EUROPEAN COMMISSION

Brussels, 4.2.2022

SWD(2022) 24 final

COMMISSION STAFF WORKING DOCUMENT

Cohesion in Europe towards 2050

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 8th Cohesion Report: Cohesion in Europe towards 2050

{COM(2022) 34 final}

CHAPTER 1 The regional dimension of the COVID-19 pandemic

·The outbreak of the COVID-19 pandemic has led to at least 872,000 more deaths in the EU compared to previous years. Excess mortality was higher in less developed regions than in transition and more developed ones. While the first wave affected primarily north-western regions and southern regions, the following waves led to the highest mortality in eastern regions.

·The restrictions put in place to contain the pandemic led to the deepest post-1945 recession. The impact was largest on southern regions, especially those dependent on tourism, where the reduction in hours worked and GDP were the most severe.

·The travel restrictions not only affected the tourism sector, but also border areas where people could no longer cross a national border to go to work or to access services.

·Thanks to job retention schemes the impact of employment and unemployment was much smaller compared to the reduction in hours worked and GDP. This allowed the EU to avoid a big spike in unemployment.

·The number of people usually working from home doubled. This increase was highest in many of the capital regions. These regions typically have a more developed service economy, host jobs that can more easily be done remotely, have a highly educated labour force and high quality IT infrastructure. All these factors facilitated the increase in working from home.

Contents

CHAPTER 1 The regional dimension of the COVID-19 pandemic    

1.1    The health impact of the pandemic    

1.2    The economic impact of the pandemic    

1.2.1    Pandemic restrictions    

1.2.2    The biggest post-war recession    

1.2.3    The tourist sector was most affected    

1.2.4    The impact on the EU labour market was muted    

1.2.5    Hours worked dropped substantially    

1.2.6    A big shift to working from home    

1.2.7    Regional impact is likely to be highly variable    

Figure 1.1 Excess mortality rate by geographic region, January 2020-July 2021    

Figure 1.2: Excess mortality per week by urban-rural typology, January 2020-November2021    

Figure 1.3 Stringency index by geographic region, January 2020 - September 2021    

Figure 1.4 Stay at home requirement index by geographic region, January 2020 – September 2021    

Figure 1.5: Change in real GDP growth relative to the previous year, 2020 and 2021    

Figure 1.6: Change in the number of nights spent in tourist accommodation, January 2020 to June 2021    

Figure 1.7 International travel restrictions index by geographic region, January 2020 – September 2021    

Figure 1.8: Annual number of tourist nights per resident in Member States, 2019-2020    

Figure 1.9: Monthly unemployment rate in the EU and United States, 2019 and 2020    

Figure 1.10: Quarterly employment rate by degree of urbanisation, 2018-2021    

Figure 1.11 Change in hours worked in Member States, 2019-2020    

Map 1.1 Excess mortality since week 9 of 2020    

Map 1.2 People fully vaccinated against COVID-19, November 2021    

Map 1.3 Population with access to green urban areas of at least one hectare within 400 metres of walking, 2018    

Map 1.4: Tourism vulnerability index, 2018, NUTS 3    

Map 1.5 Annual change in the actual number of hours worked, per week, 2020    

Map 1.6 Annual change in the share of people usually working from home, 2020    

Map 1.7: Simulated regional GDP impact of the crisis in 2020    

1.1The health impact of the pandemic

Between March 2020 and July 2021, the COVID-19 pandemic has led to excess mortality 1 in the EU of at least 872,000 deaths. In other words, compared to the average of the five previous years, the number of deaths since the start of the pandemic was 13% higher. This includes deaths directly resulting from COVID-19 and those caused indirectly because of the saturation of hospital capacity and lack of usual care. For example, half the NUTS-3 regions, for which data are available, experienced at least one week with over double the usual mortality ( Map 1. 1 ).

The excess mortality during the first wave mainly affected regions in Italy, Spain, France, Belgium and the Netherlands. During the second wave, excess mortality was highest predominantly in regions in Eastern Europe, in Poland, Bulgaria, Slovenia, Czechia, Romania and Hungary.

Figure 1.1 Excess mortality rate by geographic region, January 2020-July 2021

Source: Eurostat [demo_mexrt] and REGIO calculations

Regional excess mortality since the start of the pandemic shows the cumulative impact of the different waves ( Map 1. 1 ). It reveals hotspots in northern Italy and Madrid, which were heavily affected in the first wave, as well as in Poland, Czechia, Slovakia and Bulgaria, which were more affected in later waves. Overall 2 , less developed regions had the highest excess mortality rate (17% higher) as compared with transition (11%) and more developed regions (12%).

Map 1.1 Excess mortality since week 9 of 2020

The excess mortality rate during the first wave was highest in urban regions and peaked at 80% in April 2020, while it was lower than 40% in intermediate regions and only 20% in rural regions. During the second wave, rural regions had the highest excess rate, which peaked at 55%, while it was somewhat lower in towns and suburbs (48%) and cities (43%) ( Figure 1.2 ).

Figure 1.2: Excess mortality per week by urban-rural typology, January 2020-November2021

Source: Eurostat demo_r_mweek3 and JRC modelling

Note: Because of missing NUTS-3 data, Germany, Estonia, Ireland, Croatia, Malta and Slovenia are not included.

Because of COVID-19, life expectancy in 2020 fell in almost all Member States. The biggest reductions were in Spain (-1.6 years) and Bulgaria (-1.5 years). In only two Member States, Denmark and Finland did life expectancy increase, though only marginally 3 .

Vaccines offer the best way out of the pandemic. In November 2021, approximately 70% of the total population was fully vaccinated. Uptake of vaccinations, however, differed between and within Member States. Data reported in November indicated that in multiple regions in Romania and Bulgaria less than 20% of the population was fully vaccinated, while in many regions in Belgium, France and Spain more than 80% of the population was fully vaccinated ( Map 1. 2 .

Map 1.2 People fully vaccinated against COVID-19, November 2021

1.2The economic impact of the pandemic

The depth of the economic recession during the pandemic was affected by three main factors. First, the length and the strictness of lockdown measures implemented by national, regional and local authorities to limit the spread of the virus. The places with stricter lockdown measures tended to experience a deeper recession 7 . Second, some types of economic activities were much more affected than others. Services, notably accommodation and those relating to culture, leisure, tourism and activities requiring proximity generally have particularly suffered from the containment measures. Member States and regions that are more dependent on these sectors, have seen a bigger drop in their economic activity. Third, the policy response of Member States, regions and local authorities varied in scope and intensity, in part reflecting the differential impact of the pandemic.

1.2.1Pandemic restrictions

Restrictions imposed in response to the pandemic did not differ greatly between EU Member States ( Figure 1. 3 ). Restrictions peaked in April 2020, were relaxed in summer 2020, and were increased again during autumn and winter 2020-2021. Restrictions started to recede slowly in May 2021 and continued to do so up to September. On average, restrictions in eastern Member States were slightly less strict, while southern Member States had the tightest ones and north-western Member States were in between the two.

Figure 1.3 Stringency index by geographic region, January 2020 - September 2021

Source: Oxford COVID-19 Government Response Tracker, Blavatnik School of Government, University of Oxford. Unweighted averages of country indices.

The difference between Member States, however, was greater as regards specific kinds of restriction. For example, some Member States had long periods during which people were required not to leave their homes except for a short period of daily exercise, grocery shopping or essential trips ( Figure 1. 3 ). On the other hand, some Member States imposed no stay-at-home requirements for almost the whole period, while others imposed only modest restrictions. Eastern Member States tended to have the least restrictions and the southern ones the most. During the first wave, north-western Member States imposed similar restrictions to the eastern ones, while during the second and third waves, they had a stricter approach more similar to southern Member States.

Figure 1.4 Stay at home requirement index by geographic region, January 2020 – September 2021

Source: Oxford COVID-19 Government Response Tracker, Blavatnik School of Government, University of Oxford. Unweighted averages of country indices.

The stay-at-home requirements and the internal movement restrictions meant that people had to rely more on local facilities and amenities. The requirement to work from home and the closure of schools meant that many people in cities were crowded into small living spaces during the day. This highlighted the benefit of nearby green areas that were open to the public. In most cities the majority of the residents can reach at least one hectare of green urban area by walking a short distance. In a number of cities, however, less than half of the people have easy access to green urban spaces. This is the case in all the cities in Cyprus, Malta and Romania , and some big cities in Italy, France and Portugal, where less than half the residents have a green urban area within 400 metres of their home ( Map 1. 3 ). , The working from home requirements and remote lessons also posed challenges for households without fast internet connections, which is more often the case in rural areas.

Map 1.3 Population with access to green urban areas of at least one hectare within 400 metres of walking, 2018

1.2.2The biggest post-war recession

The COVID-19 pandemic triggered the deepest post-war recession in Europe. Real GDP growth averaged 2.1% a year between 2014 and 2019. In 2020, real GDP fell by 6.0%. All economic sectors were affected by the consequences of containment measures, the disruption of global supply chains, the sharp reduction in demand for goods and services, and the fall in tourism, business travel and recreation. Across Europe and the rest of the world, the crisis led to unprecedented policy responses to mitigate the effects of the shock and strengthen the recovery.

The economic impact of the COVID crisis varies widely across Member States. Between 2019 and 2020, there was a reduction in real GDP growth of 13 percentage points (pp) in Malta and Spain (GDP increasing by 5.5% in 2019 and falling by 7.8% in 2020 in the first and increasing by 2.0% in the second and falling by 10.8% in the second) while the reduction was less than 5 pp in Finland, Denmark and Luxemburg, and in Ireland, there was even a small increase ( Figure 1. 5 ). Economic activity rebounded in 2021, in particular in the Member States where it fell the most 8 .

Figure 1.5: Change in real GDP growth relative to the previous year, 2020 and 2021

Source: EUROSTAT table nama_10_gdp and ECFIN Spring 2021 forecast

1.2.3The tourist sector was most affected

Restrictions on movement within countries and limits on non-essential travel brought tourism to a standstill. The number of nights spent by tourists plummeted with the outbreak of the pandemic and the strict travel restrictions ( Figure 1. 6 ), falling by more than 90% compared to the same month in the previous year. The nights spent by domestic tourists recovered in the summer of 2020 but then fell again. The nights spent by international tourists remained extremely low throughout 2020 and the first half of 2021. Overall in 2020, nights spent dropped by 54% in relation to 2019, but those spent by international tourists fell by far more (70%) than those spent by domestic ones (39%).

Figure 1.6: Change in the number of nights spent in tourist accommodation, January 2020 to June 2021

Source: EUROSTAT table TOUR_OCC_NIM

These reductions were primarily caused by the restrictions on international travel that were introduced after the start of the pandemic. By the summer of 2020, all Member States had instituted some restrictions and these mostly stayed in place until summer 2021 ( Figure 1. 7 ). The restrictions on internal movement were part of the response to the first wave of the pandemic, but were loosened in summer 2020. During the second and third waves, internal restrictions remained much laxer. This allowed domestic tourism to recover somewhat during the summer of 2020, but the number of nights spent by tourists in the winter and spring of 2021 remained much less than in 2019.  

The restrictions on international travel also disproportionally affected border areas. People who usually crossed a national border for work, education, healthcare or other services were suddenly no longer able to do so. After the initial restrictions were put in place, while some borders made allowance for cross border commuting, many did not, which underlines the need for a better governance system for functional border areas.

Figure 1.7 International travel restrictions index by geographic region, January 2020 – September 2021

Source: Oxford COVID-19 Government Response Tracker, Blavatnik School of Government, University of Oxford. Unweighted averages of country indices.

The countries with the biggest reductions in the number of nights spent per resident were Cyprus, Malta, Croatia, Greece and Spain, with reductions of more than double the EU average ( Figure 1. 8 ). The reductions were much smaller in countries with generally relatively few tourist nights per resident.

 Figure 1.8: Annual number of tourist nights per resident in Member States, 2019-2020

Source: EUROSTAT TOUR_OCC_NIM, REGIO calculations.

Some regions are particularly dependent on tourism, including many of the Mediterranean islands and some coastal regions, the Alpine regions, the Black Sea Coast, Algarve and the Canary Islands. Some capitals and large cities also attract many tourists, but they are less dependent on tourism than coastal or mountain destinations because of much stronger and more diversified economies. To identify the regions most dependent on tourism, three indicators can be combined: nights spent per resident, seasonality of nights spent and the share of foreign tourists 9 (

Map 1. 4 ). Regions scoring highly on all three indicators are likely to have been more affected by the reduction in travel and nights spent. For example, the Mediterranean coastal and island regions are likely to have been particularly heavily affected.

Map 1.4: Tourism vulnerability index, 2018, NUTS 3

The tourism is not the only sector to have suffered from the economic downturn triggered by the pandemic. Contact-intensive services 10 were also severely affected. In the second quarter of 2020, activity in these sectors was 25% below pre-COVID-19 levels 11 . Other sectors were less affected but still experienced a sharp drop in activity, notably manufacturing (down by 19%) and construction (down by 15%). Services with significant scope for working remotely and with high-skilled workers, like ICT, banking and finance, contracted much less (by less than 10%) and these activities tended to rebound more quickly.

