Percentage point difference from baseline
0.0
-0.5
-1.0
-1.5
-2.0
-2.5
-3.0
Source: RHOMOLO simulations.
Figure 9.7 Impact of Cohesion Policy programmes 2014–2020 and 2021–2027 on the coefficient of variation in GDP per head in EU NUTS 2 region, 2014-2043
53.5
% of average GDP per head
53.0
52.5
52.0
51.5
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
51.0
Source: RHOMOLO simulations.
Table 9.3 Impact of Cohesion Policy programmes 2014–2020 and 2021–2027 on GDP per head in NUTS 2 regions according to the Theil index
The impact of the policy on regional disparities is confirmed by changes in the Theil index, an- other measure of dispersion, which enables be- tween-country and within-country differences to be distinguished40, which is estimated to decline by over 7 % by 2030 (Table 9.3). Both the ‘between’ and the ‘within-country’ components of the index decline, implying that disparities in GDP per head in regions within Member States are reduced (by 5.4 %), as well as disparities between Member States (by 7.9 %).
8.6Some considerations
The analysis suggests that Cohesion Policy has significant positive effects on the EU economy and those of the Member States and regions. The mag- nitude of the impact is particularly large in the less developed regions of the EU, but more developed regions also benefit from the policy, especially in the long run. This, to some extent, is explained by the strong spatial spill-over effects generated by the policy, as interventions implemented in the
benefit from Cohesion Policy investment, whether directly or indirectly.
Research suggests that investing in the less devel- oped regions tends to reduce regional disparities within countries while at the same time boosting national growth (see Box 9.8 for a review of the literature on this).
The evidence is that Cohesion Policy plays an im- portant role in reducing regional disparities in the EU in line with its mandate. It helps the less devel- oped regions to catch up with the more developed ones, while fostering aggregate growth at EU level and in all Member States.
Note: Only Member States with more than four NUTS 2 regions are included to enable the Theil index to be calculated. Source: RHOMOLO simulations.
developed ones or those with companies with a strong competitive advantage in sectors that
39See: Crucitti et al. (2023a).
40The index enables the extent of regional disparities across the EU to be decomposed into those that arise from disparities between Member States and those that arise from disparities within them.
References
Aiello, V., Reverberi, P. and Brasili, C. (2019), ‘Regional diversity in experiences of cohesion policy: The cases of Emilia–Romagna and Calabria’, Papers in Regional Science, 98(6), pp. 2275–2293.
Allard C., Choueiri N., Schadler S. and R. Van Elkan (2008), ‘Macroeconomic Effects of EU Transfers
in New Member States’, IMF Working Paper, WP/08/223.
Alonso–Carrera, J., Freire–Serén, M. J. and Manzano, B. (2009), ‘Macroeconomic effects of the regional
allocation of public Capital formation’, Regional Science and Urban Economics, 39(5), pp. 563–574.
Bachtler, J., Olejniczak, K., Smeriglio, A. and Śliwowski, P. (2016), Administrative Capacity building and EU Cohesion Policy, Outline of methodology, Inception Report to the European Parliament, EPRC.
Barbero, J., Christou, T., Crucitti, F., García Rodríguez, A., Lazarou, N. J., Monfort, P. and Salotti, S. (2024, forthcoming), ‘A spatial macroeconomic analysis of the equity–efficiency trade–off of the European cohesion policy’, Spatial Economic Analysis.
Bayar, A. (2007), Simulation of Rand D Investment Scenarios and Calibration of the Impact on a Set of Multi–Country Models, European Commission DG JRC Institute for Prospective Technological Studies.
Becker, S., Egger, P. and Von Ehrlich, M. (2013), ‘Absorptive Capacity and the growth and investment effects of regional transfers: A regression discontinuity design with heterogeneous treatment effects’, American Economic Journal: Economic Policy, 5(4).
Becker, S., Egger, P. and M. von Ehrlich (2018), ‘Effects of EU Regional Policy: 1989–2013’, Regional Science and Urban Economics, 69, pp. 143–152.
Beugelsdijk, M. and Eijffinger, S. (2005), ‘The Effectiveness of Structural Policy in the European Union:
An Empirical Analysis for the EU–15 in 1995–2001’, Journal of Common Market Studies, 43, pp. 37–51.
Bom, P. R. D. (2017), ‘Factor–biased public capital and private capital crowding out’, Journal of Macroeconomics, 52, pp. 100–117.
Bom, P. R. D. and Lightart, J. E. (2014), ‘What have we learned from three decades of research on the productivity of public Capital?’, Journal of Economic Surveys, 28(5), pp. 889–916.
Bradley J., Untiedt G. and E. Morgenroth (2003), ‘Macro–regional evaluation of the structural funds using the HERMIN modelling framework’, Scienze Regionali, 3.
Breidenbach, Ph., Mitze, T. and Schmidt, C. (2016), ‘EU Structural Funds and Regional Income Convergence – A Sobering Experience’, CEPR Discussion Papers 11210.