1.2.4The impact on the EU labour market was muted

The pandemic’s impact on the labour market was much more limited due to the many job retention schemes put in place shortly after the outbreak of the crisis. As a result, the economic slowdown did not lead to large increases in unemployment. The EU unemployment rate only went up by 0.5 pp between December 2019 and June 2021, from 6.6% to 7.1% with a peak at 7.7% in September 2020. By contrast, in the United States, which did not rely as much on job retention schemes, the unemployment rate doubled from 3% to 6% between December 2019 and June 2021, with a peak of 14% in April 2020 ( Figure 1. 9 ). 

At the EU level, employment 12  fell by 3 million, or 1.5%, between 2019 and 2020. Southern EU lost the most employment (2.7%). The reduction in eastern EU was smaller (1.2%), while in north-western EU, it fell by least (0.9%). Employment started to recover in the second quarter of 2021 but has not yet reached its 2019 level. 

Figure 1.9: Monthly unemployment rate in the EU and United States, 2019 and 2020

Source: EUROSTAT table une_rt_m

As reflected by the unemployment figures, the employment rate (of those aged 20-64) in the EU also fell by relatively little, by 0.7 pp between 2019 and 2020. The reduction was largest (1.4 pp) in the southern EU, followed by the north-western EU (0.6 pp) and t the eastern EU (0.2 pp).

Across the EU, the employment rate declined by most in towns and suburbs (1.1 pp) between 2019 and 2020, followed by cities (0.7 pp) and it barely fell at all in rural areas (by 0.3 pp). The quarterly figures show that the reduction was largest in cities in the second quarter, but it was then overtaken by the fall in towns and suburbs ( Figure 1.10 ).

Figure 1.10: Quarterly employment rate by degree of urbanisation, 2018-2021 

Source: Eurostat, LFS (non-seasonally adjusted) table lfsq_pgauws

Note: Labels show y-on-y change in pp.

1.2.5Hours worked dropped substantially

Because of the pandemic, the number of hours worked declined significantly in the EU between 2019 and 2020, though the scale of the reduction depends on the source of the data used and the working time for which hours are measured. The Labour Force Survey (LFS), which measures weekly hours, indicates a reduction in the EU of 12%, whereas the national accounts data 13 , which measures annual hours, shows a reduction of 6% ( Figure 1. 11 ). Both sources agree, however, that the biggest reductions occurred in Greece, Spain, Portugal and Italy. The LFS data also show that regions with large tourist economies were especially affected ( Map 1.5 ). More developed regions were slightly less affected (with a reduction of 10% based on LFS data) than transition and less developed regions (a fall of 13% in each).

Figure 1.11 Change in hours worked in Member States, 2019-2020

Source: Eurostat LFS ad hoc extraction and table NAMA_10_A10_E

The biggest reduction in hours worked over the period occurred in the accommodation and food services sector (by 52%) and the arts, entertainment and recreation sector (by 36%). The two most affected broad occupational groups were service and sales workers (which showed a fall of 27%) and elementary occupations (one of 23%).

1.2.6A big shift to working from home

In 2019 14 , 5.5% of the employed population in the EU usually worked from home. Because of the pandemic, and the requirement to work from home where possible, the proportion more than doubled to 12.4% in 2020. The capacity to work from home depends on the type of activity concerned. Some jobs can only be performed in person, as noted above, such as many jobs in healthcare, manufacturing and agriculture. Many of the regions with large cities saw big increases in the proportion of people working from home, reflecting the large share of economic activities which can be performed remotely (usually by high-skilled workers). In particular, the increases were over 15 pp in the Brussels, Helsinki, Dublin, Paris, Cologne and Vienna regions ( Map 1. 5 ). The distribution of critical 15 and ‘teleworkable’ jobs strongly depends on the degree of urbanisation. Rural areas tend to have a larger share of ‘non-teleworkable’ jobs than cities, towns and suburbs 16 .

1.2.7Regional impact is likely to be highly variable

Regional GDP data for 2020 is not yet available, which limits the extent to which the impact of the COVID pandemic on the economies of the EU regions can be assessed. A modelling exercise 17 using national data and the RHOMOLO regional model, however,  gives an indication of the potential regional impact. It shows a particularly severe impact on southern European regions and France, and less effect on Nordic and eastern regions ( Map 1.7 ). The model suggests that in Spain, Italy, France, and Greece, some regions are likely to experience a particularly sharp reduction in GDP. This is especially so for those with a large share of value-added in wholesale and retail trade, transport and accommodation, i.e. in the sectors where tourism is important, which is line with the actual changes in hours worked in 2020 indicated above.

Map 1.7: Simulated regional GDP impact of the crisis in 2020

(1)   https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Excess_mortality_-_statistics  

(2) Data for Ireland, Slovenia and three German regions (DE9, DEB, DED) are missing at the regional level.

(3)   Data for Ireland were not yet available for 2020.

(4) CoR, EU Annual Regional and Local Barometer, October 2021. Available online: https://cor.europa.eu/en/our-work/Documents/barometer-fullreport%20web.pdf  

(5) The territorial impact of COVID-19: Managing the crisis across levels of government, November 2020. Available online: https://www.oecd.org/coronavirus/policy-responses/the-territorial-impact-of-COVID-19-managing-the-crisis-across-levels-of-government-d3e314e1/

(6) CoR, Study: Local and regional finances in the aftermath of COVID-19. June 2021. Available online at: https://cor.europa.eu/en/engage/studies/Documents/Local%20and%20regional%20finances%20in%20the%20aftermath%20of%20COVID-19/CoR_Local_and_regional_finances_after_Covid-19.pdf  

(7) Sapir, A. (2020), “Why has COVID-19 hit different European Union economies so differently”, Bruegel, Issue 18

(8)      European Commission (2021), “European Economic Forecast Spring 2021”, Directorate-General for Economic and Financial Affairs, Institutional Paper 149, : Publications Office of the European Union, Luxemburg.

(9)      Batista e Silva F, Marin-Herrera MA, Rosina, K, Barranco R, Freire S, Schiavina M (2018), “Analysing spatiotemporal patterns of tourism in Europe at high resolution with conventional and bit data sources”. Tourism Management 68: 101-115. doi:10.1016/j.tourman.2018.02.020

(10)  Trade, transport and accommodation, and arts, entertainment and other service activities

(11) European Commission (2021), “The sectoral impact of the COVID-19 crisis”, Technical Note for the Eurogroup.

(12) Source: Eurostat, National Accounts; domestic employment

(13) 2020 data for 11 Member States is flagged as provisional.

(14) Source: Eurostat, LFS ad-hoc module 2019

(15) Critical jobs can be defined as all those occupations that need to be performed even during a pandemic in order to keep citizens heathy, safe and fed.

(16) Employment and Social Developments in Europe 2021.

(17)      Based on national figures for 2020 on employment, output in the various NACE sectors, exports and the rise in uncertainty assumed to be reflected in an increase in interest rates, Sakkas et al. (2021) used the RHOMOLO model to estimate the impact of the crisis on NUTS-2 regions. The magnitude of the shocks are calibrated so that the ranking of countries in terms of output loss is, so far as possible, in line with the latest real GDP growth figures for 2020 published in the Spring 2021 European Economy Forecast .

EUROPEAN COMMISSION

Brussels, 4.2.2022

SWD(2022) 24 final

COMMISSION STAFF WORKING DOCUMENT

Cohesion in Europe towards 2050

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 8th Cohesion Report: Cohesion in Europe towards 2050

{COM(2022) 34 final}

CHAPTER 2. A SMARTER EUROPE – part 1

·After the financial and economic crisis years and its aftermath, the EU economy is growing again, with growth being particularly high in low-income countries.

·After a long period of convergence, since the crisis in 2008 regional disparities in GDP per head have stopped shrinking. Regional disparities in employment and unemployment rates increased dramatically after the economic crisis. Since 2013, they have started shrinking again, but remain significantly greater than in 2007.

·GDP per head in the less developed regions is converging towards the EU average through both faster productivity growth and increased employment. This trend is primarily driven by developments in regions in the eastern Member States whereas many less developed regions in the southern Member States are failing to catch up and experiencing decline and divergence.

·The last two decades have witnessed a modernisation of the agricultural sector, evidenced by a long-term and ongoing increase in productivity and decrease in employment. These developments have been particularly pronounced in the less developed regions, which have experienced a sectoral restructuring of the economy.

·Transition regions, with a GDP per head between 75% and 100% of the EU average, seem stuck in a ‘development trap.’ Between 2001 and 2019, their growth in GDP per head was far below the EU average, and their productivity growth and employment creation was less than in other regions. Their manufacturing sectors are smaller than those in regions with a lower or higher GDP per head and their innovation and education systems and institutional quality are not strong enough to be competitive at the global level.

·Innovation in the EU remains highly concentrated in capital and other metropolitan regions. In north-western EU countries, good regional connections, high digital readiness, a skilled labour force and an attractive business environment have enabled surrounding regions to benefit from proximity to highly innovative ones. In southern and eastern EU countries, the most innovative regions are less strong and, accordingly, neighbouring regions reap little benefit. These patterns could lead to a widening research and innovation divide between EU regions.

Contents

CHAPTER 2. A SMARTER EUROPE – part 1    

2.1    RECENT TRENDS IN CONVERGENCE AND DIVERGENCE BETWEEN EU MEMBER STATES AND REGIONS    

2.2    PRODUCTIVITY IN LESS DEVELOPED MEMBER STATES IS CATCHING-UP    

2.2.1 Employment in agriculture and industry is shrinking while productivity is increasing    

2.2.1 Productivity is the main factor underlying growth in GDP per head    

2.2.2 Capital metropolitan regions perform better than other regions    

2.3    Development traps And related risks for European regions    

2.3.1 Regional stagnation and development traps    

2.3.2 Identifying Development Traps in EU regions    

Figure 2‑1: Growth rates of GDP per head in regions in less developed and moderately developed Member States, 2001-2008

Figure 2‑2: Growth rates of GDP per head in regions in less developed and moderately developed Member States, 2009-2013

Figure 2‑3: Growth rates of GDP per head in regions in less developed and moderately developed Member States, 2014-2019

Figure 2‑4: Regional disparities between NUTS-2 regions in the EU, 2000-2020

Figure 2‑5: Growth of GDP per head in real terms by level of development, 2001-2019

Figure 2‑6: Changes in GDP per head (PPS), 2000-2019

Figure 2‑7: Evolution of total employment (number employed) in metro and non-metro regions, 2000-2019, (index 2000=100)

Figure 2‑8: Annual growth in real GDP per head in EU regions by level of development, 2001-2019

Figure 2‑9: Share of population living in regions which experienced very low growth in GDP per head, productivity and employment, 2001-2019, by initial level of GDP per head (index, 2000=100)

Map 2‑1 GDP per head (PPS), 2019

Map 2‑2 Growth of GDP per head, 2001-2019

Map 2‑3 Growth of GDP per head in real terms between 2001-2019, main sub-periods

Map 2‑4: Growth of GDP per head, productivity, the employment rate and working-age population, 2001-2019

Map 2‑5: Transition of NUTS 2 regions between development categories, 2000- 2019

Map 2‑6: Number of years in a development trap during 2001-2019 by level of GDP per head in 2000

Table 2‑3: Employment and GVA by NACE sector and category of region, % shares in 2018 and changes, 2001-2018

Table 2‑4: Decomposition of annual average change in GDP per head, 2001-2019 and sub-periods

Table 2‑5: Changes in GDP per head, productivity and employment per head by type of region, 2001–2019

Table 2‑6: Socio-economic characteristics of development trapped regions and other regions by level of GDP per head

CHAPTER 2. A SMARTER EUROPE – PART 1

Regional economic convergence 1 has stopped in the EU and divergence could become a threat to economic progress (Iammarino et. al., 2017) at a time when globalisation poses new challenges to economic cohesion. While the evidence suggests that the EU economy as a whole has benefited, and continues to benefit, from globalisation, these benefits are not automatically and evenly transmitted to all regions.

This chapter examines recent trends in economic cohesion in regions and cities across the EU, as reflected in GDP per head and in the underlying developments in productivity and employment. It assesses the risk of regions falling into a ‘development trap’ and discusses the factors underlying regional competitiveness, including entrepreneurship, digitalisation and innovation. It also presents an aggregate indicator, the Regional Competitiveness Index, intended to summarise the different dimensions of competitiveness.

The main concern throughout the chapter is to highlight the performance of the less developed regions against the more developed ones and of rural areas compared to cities.

2.1RECENT TRENDS IN CONVERGENCE AND DIVERGENCE BETWEEN EU MEMBER STATES AND REGIONS

In 2019, over one in four people in the EU (29%) lived in a NUTS 2 region with GDP per head below 75% of the EU average in PPS terms 2 , most of them in eastern Member States 3 , Greece, Portugal, Spain, and southern Italy as well as in the outermost regions 4  ( Map 2 ‑1 ). In Bulgaria, GDP per head was below 50% of the EU average in all regions, except in Yugozapaden, the capital city region.

Over the 2001-2019 period, GDP per head in real terms increased in the vast majority of EU regions ( Map 2 ‑2 ), albeit at a modest rate in most cases. Growth was particularly high in the eastern Member States and Ireland. In most regions in Greece, however, GDP per head fell over this period, as it did in Italy, both in many of the more developed regions in the north and in many of the less developed in the south. At the same time, growth was very low in transition regions in the north of France.

Between 2001 and 2008, nearly all regions experienced growth in GDP per head ( Map 2 ‑3 ). Overall, growth was above average in both the less developed and the transition regions, with rates of over 5% a year in many of those in eastern Member States. This is in line with mainstream economic growth theories, which predict that growth will tend to be higher the lower the initial level of GDP per head. Most of these regions are in less developed and moderately developed Member States 5 , where for the most part growth was faster than the EU average ( Figure 2 ‑1 ). In Romania and Bulgaria, where the growth rate was particularly high, the 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 negative growth in the 2000s even though their GDP per head was well below the EU average.

The global financial crisis of 2007-2008 led to GDP per head in the EU declining between 2009 and 2013. Around 60% of the EU population lived in regions with a declining GDP per head ( Map 2 ‑3 , Figure 2 ‑2 ). The regions hit hardest were mainly in the southern EU countries, though also in Romania, Ireland and Finland. In most Greek regions, the reduction in GDP per head averaged over 3% a year. The crisis led to many of the less developed and transition regions growing more slowly (or shrinking faster) than the EU average during this period, so reversing the tendency towards convergence. The process of convergence was, therefore, brought to an end and disparities began to widen again. Most regions in Poland and some in Bulgaria and Romania were notable exceptions.

The 2014-2019 period shows a clear recovery from the Great Recession ( Map 2 ‑3 , Figure 2 ‑3 ). Almost all regions experienced growth in GDP per head, though at a lower rate than in the pre-crisis period. High growth rates were restored in most eastern regions, so contributing again to convergence. By contrast, growth in many north-western regions 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, more than a quarter of the EU population live in a region where by 2019 real GDP per head had still not returned to pre-crisis levels. This includes the entire population of Greece and Cyprus, 80% of Italians and a third of Spaniards, but also 75% of the Finnish population and over a third of Austrians. 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, live in regions where this is not the case.

Map 2‑3 Growth of GDP per head in real terms between 2001-2019, main sub-periods

Figure 2‑1: Growth rates of GDP per head in regions in less developed and moderately developed Member States, 2001-2008

Note: Regions are ranked by the growth rate of GDP per head over the period 2001-2019; Capital city regions are indicated in red

Source: ARDECO and Eurostat [nama_10r_2gdp], DG REGIO calculations

Figure 2‑2: Growth rates of GDP per head in regions in less developed and moderately developed Member States, 2009-2013

Note: Regions are ranked by the growth rate of GDP per head over the period 2001-2019; Capital city regions are indicated in red.

Source: ARDECO and Eurostat, DG REGIO calculations

Figure 2‑3: Growth rates of GDP per head in regions in less developed and moderately developed Member States, 2014-2019

Note: Regions are ranked by the growth rate of GDP per head over the period 2001-2019; Capital city regions are indicated in red.

Source: ARDECO and Eurostat, DG REGIO calculations

Prior to the 2007-2008 crisis, disparities in GDP per head in the EU were shrinking 6 , mainly because of regions with the lowest levels growing faster than average ( Figure 2 ‑4 ). However, in the years immediately following the crisis, regional disparities widened slightly. There are signs that the long-term process of regional convergence, which was interrupted by the crisis has resumed, although at a very slow pace. 

Figure 2‑4: Regional disparities between NUTS-2 regions in the EU, 2000-2020

Note: Disparities are measured by the coefficient of variation (CV) and the mean absolute deviation (MAD). Both are weighted by the population in each region. The analysis is based on the NUTS2 level but regions which are part of the same metropolitan area are combined.

Source: Eurostat [nama_10r_2gdp, reg_lmk], DG REGIO calculations.

Regional disparities in employment and unemployment rates 7 also narrowed from 2000 up to the financial crisis when they widened to reach a new peak in 2013. After then, they began narrowing again, but, in 2020, the disparities in both were wider than in 2008. Disparities in the employment rate remain at much the same level as in 2000.

The economic convergence of regions over the period 2001-2019, as noted above, was mainly driven by the catching up of many of the less developed ones, their GDP per head growing faster than elsewhere, except in 2010 and 2011 immediately following the global financial crisis ( Figure 2 ‑5 ). The average picture, however, hides differing trends among less developed regions. While there has been strong growth and significant catching up in those in eastern Europe, many less developed regions in southern Europe have experienced sluggish or negative growth and their GDP per head is diverging away from the EU average (Section 2.3 below examines on these trends further).

The transition regions, however, do not follow the same pattern. From 2005 onwards, growth in these regions was below the EU average, except in 2009. As a result, GDP per head, in PPS terms, diverged from the EU average instead of converging ( Figure 2 ‑ 6 a). 

Figure 2‑5: Growth of GDP per head in real terms by level of development, 2001-2019

Source: ARDECO and Eurostat, ARDECO, DG REGIO calculations

Predominantly rural regions have a GDP per head, in PPS terms, around 70% of the EU average ( Figure 2 ‑ 6 b). Over the period 2001-2019 rural regions close to cities showed convergence to the EU average. This did, however, not hold for remote rural regions where GDP per head slightly decreased relative to the EU. Remote intermediate regions also diverged from the EU average over this period.

Figure 2‑6: Changes in GDP per head (PPS), 2000-2019

Source: Eurostat [nama_10r_2gdp], ARDECO, DG REGIO calculations

The growing interdependence of the world’s economies has had a highly differentiated impact on EU regions 8 . While some have been well positioned to take advantage of the new opportunities it offers, others have been hit by job losses, stagnating wages and shrinking market shares as a result of low-cost competitors moving into more technologically advanced sectors (see also Section 2.4 below).

2.2PRODUCTIVITY IN LESS DEVELOPED MEMBER STATES IS CATCHING-UP

2.2.1 Employment in agriculture and industry is shrinking while productivity is increasing

Regions at different levels of development tend to have different economic structures. Less developed regions tend to have relatively large shares of employment in agriculture and industry ( Table 2 ‑ 3 ). In 2018, over 12% employment in these regions was in agriculture, three times more than in transition regions and 8 times more than in more developed ones. Around 21% of employment was in industry, 6 percentage points (pp) more than in transition and more developed regions. Transition and more developed regions are more comparable in terms of their employment shares, with more employed in finance and insurance and public administration.

The sectoral composition of gross value-added (GVA) follows the same general pattern as employment, but the differences between regions at different levels of development tend to be less pronounced. Notably, despite the large work force in agriculture in less developed regions, GVA from agriculture is modest, implying low productivity.

Employment in agriculture fell between 2001 and 2018, especially in the less developed regions (by over 3% a year), reflecting their economic restructuring and agricultural modernisation. The latter led to a substantial increase in productivity in the sector and an increase in GVA. Given the large share of employment in agriculture in these regions, this process is likely to continue. The same pattern is observed in the transition and more developed regions, but the reduction in employment and growth in GVA were less than half that in less developed regions.

Employment in industry also declined in each of the three types of region, though much less so than in agriculture. Despite the loss of labour, GVA increased substantially, as did productivity, especially in the less developed regions. The EU single market has created more potential for specialisation in higher value-added sectors, enabling less developed and some transition regions to maintain a larger share of employment in industry, because they have an attractive balance between labour costs, productivity and accessibility.

The construction industry showed little growth over the 2001-2018 period and even contracted slightly in transition regions. By contrast, employment and GVA in services increased in all regional groups over the period, particularly in financial activities, especially in less developed regions.

Table 2‑3: Employment and GVA by NACE sector and category of region, % shares in 2018 and changes, 2001-2018

Green bars indicate positive changes, red bars indicate negative changes.

Source: Eurostat [nama_10r_3empers], ARDECO, Cambridge Econometrics, AMECO, DG REGIO calculations

2.2.1 Productivity is the main factor underlying growth in GDP per head

Over the 2001-2019 period, GDP per head increased in the vast majority of EU regions ( Map 2 ‑4  and Table 2 ‑2 ). The increase was largely associated with productivity growth 9 , and to a lesser extent with employment growth. Working-age population as a share of the total decreased slightly in the EU and in most regions over this period. Many less developed regions, especially those located in the eastern Member States, had above average productivity and employment growth, offset only slightly by a decline in the share of working-age population, so that growth of GDP per head growth was above the EU average. This, however, masks the fact that in a number of these regions, mainly in Greece and Italy, GDP per head fell over this period, with productivity falling and the employment rate declining or increasing relatively little, combined with a shrinking share of working age population. In most of the EU outermost regions GDP per head remained at the same level or decreased. 

From 2001 to 2008, GDP per head in the EU grew by 1.8% a year in real terms, with productivity growing by 1.2% a year and an increase in the employment rate adding another 0.4% a year ( Table 2 ‑ 4 ). In many less developed regions, where GDP growth was substantially higher than the EU average, productivity growth was also the main component of growth in GDP per head, and even more so than in the EU as a whole, while the employment rate remained unchanged.

Between 2009 and 2013, GDP per head in the EU declined by 0.4% a year. Employment also declined (by 0.5% a year) as both the employment rate and the share of the working-age population fell, while productivity continued to increase, though at a slower rate. This pattern of change is mirrored in each group of regions, but it is more pronounced in the less developed regions and less pronounced in the more developed ones. Accordingly, the less developed regions, as a group, experienced the sharpest decline in GDP per head, but also in the employment rate.

Map 2‑4: Growth of GDP per head, productivity, the employment rate and working-age population, 2001-2019

Between 2014 and 2019, growth of GDP per head resumed in every regional group. Unlike in the period before the financial crisis, however, growth was strongly associated with an increase in the employment rate, which more than offset a reduction in the share of working-age population, while labour productivity grew more slowly than in the pre-crisis period. Again, this pattern of change was more pronounced in the less developed regions. On the other hand, recovery was more subdued in the transition regions, with GDP per head growth being only slightly more than half that in less developed regions, much the same as in the pre-crisis period.

Table 2‑4: Decomposition of annual average change in GDP per head, 2001-2019 and sub-periods

Green bars indicate positive changes, red bars indicate negative changes

*Workplace-based employment divided by population aged 20-64

Less developed regions exclude Mayotte

Source: Eurostat [nama_10r_3empers], ARDECO, Cambridge Econometrics, AMECO, DG REGIO calculations

2.2.2 Capital metropolitan regions perform better than other regions

In 2019, metropolitan (metro) regions accounted for 59% of population in the EU, 63% of employment and 68% of GDP. Accordingly, they are major centres of employment and business activity with higher productivity than elsewhere.

Between 2001 and 2019, real GDP per head in metro regions grew faster than in others in all parts of the EU. ( Table 2 ‑ 5 ). This was a result mainly of above average growth rates in capital city regions, though other metropolitan regions also outperformed non-metropolitan regions, except in the north-western Member States.

In regions in the eastern and north-western Member States, the growth of GDP per head was mainly associated with productivity growth. The pattern is different in southern Member States. Productivity growth was very low during this period and most of the (modest) growth in GDP per head was associated with growth in employment. In capital metro regions in the eastern and southern Member States, the contribution of employment growth to GDP growth was double the average, reflecting a continuing concentration of employment there.

Table 2‑5: Changes in GDP per head, productivity and employment per head by type of region, 2001–2019

* This combines the employment rate and working-age population as a share of the total

Source: Eurostat [reg_eco10], ARDECO, Cambridge Econometrics, AMECO, DG REGIO calculations

Employment in both metro and non-metro regions increased between 2000 and 2008, although at a faster rate in capital metro regions than in other metro regions and by more in the latter than in non-metro regions ( Figure 2 ‑ 7 ). In the following two years, it declined markedly in all regions. In the capital city regions, it began to recover in 2010, with the growth rate accelerating in 2013 and continuing at the same pace up to 2019, when total employment was significantly higher than before the 2007-2008 crisis. In other metro regions, recovery was more hesitant. Employment remained below pre-crisis levels up until 2015, and from then to 2019, its growth rate was more modest than in the capital city regions. In non-metro regions, the effect of the financial crisis was more sustained; employment only began to increase in 2013 and it grew by much less than in metro regions up to 2019, only reaching pre-crisis levels in 2018.

Figure 2‑7: Evolution of total employment (number employed) in metro and non-metro regions, 2000-2019, (index 2000=100)

Source: Eurostat [reg_eco10], ARDECO, Cambridge Econometrics, AMECO, DG REGIO calculations

2.3    Development traps 10 And related risks for European regions

2.3.1 Regional stagnation and development traps

It has become increasingly clear over recent years that not all regions in the EU with a GDP per head below the average are catching up. Regions can be categorised into different groups, defined in terms of their level of GDP 11 , but also by their rates of GDP growth.

Relating the annual growth of real GDP per head over the 2001-2019 period to the initial level of development of regions in 2000, as measured by GDP per head, reveals some striking patterns ( Figure 2 ‑8 ).