Brueckner, M., Pappa, E. and Valentinyi, A. (2023), ‘Geographic Cross Sectional Fiscal Spending Multipliers and the Role of Local Autonomy: Evidence from European Regions’, Journal of Money, Credit and Banking, 55(6), pp. 1357–1396.
Canova, F. and Pappa, E. (2022), The macroeconomic effects of EU Structural funds, Working Paper. Canton, E., Thum–Thysen, A. and Voigt, P. (2018), Economist’s musing on human capital investment:
How efficient is public spending on education in EU Member States?, European Economy Discussion Paper 081.
Chirinko, R. S. (2008), ‘Σ: The long and short of it’, Journal of Macroeconomics, 30(2), pp. 671–686.
Christou, T., Garcia Rodriguez, A., Heidelk, T., Lazarou, N., Monfort, P. and Salotti, S. (2024), ‘A RHOMOLO assessment of 2014–2027 Cohesion policy’, Territorial Development Insights Series, European Commission, JRC136790.
Coelho, M. (2019), ‘Fiscal Stimulus in a Monetary Union: Evidence from Eurozone Regions’,
IMF Economic Review, 67, pp. 573–617.
Crescenzi, R. and Giua, M. (2016), ‘Different approaches to the analysis of Cohesion Policy: leveraging complementarities for evidence–based policy learning’. In: Bachtler, J. et al. (eds.), EU Cohesion Policy: Reassessing Performance and Direction. Regions and Cities, Routledge, pp. 21–32.
Crescenzi, R. and Giua, M. (2020), ‘One or many Cohesion Policies of the European Union? On the
differential economic impacts of Cohesion Policy across Member States’, Regional Studies, 54(1),
pp. 10–20.
Crucitti, F., Lazarou, N. J., Monfort, P. and Salotti, S. (2021), A scenario analysis of the 2021–2027 European cohesion policy in Bulgaria and its regions (No. 06/2021), JRC Working Papers on Territorial Modelling and Analysis, European Commission, Seville, JRC126268.
Crucitti, F., Lazarou, N. J., Monfort, P. and Salotti, S. (2022), A cohesion policy analysis for Romania towards the 2021–2027 programming period (No. 06/2022), JRC Working Papers on Territorial Modelling and Analysis, European Commission, Seville, JRC129049.
Crucitti, F., Lazarou, N., Monfort, P. and Salotti, S. (2023a), Where does the EU cohesion policy produce its benefits? A model analysis of the international spillovers generated by the policy, Economic Systems, 47(3).
Crucitti, F., Lazarou, N., Monfort, P. and Salotti, S. (2023b), ‘The impact of the 2014–2020 European Structural Funds on territorial cohesion’, Regional Studies, Special issue: EU Cohesion Policy towards Territorial Cohesion?, pp. 2–15.
Dall’erba, S. (2005), ‘Distribution of Income and Regional Funds in Europe 1989– 1999: An Exploratory Spatial Data Analysis’, Annals of Regional Science, 39, pp. 121–148.
Dall’erba, S. and J. Le Gallo (2008), ‘Regional convergence and the impact of European structural funds 1989–1999: A spatial econometric analysis’, Papers in Regional Science, 82(2), pp. 219–244.
De la Fuente, A. and Ciccone, A. (2003), Human capital in a global and knowledge–based economy, 918,
Office for Official Publications of the European Communities.
Destefanis, S. and Di Giacinto, V. (2023), EU structural funds and GDP per capita: spatial VAR evidence for the European regions, Temi di discussione 1409, Bank of Italy, Economic Research and International Relations Area.
Di Caro P. and Fratesi, U. (2022), ‘One policy, different effects: Estimating the region–specific impacts
of EU cohesion policy’, Journal of Regional Science, 62, pp. 307–330.
Di Comite, F., Lecca, P., Monfort, P., Persyn, D. and Piculescu, V. (2018), The impact of Cohesion Policy 2007–2015 in EU regions: Simulations with the RHOMOLO interregional dynamic general equilibrium model, JRC Working Papers series on Territorial Modelling and Analysis, 03/2018, JRC114044.
Di Pietro, F., Lecca, P. and Salotti, S. (2021), ‘Regional economic resilience in the European Union: a numerical general equilibrium analysis’, Spatial Economic Analysis 16(3), pp. 287–312.
Duque Gabriel, R., Klein, M. and Pesso, A. (2020), The Effects of Government Spending in the Eurozone.
Working Paper Series 400, Sveriges Riksbank.
Durand, L. and Espinoza, R. (2021), The Fiscal Multiplier of European Structural Investment Funds:
Aggregate and Sectoral Effects with an Application to Slovenia, IMF Working Paper WP/21/118.
Ederveen, S., de Groot H. and R. Nahuis (2006), ‘Fertile Soil for Structural Funds? A Panel Data Analysis
of the Conditional Effectiveness of European Cohesion Policy’, Kyklos, 59(1), pp. 17–42.
European Commission (2022), Meta–analysis of the ESF counterfactual impact evaluations, Final Report, Publications Office of the European Union, Luxembourg.