Figure 2‑8: Annual growth in real GDP per head in EU regions by level of development, 2001-2019

Source: Eurostat [nama_10r_2gdp], DG REGIO calculations

Some of the patterns are in line with convergence theory. In particular, many of the regions with GDP per head below 75% of the EU average in 2000 displayed strong growth over the subsequent 19 years, demonstrating rapid catching up. These regions are mainly those in eastern EU Member States. Conversely, many of the southern EU regions failed to achieve comparably high growth rates. A non-negligible number of southern regions experienced a reduction in GDP per head over the period, even if their initial GDP per head was below 75% of the EU average. Consistent with convergence theory, regions with above-average GDP per head in 2000, tended to have lower rates of growth.

However, growth in the group of regions with GDP per head between 75% and 100% of the EU average (i.e. the middle category), does not show any indication of catching-up. Indeed, average growth in these regions was below that of those with above-average GDP per head. Many of them, primarily those in southern EU Member States, experienced lengthy periods of low or negative growth, weak productivity increases and low employment creation or even job losses.

Iammarino et al. (2020) develop a concept of ‘development traps’, which is based on more dimensions than just a slowdown in GDP growth. It covers three dimensions of the economic dynamism of a region: GDP per head, productivity and employment. Some 45% of the population of the above mentioned middle category regions in 2000 were in regions where growth was very low 12 over the 2001-2019 period ( Figure 2 ‑ 9 ). Moreover, a third of the population were in regions where productivity growth was very low and 40% in regions with very low employment creation relative to the change in population. All of these population shares are higher, in some cases considerably, than in other regions.

Figure 2‑9: Share of population living in regions which experienced very low growth in GDP per head, productivity and employment, 2001-2019, by initial level of GDP per head (index, 2000=100)

Very low growth is defined here as annual average growth over the period in the bottom quartile of regions.

Source: Eurostat [nama_10r_2gdp], ARDECO, Cambridge Econometrics, AMECO, DG REGIO calculations

Since 2000, an increasing number of regions have experienced stagnating economic development after reaching a level of GDP per head of 75-100% of the EU average ( Map 2 ‑5 ). As this group has grown larger over time, transition out of it has become rarer. Indeed only one region (Zahodna Slovenija) out of a total of 53 regions in the middle category in 2000 managed to achieve above-average GDP per head by 2019. 13 On the other hand, in 18 of these regions, mainly in the southern EU, GDP per head fell below 75% of the EU average, implying divergence and increasing disparities.

Map 2‑5: Transition of NUTS 2 regions between development categories, 2000- 2019

The low growth of regions in the middle category suggests that they may have fallen into a development trap. Many of them are less cost-competitive than less developed regions, characterised by low-cost of capital and labour, and less innovative or productive than more developed regions. Accordingly, their costs tend to be too high to compete with less developed regions and their innovation systems not strong enough to compete with more developed regions. This makes it very difficult for them to escape the development trap and achieve higher GDP per head. While some of these regions had low GDP per head earlier and were catching-up until some years ago, others were formerly relatively prosperous but have moved into a prolonged period of relative economic decline. Indeed, in a quarter of the regions with above average GDP per head in 2000, mainly in north-western but also in southern Member States, GDP per head had fallen below the EU average by 2019 ( Map 2 ‑5 ).

2.3.2 Identifying Development Traps in EU regions

In Iammarino et al. (2020) the risk of a region being in a development trap in a specific year is assessed in terms of the pattern of growth of GDP per head, productivity and employment, as well as their growth relative to that of the Member State the region is located in and the EU average.

Analysis based on this approach shows that the number of years that regions were in a development trap over the 2001-2019 period varies greatly between them ( Map 2 ‑6 ). In general, regions that were in a development trap in 15 years or more during this period (henceforth called ’development trapped‘ regions) are concentrated in southern EU Member States (especially in Greece and Italy) or are rural or old industrial regions in France. Some of the regions, however, are also located in many of the north-western Member States, and so include regions at different levels of initial development. Accordingly, three types of development-trapped region can be identified in terms of their GDP per head in 2000.

-Development-trapped regions with very low GDP per head, which receive substantial Cohesion policy support, but which, unlike most of the other less developed EU regions, have struggled to sustain long term growth, so consistently lag behind other regions in the EU. Regions in this group include Calabria in Italy, and Anatoliki Makedonia, Thraki and Ipeiros, and Dytiki Ellada in Greece.

-Development-trapped regions with below average GDP per head between 75% and 100% of the EU average in 2000, but where economic dynamism has since stagnated. Accordingly, they have struggled to improve their standing, often in both relative and absolute terms. This group includes a number of regions in the Italian Mezzogiorno and regions in Portugal, Greece and Cyprus, as well as several regions in France and Wallonia in Belgium.

-Development-trapped regions with above-average GDP per head, which despite still being relatively prosperous have experienced frequent or long periods of below average growth in GDP, productivity and employment, often because of the demise of industries that used to be their main source of wealth. This group includes a number of regions in northern and central Italy, various regions in France, and a few in Spain, Portugal, Germany, Denmark, Austria and the Netherlands.

Map 2‑6: Number of years in a development trap during 2001-2019 by level of GDP per head in 2000

Source: ARDECO, DG-REGIO

The reasons for falling into a development trap differ between regions. However, there are a number of common traits, including, for example the levels of value-added in industry, human capital, innovation endowment and institutional quality.

EU regions that were development-trapped in 2000-2019 tend to have a smaller share of industrial output in total production, smaller endowments of human capital, (fewer workers with tertiary education,) and lower levels of support for science and technology ( Table 2 ‑ 6 ). Regions with a better quality of local government, and so a more favourable institutional environment, tend to fare better than those with low government efficiency, limited transparency and accountability, and more corruption. Development-trapped regions also tend to have higher old-age dependency rates and less demographic dynamism, though this is likely to be as much a consequence as a cause of being trapped.

Table 2‑6: Socio-economic characteristics of development trapped regions and other regions by level of GDP per head

Source: Eurostat [nama_10r_3gva, rd_e_gerdreg, edat_lfse_04], ARDECO, Cambridge Econometrics, AMECO, World Bank, DG REGIO calculations

The differing characteristics of the regions suggest different approaches to avoiding being development-trapped, depending on a region’s level of development. The chances of a region with below average GDP per head in 2000 avoiding being trapped are improved by having a better quality of government and larger industrial output. The latter would also improve the chances of transition regions in this respect. For more developed regions, the chances of staying out of a development trap are better if they have higher R&D investment and a more highly educated work force. In all regions, the chances could be improved by increasing the share of working-age population with tertiary education.

Regional development traps are a serious risk for the future of the EU. Springing these traps and so liberating the untapped economic potential of the many struggling and stagnating regions in the EU would not only increase their GDP, productivity and employment, but would also boost the growth potential of the EU as a whole. This is not just an economic matter; the sub-par economic performance and lack of employment opportunities are causing social costs and political resentment towards what is increasingly regarded as a system that does not benefit areas that are left behind, leading to a growing geography of discontent. 14  

Since development traps can occur at different levels of development, and appear to be a particular risk for transition regions, they may require policy responses that go beyond the poorest regions. Assisting all regions that are development-trapped to become more dynamic will help to reduce regional inequalities and counter the threat of rising discontent in EU societies.

(1) In this report ‘economic convergence’ primarily refers to a decrease in regional disparities in gross domestic product per capita. However, the chapter also discusses trends in disparities in related concepts such as productivity and employment.

(2)

GDP per head in PPS (Purchasing Power Standards) terms is the total value of goods and services produced per inhabitant adjusted for differences in in price levels.

(3) Eastern Member States are those in central and eastern Europe, which have joined the EU since 2004.

(4)

 The EU includes nine Outermost Regions: Guadeloupe, La Réunion, Mayotte, Guyane, Martinique, Saint-Martin (France), Madeira and Açores (Portugal) and Canarias (Spain).

(5) See the Lexicon section for the list of less developed and moderately developed Member States.

(6) The coefficient of variation, weighted by total regional population, fell by 12% during 2001-2008

(7) As measured by the mean absolute deviation weighted by total regional population.

(8) European Commission (2017)

(9) Note that this productivity growth, as being measured by GDP per person employed, does not reflect the decrease in the average hours worked per person employed during this period.

(10)  Prof. Simona, Prof. Andrés Rodríguez-Pose, and Prof. Michael Storper contributed substantially to the content of this section.

(11) Throughout this section regions are classified based on their GDP per head relative to the EU in 2000. The thresholds applied correspond to those currently used to classify regions as “less developed”, “transition” or “more developed”, but differ from those used in 2000. These group labels are therefore not used in this section.

(12)

Here, very low growth is defined as annual average growth over the period in the bottom quartile of regions ranked by the rate of growth (i.e. in the 25% with the lowest growth over the period 2001-2019).

(13) It is worth noting that Zahodna Slovenija improved its performance over the period in terms of the indicators identified here as determining factors of the risk of being ‘development-trapped’, with a larger than average share of industry in GVA, higher than average R&D expenditure relative to GDP and a larger than average share of working-age population with tertiary education. Institutional quality, however, remains below the EU average.

(14) See Dijkstra et al. (2020), who show that political discontent with the EU in Member States and regions is linked to an important extent to economic and industrial decline.

EUROPEAN COMMISSION

Brussels, 4.2.2022

SWD(2022) 24 final

COMMISSION STAFF WORKING DOCUMENT

Cohesion in Europe towards 2050

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 8th Cohesion Report: Cohesion in Europe towards 2050

{COM(2022) 34 final}

CHAPTER 2. A SMARTER EUROPE – PART 2

Contents

CHAPTER 2. A SMARTER EUROPE – PART 2    

2.4 COMPETITIVENESS OF EU REGIONS    

2.4.1 Innovation, digitalisation and smart specialisation    

The Regional Innovation Scoreboard    

Expanding digitalisation    

2.4.2 Firm dynamics in EU regions    

2.4.3 Regional competitiveness in Europe    

Figure 2‑10: Patent applications to the European Patent Office by type of region, 2016-17

Figure 2‑11: GVC profile, R&D expenditure (% of GDP) and Gini-coefficient by Member State

Figure 2‑12: Total expenditure on R&D as a % of GDP, 2001 and 2019

Figure 2‑13: Total expenditure on R&D as % of GDP, 2019

Figure 2‑14: Share of EU population by RIS category and level of development, 2016 and 2021

Figure 2‑15: EU enterprises take-up of digital technologies by Member State level of development, 2020

Figure 2‑16: EU enterprise take-up of e-commerce and e-business technologies by Member State level of development, 2020

Figure 2‑17: Composition of investments (in %), by level of development

Figure 2‑18: Employer firm birth rates by type of region, 2018

Figure 2‑19: Number of high growth firms by type of region, 2018

Figure 2‑20: Creative destruction and GDP growth in EU regions, 2008-2018

Figure 2‑21: Distribution of regional RCI 2019 scores by Member State

Map 2‑7: Patent applications to the European Patent Office, average 2016-2017

Map 2‑8: Total expenditure on R&D as a % of GDP, 2019

Map 2‑9: Regional Innovation Scoreboard, 2021

Map 2‑10: Active employer businesses per 1000 inhabitants, 2018

Map 2‑11: Regional Competitiveness Index scores, 2019                24

Table 2‑7: Total R&D expenditure and the distance to the Europe 2020 target, EU-27 regions, 2019

2.4 COMPETITIVENESS OF EU REGIONS

2.4.1 Innovation, digitalisation and smart specialisation

Innovation is an important driver of long-run productivity growth and, as such, is a key factor in supporting the competitiveness of firms. This is especially important for firms in the EU, which increasingly have to compete with firms in developing regions of the world, such as in South East Asia, which benefit from cheaper labour, less labour market regulation, and fast technological catch-up (World Economic Forum, 2019). The capacity to innovate, and to take up innovation produced elsewhere, is of prime importance, especially since, unlike cost-reduction strategies, innovation is, in principle, without bounds, and so is central to sustaining growth over the long-term.

However, concern has risen about a growing research and innovation divide, linked to geographical concentration, both within Member States and across the EU, of the most innovative firms and research centres. While concentration can result in positive externalities of research and innovation, the core areas are very often located in more developed regions 1 , so widening geographic disparities. 2  This research and innovation divide may be further fuelled by the ongoing process of digitalisation.  

Measuring innovation is widely recognised as challenging (OECD and Eurostat, 2018). 3 The most commonly used indicator, the number of patent applications, gives only an approximate measure of the real innovation activity because it captures only innovations registered at the European Patent Office. These relate mainly to technological innovation in industry, while many if not most innovations in services, which are often intangible, remain unpatented. 4 Nevertheless, though limited, patents provide a useful means of comparing performance of technological innovation across regions.

Over the period 2016-2017, 122 patent applications per million inhabitants were registered at the European Patent Office ( Map 2 ‑7 ). These show a distinct spatial pattern, regions with most applications being located mostly in the north-western Member States and in northern Italy. At the NUTS 3 level, Ludwigshafen in Germany, home to BASF, had the highest number (3 224 per million inhabitants in the period), followed by Erlangen, home to a major Siemens site (2 558) and Zuidoost-Noord-Brabant in the Netherlands (2 529), home to Philips. The degree of concentration suggests a regional innovation divide between the most advanced Member States and regions and the others.

Metropolitan areas tend to offer an environment that is particularly conducive to the development of new ideas, products and processes. A vast literature explains the reasons for this – the presence of a creative and skilled work force, specialised clusters of economic activity, universities and research institutes. 5  There are clear differences in patenting activity between metropolitan (metro) regions (around 167 applications per million inhabitants) and non-metro regions (around 58 per million inhabitants) ( Figure 2 ‑ 10 ). In quite a few metro regions, however, applications are less than in non-metro regions in the same country, indicating that not all metro regions offer a favourable innovation environment. Still, the distinct spatial pattern and concentration in metropolitan areas of patent applications are further indications of a research and innovation divide in the EU.

Figure 2‑10: Patent applications to the European Patent Office by type of region, 2016-17

Source: OECD REGPAT, DG REGIO calculations

Map 2‑7: Patent applications to the European Patent Office, average 2016-2017

Source: Eurostat, REGIO-GIS

A widely used indicator of innovation capacity, rather than performance, is expenditure on R&D relative to GDP, which is a measure of input into the innovation process, or the effort made, rather than of output. As in the case of patents, however, R&D expenditure is likely to underestimate innovation activity, particularly in sectors outside industry where non-technological and non-research-based innovation is common.

Expenditure on R&D in the EU amounted to 2.2% of GDP in 2019 ( Figure 2 ‑ 11 ) and increased only marginally over the previous two decades (from 1.8% of GDP in 2001). The expenditure rate increased in all Member States, except for Sweden and Finland, where it had already reached a high level in 2001, and Luxembourg 6 . Despite the overall increase, in most Member States the expenditure rates for the most part remain well below those in other highly developed economies, especially Japan (where expenditure was 3.2% of GDP in 2019) or the US (where it was 3.1%). There is also no evidence of convergence in rates within the EU, countries with comparatively low R&D expenditure in 2001 having the smallest increase in spending over the 2001-2019 period, suggesting a widening research and innovation divide between Member States.

R&D expenditure in the EU is highest in the north-western regions (at an average of 2.7% of GDP in 2019) and lowest in the east (1.3%) and south (1.4%). At the NUTS2 level, spending is highest, at over 7% of GDP, in Braunschweig and Stuttgart in Germany and Brabant Wallon in Belgium ( Map 2 ‑8 ).

In general, regions with the highest R&D expenditure tend to be the most developed and often include capital cities (Belgium and Germany are notable exceptions) ( Figure 2 ‑ 13 ). Of the 20 regions with the highest expenditure, 19 are more developed with GDP per head above the EU average, while two third of the 50 regions with the lowest expenditure are less developed with GDP per head below 75% of the average.

Figure 2‑12: Total expenditure on R&D as a % of GDP, 2001 and 2019

The 2001 figure for LU relates to 1999, for MT and HR to 2002.

Source: Eurostat [rd_e_gerdreg], DG REGIO calculations.

Map 2‑8: Total expenditure on R&D as a % of GDP, 2019

Source: Eurostat, REGIO-GIS

Figure 2‑13: Total expenditure on R&D as % of GDP, 2019

Note: BE (except BE10) and IE relate to 2017. FR relates to 2013.

Source: Eurostat [rd_e_gerdreg], DG REGIO calculations

In 2019, expenditure on R&D relative to GDP exceeded the Europe 2020 target of 3% only in a small number of NUTS 2 regions, accounting for just 12% of EU population ( Table 2 ‑7 ). These are all more developed regions in the north-west of the EU, except Dresden (Germany) which is a transition region. None of the less developed regions met the 3% target, with expenditure on average over 2 pp below the target.

Table 2‑7: Total R&D expenditure and the distance to the Europe 2020 target, EU-27 regions, 2019

BE (except BE10) and IE relate to 2017. FR relates to 2013.

* Includes only regions for which data are available

Source: Eurostat, DG REGIO calculations

The Regional Innovation Scoreboard

The Regional Innovation Scoreboard (RIS) 2021 highlights the key role innovation plays in regional development. 7 The RIS, an extension of the European Innovation Scoreboard (EIS), assesses the innovation performance of regions on the basis of a subset of the indicators included in the EIS. In 2021, it covers 215 regions in the EU 8 , plus 30 regions in Norway, Serbia, Switzerland and the UK.

The most innovative regions in the EU by this measure are Oberbayern (Germany), Hovedstaden (Denmark), Etelä-Suomi (Finland) and Stockholm ( Map 2 ‑ 9 ). Despite some regional variation within countries, the ranking of regions largely matches that of Member States, suggesting that indicator values at the regional level are affected by national characteristics. Most regional ‘innovation leaders’ are in countries which are also identified as ‘innovation leaders’ or as ‘strong innovators’, and almost all of the regional ‘moderate ‘and ‘modest’ innovators are in countries categorised in the same way. However, regional ‘pockets of excellence’ are evident in some ‘moderate innovator’ countries, including capital city regions in Czechia, Spain, and Lithuania as well as País Vasco in Spain, while some regions in ‘strong innovation’ countries lag behind.

Map 2‑9: Regional Innovation Scoreboard, 2021

Source: Regional Innovation Scoreboard 2021, REGIO-GIS

There is a close relationship between the level of development of regions and the innovation score ( Figure 2 ‑ 14 ). In 2021 about 70% of the population of less developed regions live in an ‘emerging innovator’ region, which is twice as much as in 2016. This indicates that a large number of less developed regions that used to be moderate innovators have become emerging innovators. Furthermore, none of them live in a ‘strong’ or a ‘leader innovator’ region. Accordingly, during the last five years, the less developed regions have fallen further behind in terms of innovation, rather than catching up with the other regions. At the other end of the spectrum, ‘leader innovators’ are almost exclusively in the group of more developed regions, with only 2% of the population in transition regions living in a region in this category in 2021. The majority of ‘strong innovators’ are also in the more developed regional group, with 84% of the population of these regions in 2021 living either in a ‘strong’ or a ‘leader innovator’ region, up from 70% in 2016.  

Figure 2‑14: Share of EU population by RIS category and level of development, 2016 and 2021

Note: In cases where the RIS score is only available at NUTS1 level, it is assumed that the same score applies to the latter’s constituent NUTS2 regions. Calculations for both years are based on 2021 population data and level of development classification.

Source: Regional Innovation Scoreboard 2021, DG REGIO

In general, the RIS confirms the wide diversity of EU regions in terms of innovation performance, so highlighting the fact that innovation has a strong regional dimension. Because of this, measures supporting innovation, including Cohesion policy programmes, need to take explicit account of the regional or local context when devising the kind of support to provide. As it is inherently place-based, the Smart Specialisation approach helps in this regard.

Expanding digitalisation

Digital technologies have the potential to boost more inclusive and sustainable growth by spurring innovation, generating efficiencies and improving services. 9  The current Commission has put the green and digital transition, the so-called ‘twin transition’, on top of the political agenda as the two trends that will shape Europe and its future. A goal of the EU is to boost the digital transformation of businesses by encouraging the take-up of three digital technologies 10 : cloud computing services, use of big data and Artificial Intelligence (AI). The objective is that 75% of European enterprises 11  will have taken these up by 2030.

The take-up of Cloud Computing in 2020 was greater than for the other two technologies ( Figure 2 ‑ 15 ), and the share of enterprises using it was twice as large as in 2014, a rate of increase, which, if it continues, will enable the 2030 target to be achieved. The take-up of big data and AI remains much smaller, which might be a result of these being newer and possibly less generally applicable from a business perspective.

The take-up of digital technologies in the EU masks pronounced differences between Member States. For each of the three technologies, businesses in less developed countries lag behind, the take-up being highest in highly developed Member States.

Figure 2‑15: EU enterprises take-up of digital technologies by Member State level of development, 2020

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], DG REGIO calculations

A similar pattern is seen for the take-up of e-commerce and e-business technologies ( Figure 2 ‑16 ). A sufficiently fast internet connection is required for such take-up. On average, some 46% of enterprises in the EU have broadband with a speed at least 100 Mb/s, but the figure is smaller in less developed Member States. Businesses in less developed Member States also lag behind in terms of the take-up of two specific e-business solutions, namely the use of business processes which are automatically linked to those of their suppliers or customers and the use of ERP (Enterprise Resource Planning) software to share information between different functional areas. The same is the case for e-commerce sales and online purchases. Both the share of enterprises with e-commerce sales of at least 1% of turnover and the share with online purchase of at least 1% of the total are smaller in less developed Member States, although the difference with other Member States is larger for the latter share

Figure 2‑16: EU enterprise take-up of e-commerce and e-business technologies by Member State level of development, 2020

All EU enterprises outside the financial sector with 10 or more persons employed (Eurostat code 10_C10_S951_XK) are covered.

The full definitions of the five indicators are: (1) the maximum contracted download speed of the fastest fixed line internet connection of at least 100 Mb/s; (2) enterprises with business processes automatically linked to those of their suppliers and/or customers; (3) enterprises with ERP software package to share information between different functional areas; (4) enterprises with e-commerce sales of at least 1% turnover; (5) enterprises purchasing at least 1% of the total online.

Data on ERP software relate to 2019; auto-linked business processes to 2017; online purchases to 2018, except AT, DE, IT, SE, EU27: 2017; EE, HR, SI: 2016; FI, MT: 2015.

Source: Eurostat [isoc_eb, isoc_ec], DG REGIO calculations

These results confirm that digitalisation may further fuel the research and innovation divide, at least between Member States. Given the increasing importance of digital technologies for enterprises to remain competitive, this is a cause for concern from a cohesion perspective. Since technology take-up is an important driver of economic convergence, less developed Member States risk falling further behind rather than catching-up, if their businesses do not innovate by adopting digitalisation. Moderately developed Member States may also see their capacity to compete diminished if they fail to do likewise, so risking falling into, or remaining in, a development trap (as indicated in Section 2.3 above).

2.4.2 Firm dynamics in EU regions

In 2018, the number of firms 15 with at least one employee – termed ‘employer firms’ here – was largest relative to population in Greece, Cyprus, Luxembourg, Slovenia (for which only national data are available) and most parts of Hungary and Estonia ( Map 2 ‑10 ). This may reflect a relative absence of large firms. Although the number of firms varies greatly between regions within Member States, the national context appears to be an important factor. In most countries, the number of firms relative to population is highest in the capital metro regions, except for France, Italy, Austria, and Spain. This is in part because many firms, especially large ones, have their headquarters there. The headquarter function also contributes to the higher number of employees per firm in the capital metro region 16 . In general, non-metro regions tend to have fewer employer firms per inhabitant than metro regions. 

Firms may locate in more urbanised areas to benefit from agglomeration economies, from ‘matching’, ‘sharing’ and ‘learning’ (Duranton and Puga, 2020). Cities tend to have larger labour markets, allowing better matching between labour demand and supply, and enable better sharing of inputs and infrastructure, while the fact that people work and live in close proximity facilitates learning from each other.

Map 2‑10: Active employer businesses per 1000 inhabitants, 2018

New enterprise creation is one of the main drivers of economic development and employment creation. New firms can help to open new sectors and higher value-added markets, so contributing to the structural transformation of an economy (Dent et al., 2016). They may also help to increase competitiveness by pushing incumbent enterprises to become more efficient.

In 2018, the number of newly-created employer firms relative to population tended to be higher in capital metro regions in both more developed and less developed Member States, with birth rates in Budapest and Tallinn being particularly high ( Figure 2 ‑ 18 ). Paris, Rome and Madrid are exceptions, birth rates being lower than in other metro regions in the countries concerned. In many sectors, firms operating in metro regions tend to face more competition because of the larger market and so a greater risk of being forced out of business if they are uncompetitive (Melitz and Ottaviano, 2008; Combes et al., 2012). High death rates, therefore, often go with high birth rates, as in Budapest and Tallinn, though death rates tend to be lower than birth rates, particularly in metro regions 17 .

High growth enterprises 18 play an important role in the economic growth of cities and regions through their contribution to productivity and innovation (Acs et al., 2008). In 2018, capital metro regions typically had the highest number of high growth firms per head. The only exceptions were Lisbon, Amsterdam, Rome, Paris and Vienna, but even there the number was still above the country average ( Figure 2 ‑ 19 ). In all Member States, the number was higher in metro regions than non-metro regions.

Figure 2‑18: Employer firm birth rates by type of region, 2018

BE: 2017

Source: Eurostat [bd_esize_r3], DG REGIO calculations

Figure 2‑19: Number of high growth firms by type of region, 2018

Data covers NACE sectors B-S (except K), apart from BE, CY, DE, EL, IE, LU, and SI, where they cover B-N (except K) and S95.

Source: Eurostat [bd_hgnace2_r3], DG REGIO calculations

2.4.3 Regional competitiveness in Europe

Regional competitiveness indicates the ability of a region to offer an attractive and sustainable environment for firms and residents to live and work in. Launched in 2010 and updated regularly since, the Regional Competitiveness Index (RCI) is designed to capture the different dimensions of competitiveness for EU NUTS 2 regions 19 . It allows regions to monitor and assess their development over time as compared with other regions 20 . The most recent edition of the RCI was published in 2019. It shows that more than 10 years after the crisis, there is still a clear north-west – south-east divide across the EU ( Map 2 ‑ 11 ).