European Commission (2023), Cohesion 2021–2027: forging an ever stronger Union, Report on the outcome of 2021–2027 Cohesion Policy programming, SWD(134) final.
European Commission (2024), European Structural and Investment Funds 2023: Summary report of the annual programme implementation reports covering implementation in 2014–2020, COM(2024) 6 of 15 January 2024.
European Union (2021), Regulation (EU) 2021/241 of the European Parliament and of the Council of 12 February 2021 establishing the Recovery and Resilience Facility.
Ilzetzki, E., Mendoza, E. and Vegh, C. (2011), How Big (Small?) are Fiscal Multipliers?, IMF Working Paper, 11/52.
Kancs, D. and Siliverstovs, B. (2016), ‘Rand D and non–linear productivity growth’, Research Policy, 45(3), pp. 634–646.
Korzhenevych, A. and Bröcker, J. (2020), ‘Investment subsidies and regional welfare: a dynamic
framework and its application to the European regional policy’, Regional Studies, 54(9), pp. 1262–1274.
Lecca, P., Barbero, J., Christensen, M.A., Conte, A., Di Comite, F., Diaz–Lanchas, J., Diukanova, O., Mandras, G., Persyn, D. and Sakkas, S. (2018), RHOMOLO V3: A spatial modelling framework. JRC Technical Reports JRC111861, Publications Office of the European Union, Luxembourg.
Lecca, P., Christensen, M., Conte, A., Mandras, G. and Salotti, S. (2020), ‘Upward pressure on wages and
the interregional trade spillover effects under demand–side shocks’, Papers in Regional Science 99(1),
pp. 165–182.
Leon–Ledesma, M. A., McAdam, P. and Willman, A. (2010), ‘Identifying the elasticity of substitution with biased technical change’, American Economic Review, 100(4), pp. 1330–1357.
McCann, Ph. (2023), How Have Place–Based Policies Evolved to Date and What Are They For Now?,
Paper presented at OECD–EC High–Level Workshop Series: Place–Based Policies for the Future.
Monfort, P., Piculescu, V., Rillaers, A., Stryczynski, K. and and Varga, J. (2017), The Impact of Cohesion Policy and Rural Development Policies 2007–2013: Model Simulations with QUEST III, DG REGIO Working Papers Series, Working Paper 05/201.
Németh, G., Szabó, L. and Ciscar, J. C. (2011), ‘Estimation of Armington elasticities in a CGE economy– energy–environment model for Europe’, Economic Modelling, 28(4), pp. 1993–1999.
Olekseyuk, Z. and Schürenberg–Frosch, H. (2016), ‘Are Armington elasticities different across countries
and sectors? A European study’, Economic Modelling, 55, pp. 328–342.
Pellegrini, G., Terribile, F., Tarola, O., Muccigrosso, T. and Busillo, F. (2013), ‘Measuring the effects of European Regional Policy on economic growth: A regression discontinuity approach’, Papers in Regional Science, 92, pp. 217–233.
Persyn, D., Barbero, J., Díaz–Lanchas, J., Lecca, P., Mandras, G. and Salotti, S. (2023), ‘The ripple effects
of large–scale transport infrastructure investment’, Journal of Regional Science, 63(4), pp. 755–792.
Persyn, D., Díaz–Lanchas, J. and Barbero, J. (2022), ‘Estimating distance and road transport costs between and within European Union regions’, Transport Policy, 124, pp. 33–42.
Ramey, V.A. (2020), The macroeconomic consequences of infrastructure investment, National Bureau of Economic Research, w27625.
Rodríguez–Pose, A. and Garcilazo, E. (2015), ‘Quality of government and the returns of investment: Examining the impact of cohesion expenditure in European regions’, Regional Studies, 49(8), pp. 1274– 1290.
Smets, F. and Wouters, R. (2003), ‘An estimated dynamic stochastic general equilibrium model of the Euro area’, Journal of the European Economic Association, 1(5), pp. 1123–1175.
Tesfaselassie, M. (2013), ‘Trend productivity growth and the government spending multiplier’, Journal of Macroeconomics, 37, pp. 197–207.
Thissen, M., Ivanova, O., Mandras, G. and Husby, T. (2019), European NUTS 2 regions: construction of interregional trade–linked Supply and Use tables with consistent transport flows, JRC Working Papers on Territorial Modelling and Analysis No. 01/2019, European Commission, Seville, JRC115439.
Varga, A. (2017),’Place–based, spatially blind, or both? Challenges in estimating the impacts of modern development policies: The case of the GMR policy impact modelling approach’, International Regional Science Review, 40, pp. 12–37.
Varga, J. and in ’t Veld, J. (2011a), ‘Cohesion Policy spending in the New Member States of the EU in an endogenous growth model’, Eastern European Economics, 49(5), pp. 29–54.
Varga, J. and in ’t Veld, J. (2011b), ‘A Model–based Analysis of the Impact of Cohesion Policy Expenditure 2000–06: Simulations with the QUEST III endogenous Rand D model’, Economic Modelling, 28, pp. 647–663.