Map 2 ‑ 11 : Regional Competitiveness Index scores, 2019  

In line with previous editions, the 2019 RCI shows a polycentric pattern with strong performance of most capital city regions and others with large cities, which benefit from agglomeration economies, better connectivity and high levels of human capital.

Capital city regions tend to be the most competitive, except in the Netherlands (where the capital city region is ranked second), Italy (where Lombardia is the most competitive region) and Germany ( Figure 2 ‑ 21 ).

Figure 2‑21: Distribution of regional RCI 2019 scores by Member State

Source: DG-REGIO

The gap between the capital city region and others is particularly wide in France, Spain, Portugal and many of the Eastern Member States. This can be a reason for concern as it puts pressure on the capital city region while possibly leaving resources under-used in other regions.

In general, higher levels of GDP per head are associated with higher levels of competitiveness. However, this relationship is stronger at lower levels of GDP – among more prosperous regions there is more variation in competitiveness.

References

Acs Z., Parsons, W., Tracy, S. (2008), High Impact Firms: Gazelles Revisited, Office of Advocacy, U.S. Small Business Administration.

Aghion, P. and Howitt, P. (1992) A Model of growth through Creative Destruction.  Econometrica  60(2), 323–51.

Altomonte, C., Di Mauro, F., Ottaviano, G., Rungi, A., & Vicard, V. (2012). Global value chains during the great trade collapse: a bullwhip effect? Firms in the international economy: Firm heterogeneity meets international business, 277-308.

Annoni, P., Kozovska, K. (2010), EU Regional Competitiveness Index 2010, EUR 24346, Publications Office of the European Union, Luxembourg.

Dijkstra, L., Annoni P., Kozovska, K. (2011), A new European Regional Competitiveness Index: theory, methods and findings, Working Papers 02/2011, Directorate General for Regional and Urban Policy, European Commission

Annoni, P., Dijkstra, L. (2017) Measuring and monitoring regional competitiveness in the European Union. In Huggins, R. and Thompson P. (Eds.): Handbook of Regions and Competitiveness - Contemporary Theories and Perspectives on Economic Development, Edward Elgar Publishing.

Annoni P. and Dijkstra L. (2019) The EU regional competitiveness index 2019. European Commission. Regional and Urban Policy Papers.

Annoni, P. and Catalina Rubianes, A. (2016) “Tree-based approaches for understanding growth patterns in the European regions”, REGION, 3(2), 23-45.

Annoni, P., de Dominicis, L., and Khabirpour, N., (2019), Location matters: A spatial econometric analysis of regional resilience in the European Union, Growth and Change, 50 (3): 824-55.

Combes, P.P, Duranton, G., Gobillon, L., Puga, D., Roux, S. (2012), The productivity advantages of large cities: distinguishing agglomeration from firm selection, Econometrica 80 (6), 2543-94.

Dijkstra L., Poelman, H. and Rodríguez-Pose, A. (2020), The geography of discontent. Regional Studies 54(6), 737-753.

Duranton, G. and Puga, D. (2020). The economics of urban density. Journal of Economic Perspectives 34(3), 3-26.

European Commission (2017), 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) EC Communication on a long-term vision for the EU’s rural areas – Towards stronger, connected, resilient and prosperous rural areas by 2040. COM (2021) 345 final.

European Commission (2021b) EC Communication on the 2030 Digital Compass: the European way for the Digital Decade. COM (2021) 118 final. 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.

Guzzo, F., and Gianelle C. (2021), Assessing Smart Specialisation: Governance, Publications Office of the European Union, Luxembourg, DOI: 10.2760/48092, JRC123984.

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, DOI: 10.2760/554632, JRC123918.

Iammarino, S., Rodriguez-Pose, A., 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. Luxembourg: Publications Office of the European Union, 2020.

Iammarino, S., Rodriguez-Pose, A., Storper, M. (2017), Why Regional Development Matters for Europe’s Economic Future, Working Papers 07/2017, Directorate General for Regional and Urban Policy, European Commission.    

Krugman, P. (2007). The ‘new’ economic geography: Where are we?. In Regional Integration in East Asia (pp. 23-34). Palgrave Macmillan, London.

Lesage, J.O., and Fischer, M. (2008) Spatial Growth Regressions: Model Specification, Estimation and Interpretation, Spatial Economic Analysis 3(3), 275-304.

Melitz, M.J., Ottaviano, G.I.P. (2008), ‘Market Size, Trade, and Productivity’, Review of Economic Studies, 75(1): 295-316.

OECD (2018), Hungary: Trade and Investment Statistical Note

OECD (2019), Luxembourg. OECD economic surveys.

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, https://doi.org/10.1787/9789264304604-en.

Pike A., Rodríguez-Pose A., Tomaney J. (2017) Shifting horizons in local and regional development, Regional Studies, 51:1, 46-57.

Prognos and CSIL (2021, forthcoming), Study on Prioritisation in Smart Specialisation Strategies in the EU

Schumpeter, Joseph A. (1942). Capitalism, Socialism and Democracy. New York: Harper & Bros.

Rodríguez-Pose A. (2020). The research and innovation divide in the EU and its economic consequences. European Commission R&I Paper Series. Working Paper 2020/03.

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).

World Economic Forum (2019), The Global Competitiveness Report 2019.

(1) See (Rodríguez-Pose, 2020) for an analysis of the economic consequences of the research and innovation divide in the EU.

(2) For example, European Commission (2020) concludes that increasing concentration of economic and innovative activities in capitals and metropolitan areas, on the one hand, and declining or peripheral areas on the other lead to negative developments in regions with low capacity to exploit innovation.

(3) This is particularly true in a sub-national context, which highlights the need to work on better territorial innovation data as mentioned for example in the Commission’s Communication on a long-term vision for rural areas (COM(2021) 345 final).

(4) This also holds for practices in primary production and organisational and social forms of innovation that can contribute to social capital.

(5) European Union and UN-HABITAT (2016).

(6)  The decrease in Luxembourg is linked to the fact that business R&D spending strongly decreased over the past decade. This is possibly related to the potentially large impact of the behaviour of few multinational companies on official business R&D statistics (see OECD 2019).

(7)

Regional Innovation Scoreboard 2021, available at:

https://ec.europa.eu/info/research-and-innovation/statistics/performance-indicators/regional-innovation-scoreboard_en

 

(8) All Member States are covered at the NUTS 2 level except for Austria, Belgium and France, which are covered at the NUTS1 level.

(9) OECD (2021) https://www.oecd.org/g20/topics/digitalisation-and-innovation/

(10) European Commission (2021b) EC Communication on the 2030 Digital Compass: the European way for the Digital Decade. COM (2021) 118 final.

(11)  All enterprises outside the financial sector with 10 persons or more employed (Eurostat code 10_C10_S951_XK).

(12) Available at this link: https://www.eib.org/en/publications-research/economics/surveys-data/eibis/index.htm .

(13) For further information on the methodology see: Delanote and Wruuck (2021), Regional Cohesion in Europe 2020-2021: Insights from the EIB Investment Survey, European Investment Bank, Luxembourg; available at this link: https://www.eib.org/attachments/publications/eibis_2020_regional_cohesion_en.pdf .

(14) Based on EIBIS 2020, available at this link: https://www.eib.org/en/publications/econ-eibis-2020-eu .

(15)

All firms in the business economy, as defined by NACE Rev.2, are covered, except insurance activities of holding companies (sector K642).

(16) Some caution is needed in interpreting this result. Some large enterprises may be composed of multiple local units, located in different regions, but with their employment registered in the head office often located in the capital city. This may inflate the number of employees counted as working there.

(17) This may reflect the fact that firms can cease operating without being formally closed down.

(18) High-growth enterprises are those which had at least 10 persons employed at the beginning of the period and where employment increased by over 10% a year over the subsequent three years.

(19) See Annoni and Kozovska (2010), Dijkstra, Annoni and Kozovska (2011), Annoni and Dijkstra (2017) and Annoni and Dijkstra (2019)

(20) All RCI editions are built on the same approach as the Global Competitiveness Index of the World Economic Forum.

EUROPEAN COMMISSION

Brussels, 4.2.2022

SWD(2022) 24 final

COMMISSION STAFF WORKING DOCUMENT

Cohesion in Europe towards 2050

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 8th Cohesion Report: Cohesion in Europe towards 2050

{COM(2022) 34 final}

3CHAPTER 3 A Greener, low-carbon Europe – PART 1

·The EU has adopted the European Green Deal with the goal to make the EU economy climate-neutral by 2050. This will require a rapid reduction of greenhouse gas (GHG) emissions, more investments in green technologies and protecting the natural environment.

·GHG emissions dropped by 24% between 1990 and 2019. This suggests the EU will meet its 2020 target of reducing GHG emissions by 20%. The new 2030 target as part of the ‘Fit for 55’ is a reduction of 55%. This will imply large reductions in emissions both within and outside the emissions trading scheme.

·Energy consumption has decreased significantly in the EU over the past decades. Nevertheless, the latest figures indicate that the 2020 energy efficiency target will be missed. The 2030 target is more ambitious and will require additional efforts.

·Renewable energy consumption in the EU rose steadily from 11% in 2006 to 19% in 2018, close to its 2020 target of 20%, but some Member States are lagging behind their 2020 national targets. The target of 40% by 2030 will require combined efforts for boosting production of renewable energy and reducing total energy consumption.

·Climate change affects a growing number of EU regions but the impact differs depending on their geography and the structure of their economy. Sectors such as tourism and agriculture are likely to be particularly affected.

·Only 40% of EU water bodies are in a good ecological state. Despite significant progress, several rural areas and less developed regions still need important investment in waste water treatment.

·The share of waste recovered increased from 46% in 2004 to 54% in 2018 in the EU. This helps to protect environment, recycle raw materials and recover energy. Nevertheless, recycling and incineration with energy recovery remain low in several Member States.

·The emissions of most major air pollutants have significantly shrunk in the EU. Exposure to air pollutants, however, is still high in many cities. One out of three city residents lives in a city where at least one of the air pollution thresholds is exceeded.

·Biodiversity loss and the degradation of ecosystem services continue in the EU across terrestrial, freshwater and marine ecosystems. Protecting and restoring biodiversity can help to improve the flow of ecosystem services and to mitigate climate change and its impacts. For example, investing in urban vegetation or wetlands can reduce the impact of heat waves and floods, provide more habitat for endangered species, reduce air and noise pollution and provide spaces for leisure, thus improving urban quality of life. In rural areas, fostering high-diversity landscapes can increase ecological connectivity and help species to adapt to climate change, while at the same time enhancing ecosystem services such as pollination, climate and water regulation, and erosion protection.

Contents

3    CHAPTER 3 A Greener, low-carbon Europe – PART 1    

3.1    Introduction    

3.2    EU climate action and the European Green Deal    

3.2.1    Reducing greenhouse gas emissions    

3.2.2    Increasing energy efficiency    

3.2.3    Boosting renewable energy    

3.2.4    Achieving low-carbon transport    

3.3    Reducing the impact of climate change    

3.3.1    The threat of floods from climate change    

3.3.2    Protecting Europe’s coasts against rising seas    

3.3.3    Infrastructure is also at risk    

3.3.4    Unevenly distributed impact of extreme temperature events    

3.4    Improving our environment    

3.4.1    More investment needed to improve water quality    

Figure 3‑1 Change in greenhouse gas emissions outside the Emissions Trading Scheme, 2005-2018 and Europe 2030 targets    

Figure 3‑2 Primary energy consumption, % change 2005-2019 and 2020 target    

Figure 3‑3 Share of renewables in gross final energy consumption, 2006, 2018 and 2020 target    

Figure 3‑4 GHG emissions in transport 1990 to 2019 and projections to 2035, Metric tons of carbon dioxide equivalent (MtCO2e), EU-27    

Figure 3‑5 Passenger travel by transport mode, 2019    

Figure 3‑6 Freight transport by mode, 2019    

Figure 3‑7 Estimated damage to coastlines in 2100 without and with adaptation measures (high emissions scenario)    

Map 3‑1 CO2 emissions from fossil fuels per head, NUTS2, 2018    

Map 3‑2 Change in total CO2 emissions from fossil fuels between 1990 and 2018    

Map 3‑3 Employees in ETS stationary installations, 2018    

Map 3‑4 Decarbonizing employment potential in coal regions under the EUCO3232.5 energy scenario    

Map 3‑5 Economic damage due to floods in 2100 under the 3°C warming scenario relative to the baseline    

Map 3‑6 Benefit to cost ratios of elevating dykes in NUTS2 regions under a moderate mitigation and high emissions scenario    

Map 3‑7 Expected annual damage to infrastructure due to inland flooding under a global warming scenario of 3°C    

Map 3‑8 Projected changes in human exposure to heat and cold waves events for a 3.0°C levels of global warming    

Map 3‑9 Urban wastewater receiving more stringent treatment    

Table 3‑1 Key EU climate and energy targets    

3.1Introduction

Recent extreme events such as deadly flooding in Germany and Belgium or uncontrollable forest fires in Greece illustrate the challenges faced by the EU in tackling the consequences of climate change. According to the last report from the Intergovernmental Panel on Climate Change (IPCC), almost the entire 1.1 degrees C of warming since the pre-industrial era is due to human activity 1 . The IPCC gives a 50% chance that a 1.5 degrees C warming could be reached before 2040. As a result, the negative impacts of climate change will become more frequent and more severe and all regions in the EU will be affected.

At the same time, the world is facing a massive extinction episode. This translates into a rapid fall in biodiversity which affects all parts of the world. At present, one million of the eight million species known on the planet are at risk of being lost due to the impact of human activities, including land and sea use changes, over-harvesting, climate change, pollution and invasive alien species. Biodiversity loss due to human pressures continues also in the EU, undermining the capacity of ecosystems to deliver benefits to humans. Yet, the quality of our environment is essential to human wellbeing and to maintain the provision of key ecosystem services such as climate regulation, flood protection, air and water quality, soil fertility, pollination and the production of food, fuel, fibre and medicines.

This chapter looks at the main trends related to climate change and environment. It assesses the extent to which the EU has or has not reached some of its key policy targets in the area. It also analyses how and to what extent EU regions are affected by the consequences of climate change and how they perform in preserving the quality of their environment.

3.2EU climate action and the European Green Deal 

Climate change and environmental degradation are the most challenging threats to living conditions in Europe and, indeed, in the world as a whole. In response, the EU has adopted the European Green Deal, a new growth strategy, with ambitious targets for resource-efficiency, competitiveness, greenhouse gas (GHG) emissions and inclusiveness. The goal is to make the EU economy and society climate-neutral by 2050 by cutting emissions, investing in green technologies and protecting the natural environment. A European Climate Law has been proposed by the Commission to make the goal legally binding 2 .

Over the past decades, the EU has adopted a series of targets for GHG emissions, energy efficiency and the share of renewables in energy consumption with the aim of achieving the transformation to a low carbon economy. The EU key targets were set in following frameworks:

-The 2020 climate and energy package adopted in 2007 and which aimed at a 20% cut in GHG emissions (from 1990 levels), a 20% share of renewables in energy consumption and a 20% improvement in energy efficiency by 2020;

-the 2030 climate and energy framework adopted in 2014 which upgraded the 2020 targets to respectively 40%, 32% and 32.5%;

-the European Green Deal, in which the Commission proposed an update of the 2030 target for reducing GHG emissions by 55% and raise the targets relative to renewables and energy efficiency to 40% and 36% respectively;

-the 2050 long-term strategy aiming at making the EU climate-neutral by 2050.  

Table 3 ‑1  summarises the most recent steps taken by the EU in setting climate and energy targets.

Table 3‑1 Key EU climate and energy targets

In July 2021, the European Commission adopted a series of legislative proposals setting out how it intends to achieve climate neutrality in the EU by 2050, including the intermediate target of an at least 55% net reduction in greenhouse gas emissions by 2030. The so-called ‘Fit for 55’ package combines the application of emissions trading to new sectors and a tightening of the existing EU Emissions Trading System, accelerating the use of renewable energy and greater energy efficiency, a faster roll-out of low emission transport modes, an alignment of taxation policies with the European Green Deal objectives, measures to prevent carbon leakage and tools to preserve and grow the EU’s natural carbon sinks. At the same time, a more transparent and dynamic governance process has been set up to help meet the 2030 targets and the EU’s international commitments under the Paris Agreement, involving an integrated monitoring system and reporting rules.

For these plans to succeed, action in all parts of the EU economy is needed, notably investment in environmentally-friendly technologies, targeted R&D and innovation, cleaner, cheaper and healthier forms of private and public transport, decarbonisation of the energy sector and improvements in the energy efficiency of buildings.

3.2.1Reducing greenhouse gas emissions 

Under the 2020 climate and energy package  3 , the EU committed to reducing GHG emissions by 20% by 2020 relative to 1990. The pursuit of this objective was supported by two instruments, the EU Emissions Trading System (ETS) and the Effort Sharing Decision (ESD).

The ETS is a market based tool for cutting emissions from large-scale power and industrial plants and aviation. It covers around 45% of EU total emissions and the target at the time was to reduce these emissions by 21% below the 2005 level by 2020. The ESD covers sectors not included in the EU ETS, such as transport, buildings, agriculture (non-CO2 emissions) and waste, which account for around 55% of EU emissions. Member States have committed to national 2020 targets, set according to their levels of development – from a 20% cut for the most developed countries to a maximum increase of 20% for the least developed relative to 2005. The ESD objective is to reduce emissions in the sectors it covers by 10%.

According to the latest figures available, the EU is likely to have met its 2020 target. Between 1990 and 2019, GHG emissions were reduced by 24%, while EU GDP grew by around 60%. Accordingly, the GHG emission intensity of the economy, defined as emissions relative to GDP, fell to less than half of the 1990 level 4 . EU-27 emissions covered by the ESD were 10% lower in 2019 than in 2005, so the 2020 target is likely to have been achieved. 

In 2014, the EU has enacted legislation to reduce emissions by at least 40% by 2030. National emission targets for ESD sectors have been revised to achieve a reduction of 30% by 2030 relative to 2005. These targets, enshrined in the Effort Sharing Regulation 5 , range from a reduction of 0 to 40%. Although all Member States have committed to not increasing emissions from their ESD sectors, they have risen in Malta, Latvia, Lithuania and Poland ( Figure 3 ‑1 ) 6 .

In 2018, levels of ESD emissions were lower than the 2030 target only in Greece, Hungary and Croatia and were well above it in a number of countries, either because the target was set at a high level (as in Luxembourg, Finland, Germany, and Belgium - a cut of 35% or more in all cases) or because emissions have been reduced only slightly (as in Ireland) or have increased (as in Bulgaria, Latvia, Lithuania , Malta and Poland).

Figure 3‑1 Change in greenhouse gas emissions outside the Emissions Trading Scheme, 2005-2018 and Europe 2030 targets

Source: Source: EUROSTAT, regulation(EU) 2018/842 and Commission Implementing Decision (EU) 2020/2126. Member States with actual changes above their target are highlighted.

Under the European Green Deal, as noted above, the EU launched the 2030 Climate Target Plan under which it set a more ambitious target of cutting emissions by at least 55% below 1990 levels by 2030, instead of 40%, on the way to becoming climate neutral by 2050.

GHG emissions per head vary substantially within countries. This is notably the case in Spain, Portugal, Germany, Greece, Bulgaria and Poland, where some regions are emission hotspots ( Map 3 ‑1 ) 7 . Many factors can explain differences in high emission levels, including, in particular, the level and composition of economic activity, the energy efficiency of production plants and buildings and the use of renewable energy as well as land use, climate and geography 8 .

Map 3‑1 CO2 emissions from fossil fuels per head, NUTS2, 2018

Map 3‑2 Change in total CO2 emissions from fossil fuels between 1990 and 2018

Between 1990 and 2018, GHG emissions were reduced in most EU regions but they significantly increased in some of them, notably in Cyprus, Ireland, Spain and Poland where they soared by more than 30% ( Map 3 ‑2 ).

Box 3‑1 Employment in EU ETS installations

3.2.2Increasing energy efficiency 

Increasing energy efficiency is key to protecting the environment, reducing GHG emissions and improving the quality of life. The EU has set ambitious targets for 2020 and 2030, focusing on the sectors where the potential for savings is the greatest, such as buildings.

As part of the 2020 climate and energy package, the objective set in 2007 was to improve energy efficiency by 20% by 2020 11 compared to the projections made at that time. To achieve this objective, Member States were asked to set their own indicative national energy efficiency targets 12 .

In 2018, the Energy Efficiency Directive 13 was amended to establish a target for 2030 of reducing EU energy consumption by at least 32.5% 14 . A reduction in energy consumption, however, does not necessarily signify an improvement in energy efficiency. The main determinants of energy use are GDP growth and the share of manufacturing in the economy. Changes in energy consumption, therefore, reflect not only changes in energy efficiency but also fluctuations in economic activity as well as changes in the structure of the economy.  

In 2019, primary and final energy consumption 15  had decreased by 9.7% and 5.5% respectively compared to their 2005 levels. However, primary and final energy consumption levels were respectively 3.0% and 2.6% above the 2020 targets and 19.9% and 16.3% above the 2030 targets. It is therefore likely that the EU will miss its 2020 targets while it is still far from the 2030 targets, implying a need for additional efforts to make the EU economy more energy efficient.

Progress in reducing energy use varies markedly between Member States. In 2018, only 11 of the 27 Member States had lowered primary energy consumption below their 2020 target and only 9 had reduced final consumption below the target. In a number of Member States, the reduction required to meet the targets was still considerable (Cyprus, Malta, Bulgaria and France in respect of primary energy consumption and Lithuania, Hungary, Malta and Slovakia in respect of final consumption) ( Figure 3 ‑2 ). 

Figure 3‑2 Primary energy consumption, % change 2005-2019 and 2020 target

Source: EUROSTAT.

3.2.3Boosting renewable energy 

Renewable sources play an increasing role in the production of energy in the EU. The share of renewables in gross final consumption of energy in the EU rose steadily from 11% in 2006 to 19% in 2018. In the 2020 climate and energy package of 2007, the objective was to raise this share to at least 20% by 2020, with a 10% share of renewables in transport. EU Member States have committed to meeting binding national targets for the share of renewables in energy consumption under the Renewable Energy Directive 16 of 2009. These range from 10% in Malta to 49% in Sweden. 

The 2030 climate and energy framework of 2014 set the target of reaching a share of 32% of renewables in energy consumption by 2030 but, as part of the ‘Fit for 55’ package, the Commission has proposed to increase this target to 40% 17 . For this target to be reached, the share of renewables would have to double compared to levels of 2018.

The share of renewables in energy consumption varies substantially across the EU. In 2018, it was over 40% in Finland and Latvia and close to 55% in Sweden ( Figure 3 ‑3 ). It is much smaller in other countries – below 10% in Malta, Luxemburg, Belgium and the Netherlands – though it has increased significantly in recent years. In 2018, 13 Member States had reached their national target set for 2020, Sweden, Estonia and Denmark exceeding it by over 5 percentage points. At the same time , some countries are still far from their target, like Belgium, France and Ireland where the share of renewables in 2018 was still less than 75% of the national 2020 target. For the Netherlands to meet their target, the share of renewables would need to have almost doubled between 2018 and 2020.

The capacity to produce renewable energy is closely linked to the geography of countries and regions. The production of wind energy is easier in coastal regions, like those of north-western Europe and Baltic Seas, the Atlantic and some Mediterranean coasts. The production of hydroelectricity requires suitable geo-physical features, while the potential for solar energy production is higher in southern European regions where there are many more days of sunshine. For instance, in 2018, the photovoltaic (solar panel) capacity per head in the EU was largest in Germany (590 watts per inhabitant), followed by the Netherlands (401) and Belgium (394) 18 . In Spain (197 watts per inhabitant) and Portugal (88), it was much less despite the potential production of electricity by this means being among the highest in the EU.

Figure 3‑3 Share of renewables in gross final energy consumption, 2006, 2018 and 2020 target

Source: EEA, EUROSTAT T2020_31.

Box 3‑2 Coal regions in transition

3.2.4Achieving low-carbon transport

After a sharp drop between 2008 and 2014 as a consequence of the 2008 economic crisis, GHG emissions from transport in the EU increased from 2014 to 2019 at rates similar to those in the period 1990-2008, just under 2% a year 22  ( Figure 3 ‑4 ). This implies that transport has not followed the general tendency for GHG emissions to decline in recent years. Its contribution to overall GHG emissions in the EU has therefore become more significant.

Projections suggest that GHG emissions from transport will decline relatively little over the next few years and will remain higher than in 1990, even with measures currently planned in Member States. Further action is therefore needed, particularly in road transport but also in aviation and shipping where demand is pushing emissions up in both absolute and relative terms. Emission reduction in all transport sub-sectors will need to be much more ambitious if the sector as a whole is to contribute its fair share to the goals set out in the European Green Deal.

Figure 3‑4 GHG emissions in transport 1990 to 2019 and projections to 2035 23 , Metric tons of carbon dioxide equivalent (MtCO2e), EU-27

Source: EEA.

The new EU Strategy on Sustainable and Smart Mobility 24  includes measures aimed at significantly reducing CO2 and polluting emissions in all modes of transport with the objective of reducing emissions by 90% by 2050. As part of the strategy, the Commission will foster the use of more sustainable transport modes such as rail and inland waterways.

The use of cars remains predominant for passenger travel and has even expanded slightly in recent years ( Figure 3 ‑5 ). In 2014, cars were used for 82.2% of inland travel and in 2019 for 82.8%. The share of passenger travel by train increased slightly from 7.7% to 8.0%, meaning that the share by buses, trams and trolleybuses fell from 10.1% to 9.2%. Cars account for less than 80% of passenger travel in only 5 Member States (Romania, Austria, Slovakia, Czechia and Hungary), while in Lithuania the share is over 90%.

Figure 3‑5 Passenger travel by transport mode, 2019

Source: EUROSTAT

These trends are a matter of concern as transport is responsible for almost a quarter of EU GHG emissions and is the main cause of air pollution in cities. Roads are by far the biggest emitter accounting for over 70% of all GHG emissions from transport in 2019. Emissions from road transport, however, are expected to diminish as it decarbonises faster than other modes. The largest increases are expected in aviation and international maritime transport, which are likely to account for a bigger share of transport emissions in coming years.

As in the case of passenger travel, most goods in the EU are transported by road (

Figure 3 ‑6 ). In 2019, 76.6% of freight was carried by road, up from 73.9% in 2014. In 8 countries, the share is over 80%, peaking at 98 and 99 in Greece and Ireland, respectively (Malta and Cyprus have no inland waterways and railways transport, therefore the share freight carried by road is 100%). At the other end of the scale, over half of freight is transported by rail or inland waterways in Bulgaria, Romania, Latvia and Lithuania.

Figure 3‑6 Freight transport by mode, 2019

Source: EUROSTAT.

3.3Reducing the impact of climate change

Climate change is recognised as the most serious threat to human societies around the world. Scientists see an increase in global temperature of 2°C relative to pre-industrial times as the threshold beyond which there is a very real risk that dangerous and possibly catastrophic changes in the global environment will occur. The past three decades have been warmer than any previous decade since records began in 1850. All parts of the world are potentially affected by the consequences of a rapid rise in temperature and the various climatic changes that are associated with it. Southern and part of Eastern Europe will experience more frequent and severe heat waves, forest fires and droughts. Already Northern Europe is becoming much wetter, with increasing risk of floods and extreme weather events, while coastal areas face the devastating consequences of rising sea levels from the melting of polar ice sheets and glaciers. The marine environment is also heavily affected by climate change and these impacts are projected to increase dramatically with severe implications for marine currents, vulnerable ecosystems such as coral reefs, biological resources and food chains.

The effects of climate change pose a major challenge for a growing number of EU regions. Around 7% of EU population live in areas at high risk of floods and over 9% live in areas where there are already over 120 days a year without rain. The exposure of EU regions to the damaging effects of climate change, however, differ widely between them, depending on their location but also the structure of their economies, given that sectors such as tourism or agriculture are likely to be particularly affected.

3.3.1The threat of floods from climate change

Flooding is a major cause of economic damage and loss of life in Europe and other parts of the world 25 . Despite considerable efforts to reduce the risk, the damage from floods appears to have increased over recent decades 26 . Ongoing climate change coupled with growing land take, especially in flood plains, is likely to further increase the social and economic damage in the EU.

The greater risk of floods for future societies makes it important to identify adaptation strategies that are effective and sustainable in economic, social and environmental terms. In particular, such strategies need to be assessed in terms not only of their effectiveness in reducing the potential damage, but also of the economic costs involved (e.g. for building and maintaining defences). According to recent estimates of the consequences of river flooding 27 , if no mitigation and adaptation measures are taken and the global temperature rises by 3°C by the end of the century, economic losses from river flooding will grow to nearly €50 billion a year, or over 6 times more than at present, and nearly three times as many people would be exposed to flooding 28 . The damaging effects are projected to increase with higher temperatures and economic growth in almost all EU regions, although countries in Eastern Europe would suffer larger losses relative to their GDP (

Map 3 ‑5 ). Limiting global warming to 1.5°C would halve economic losses and the population exposed to river flooding in the EU.

Map 3‑5 Economic damage due to floods in 2100 under the 3°C warming scenario relative to the baseline

Flood risk reduction strategies can substantially reduce the projected losses due to climate change. However, these strategies have different costs as well as benefits, as illustrated by a recent study which assessed four different approaches to limit the damages from coastal flooding 29 :

-Strengthening existing dyke systems, which is likely to have larger benefits than costs but tend to transfer risks downstream by stimulating further the development of human settlements and activities in risk zones behind flood barriers, which can result in catastrophic effects in case of failure; 

-Retention areas and dykes require large investment but can reduce the economic and human losses substantially; 

-Flood proofing buildings can markedly reduce losses with limited investment, but they do not prevent floods from happening and so can only partly prevent flood damage.

-Relocation can produce the largest benefits but tends to be the least cost-effective, though the costs involved vary substantially; it also tends to have low social acceptance.

Results suggest that reducing flood peaks using retention areas has strong potential for lowering the effects in a cost-efficient way in most EU countries (see section 3.3.2). Implementing such a strategy at EU level could reduce the economic damage and population exposed to flooding by over 70% by 2100. Moreover, retention areas have many additional benefits, such as restoring the natural functioning of flood plains and improving the ecosystem by improving nutrient removal, water filtration and the replenishment of groundwater reservoirs, providing fish-spawning habitat as well as opportunities for recreation and nature-based activities. Depending on local circumstances, other strategies than creating retention areas may be more suitable.

3.3.2Protecting Europe’s coasts against rising seas

Coastal zones are areas of high interest. Over 200 million people in the EU live within 50 km of the coast, stretching from the north-east Atlantic and the Baltic to the Mediterranean and Black Sea and in the EU outermost regions, and the evidence is that migration to coastal zones is continuing. Such areas in many cases are locations for major commercial activities and support diverse ecosystems with important habitats and sources of food.

Coastal zones are particularly vulnerable to climate change due to the combined effects of rising sea levels and the increasing frequency and intensity of storms, adding to already significant pressures from human activities. The mean global sea level has increased by 13-20 cm since pre-industrial times 30 and at an accelerating rate since the 1990s, the rise since 1950 being explicable by global warming 31 . This has already contributed to coastal erosion and made Europe’s coasts more susceptible to hazards. The continued rise in sea levels from global warming could result in unprecedented coastal flood losses in the EU unless additional coastal protection and measures to reduce risks are implemented.

This is affirmed by a recent study 32 , which assesses the costs and benefits of applying additional protection through dyke improvements. The largest amounts of damage are projected for France, Denmark, Italy, the Netherlands and Germany ( Figure 3 ‑7 ), though for some countries the potential damage is larger in relation to GDP, such as for Cyprus (5%), Greece (3%) and Denmark (2%). Appropriate adaptation measures are therefore needed to lessen these damaging effects. 

Figure 3‑7 Estimated damage to coastlines in 2100 without and with adaptation measures (high emissions scenario)  33

Source: Vousdoukas et al. (2020).

As argued by the authors, raising dyke levels along the EU coast could significantly reduce damages from flooding. The costs and benefits involved, however, vary markedly along coastal sections. The presence of human settlements makes investing in dykes economically beneficial, typically when population density exceeds 500 people per square km. In urbanised and major economic areas, the benefits of raising dykes tend to be several times the costs. Under a high emissions scenario, this would be the case for around 23% of the EU coastline. For the remainder, additional protection against coastal flooding is not needed or is not economically beneficial. This is either because natural barriers will provide sufficient protection against the rise in sea levels or because the costs of increasing dyke levels outweigh the benefits, such as in almost inhabited areas or along winding coastlines.

The analysis suggests that the average increase in the height of coastal defences needed where further protection is required is one meter under a high emissions scenario. In Slovenia, Latvia, Poland, Germany and the Netherlands it is well above this, and in Belgium it is over 2 meters. This implies that along many such areas, the shoreline might well become disconnected from hinterland areas.

When benefits and costs are aggregated across coastal sections of NUTS2 regions, the benefits to cost ratio (BCR) is highest in urban centres ( Map 3 ‑6 ). Adaptation brings large net economic benefits in the Ionian Islands (a BCR of 30 under a high emissions scenario), País Vasco (27), Aquitaine (16), Calabria (11.3), Basse-Normandie (14), Pays de la Loire (13), Puglia (11) and Alentejo (11).

Map 3‑6 Benefit to cost ratios of elevating dykes in NUTS2 regions under a moderate mitigation and high emissions scenario

Aggregating the results for coastal sections to the country level shows the Netherlands to have the highest BCR under a high emissions scenario (18), followed by Greece (12), France and Belgium (11 for each). By contrast, the BCR is low – though still over 1 – in Bulgaria, Finland, Romania, Croatia and Malta (3 or less in each case).

Investments in green infrastructure can also provide an efficient mean to enhance EU coastal defences against sea level rise. In particular, protecting and restoring costal ecosystems such as seagrass meadows and coral reefs can buffer the impacts of storms and help to reduce coastal erosion while bringing simultaneous benefits for biodiversity and natural resources.

3.3.3 Infrastructure is also at risk

The EU has an extensive transport network, with around 5 million km of paved roads, 0.5 million km of railways, over 2 400 airports, and almost 2 000 seaports, with a combined estimated value of around EUR 9 trillion. This is particularly susceptible to climate hazards and so is generally built to withstand the variations in temperature as indicated by historical observations, or according to regional standards of construction. However, rises in average temperatures or greater frequency of extreme weather events as a result of increased GHG emissions are likely to lead to increased economic losses.

A recent study 34  estimates the direct effects of flooding and heatwaves (two of the most damaging climate-related hazards according to a 20 year review by the UN Office for Disaster Risk Reduction 35 ) on the transport network in the EU, covering the modes of roads, railways, airports, and seaports. For each hazard, the effect is estimated as the change in expected annual damage for global warming levels of 1.5, 2, and 3°C relative to 1981–2010.

As would be expected, flood risk is concentrated in areas prone to flooding with high-value infrastructure, such as motorways and electrified railways. Some 95% of potential flood damage comes from roads and railways, with airports and seaports accounting for only 4%. The estimated cost of potential damage to railways is particularly high, at almost twice that of roads, reflecting the much higher costs of reconstruction and their location in lower lying terrain.

Nearly all regions in the EU are expected to experience increasing flood damage to their infrastructure as a results of climate change, particularly those prone to flooding in north-western and Eastern Europe, where the damage could in some case be over 6 times the present damage with global warming of 3°C. For most southern regions, damage to transport infrastructure from floods is projected to increase less dramatically, but could still be over twice as high as today ( Map 3 ‑7 ).

Map 3‑7 Expected annual damage to infrastructure due to inland flooding under a global warming scenario of 3°C

Road maintenance costs are also projected to rise in all EU regions as a result of more frequent spells of extreme heat. The most significantly affected countries in terms of additional cost to maintenance are Bulgaria, Poland, Greece, Ireland and Romania. Future risk can be alleviated by upgrading roads or doing more frequent maintenance.

Most of the increased maintenance costs are on tertiary and rural roads, which are generally managed by local authorities. Since their road maintenance budgets already tend to be constrained, damages from climate change could be particularly problematic for them.

The buckling of railway lines is also likely to occur more frequently with global warming so increasing maintenance costs. The biggest increases (of up to 10% with global warming by 3°C) are projected for regions in Germany and southern Spain, because of stress-free temperatures 36 being likely to be exceeded most often. Significant increases are also likely in regions in Belgium, France, Sweden, Finland, Poland and Czechia.

3.3.4Unevenly distributed impact of extreme temperature events

Extreme heat events are projected to happen more frequently and become more intense with climate change. The number of people exposed to heatwaves in the EU is projected to grow from 10 million/year (average 1981-2010) to nearly 300 million/year in a scenario with 3°C global average warming by the end of this century 37 . As a result, the number of fatalities from extreme heat could increase up to nearly 100,000 per year if no mitigation measures are taken, which is significantly higher than the current 2,750 annual deaths.

The exposure of the population to the risk of extreme temperature considerably varies across EU Member States and regions. Risks of being exposed to extreme heat should increase in southern Europe while milder winters could reduce significantly exposure to extreme cold, nearly 10-fold with 3°C global average warming by the end of this century ( Map 3 ‑8 ). Heatwaves, human exposure and fatalities are projected to increase everywhere in Europe but Cyprus, Greece, Malta and Spain could see a 40-fold increase in mortality from heatwaves if no adaptation and mitigation actions are taken.

Map 3‑8 Projected changes in human exposure to heat and cold waves events for a 3.0°C levels of global warming

In order to limit exposure and the increase of fatalities linked to extreme heat, a wide range of measures can be taken, including improved design and insulation of houses, schools and hospitals, education or early warning systems. This risk also needs to be taken into consideration in urban planning in order to minimise the urban heat island effect 38 . In that perspective, urban green infrastructure can play an important role, notably by increasing tree and vegetative cover, installing green or reflecting roofs, or using cool pavements (see section 3.5).

3.4Improving our environment

The EU faces unprecedented challenges of environmental sustainability, notably from accelerating biodiversity loss, degradation of ecosystem services, depletion of scarce resources and various forms of pollution, with the associated risk to human health and well-being.

As pointed by a series of recent scientific reports from the EEA, IPCC, IPBES, IRP and UN Environment 39 , current trends in production and consumption are fundamentally unsustainable. 

The EU has launched many policy initiatives to address these challenges, putting in place a broad range of legislation to reduce air, water and soil pollution. These have produced substantial benefits over recent decades. EU citizens enjoy some of the best water quality in the world and over 18% of the EU land area has been designated as protected for nature. As part of the European Green Deal, the European Commission adopted the EU Biodiversity Strategy 2030, which acknowledges nature restoration as a key contribution to both climate change mitigation and adaptation, the Farm to Fork Strategy 40 , the Zero pollution action plan 41 , the EU forest strategy 42 and the EU Soil Strategy 43 . The 8th Environmental Action Plan is designed to support the objectives of the European Green Deal and the transition towards a climate-neutral, resource-efficient and regenerative economy while improving the status of ecosystems.