ECONOMIC IMPACT OF MIGRATION

Statistical Briefings

Call: H2020-SC6-MIGRATION-2019

Work Programmes:

H2020-EU.3.6.1.1. The mechanisms to promote smart, sustainable and inclusive growth H2020-EU.3.6.1.2. Trusted organisations, practices, services and policies that are necessary to build resilient, inclusive, participatory, open and creative societies in Europe, in particular taking into account migration, integration and demographic change

Deliverable 4.2 – Statistical briefing for Austria on economic impact of TCNs in MATILDE re- gions

Authors: Birgit Aigner-Walder, Albert Luger, Rahel M. Schomaker with contributions from Ingrid Machold

Approved by Work Package Manager of WP4: Simone Baglioni, UNIPR Approved by Scientific Head: Andrea Membretti, UEF Approved by Project Coordinator: Jussi Laine, UEF

Version: 28.05.2021

DOI: 10.5281/zenodo.4817376

This document was produced under the terms and conditions of Grant Agreement No. 870831 for the European Commission. It does not necessary reflect the view of the European Union and in no way anticipates the Commission’s future policy in this area.

2

Introduction

This document presents the results of an assessment of the impact of migration and a measure- ment of the contribution provided by ‘third country nationals’ (TCNs) to the economic systems in the receiving contexts. In total, 10 statistical briefings for the MATILDE regions have been compiled, including the following countries: Austria, Bulgaria, Finland, Germany, Italy, Norway, Spain, Sweden, Turkey and the United Kingdom. Each document focuses on the following dimen- sions of economic development: economic growth, labour markets, innovation, and entrepre- neurship. Moreover, a short introduction on the relevance of TCNs in the respective country as well as the population development and structure are considered. The results of a literature review and a comparative analysis of the 10 statistical briefings are published separately.

Within the Matilde project there is a special focus on so-called ‘third country nationals’ (TCNs). MATILDE considers TCNs as Non EU citizens, who reside legally in the European Union (EU) and who are the target of EU integration policies. A TCN is “any person who is not a citizen of the European Union within the meaning of Art. 20(1) of TFEU and who is not a person enjoying the European Union right to free movement, as defined in Art. 2(5) of the Regulation (EU) 2016/399 (Schengen Borders Code).“ (European Commission 2020a). According to this definition, nationals of European Free Trade Association (EFTA) member countries Norway, Island, Liechtenstein and Switzerland are not considered to be TCNs. TCNs cover economic migrants, family migrants, stu- dents and researchers, highly skilled migrants, and forced migrants.

3

AUSTRIA Figure 1: Origin and Share of Total Population of Different Nationalities in Austria

Source: Statistics Austria (2020a), own illustration Third Country Nationals (TCNs) in Austria On January 1, 2020, a total of 1,486,223 persons with non-Austrian citizenship were living in Aus- In 1961, the number of migrants, i.e. foreign na- tria. This corresponded to a share of around tionals living in Austria, slightly exceeded the num- 16.7 % vs. 83.3 % of Austria's total population. ber of 100,000. This corresponded to a share of Among non-Austrian nationals, slightly more than about 1.4 % of the total population. In the second half (778,443 persons) came from EU-28 (incl. UK) half of the 1960s and at the beginning of the and EFTA countries, including a total of 1970s, the number and share of the foreign popu- 199,993 Germans, who formed the largest group lation increased relatively strongly due to the tar- of foreigners in Austria with a share of 13.5 % (or geted recruitment of workers from the former Yu- 2.2 % of total population). A total of 707,780 indi- goslavia and Turkey. In 1974, a temporary peak viduals (or 8.0 % of total population) were third was reached with about 311,700 foreign nationals country nationals (TCN), having a nationality that (4.1 % of the total population at that time). It was is neither Austrians, EU citizen nor EFTA citizens. not until the strong wave of immigration in the By comparison, the share of TCN in EU- early 1990s that the proportion of foreigners 27 (excl. UK) and EU-28 (incl. UK) is 5.7 % and jumped to over 8 %. After a brief stagnation in the 3.8 %, Austria is therefore lying above the average. second half of the 1990s, the number of foreign nationals in Austria has increased again since the Considering the nationalities of TCN, Serbians turn of the millennium, with the 10 % threshold (122,115 persons; share of 1.4 %) are making up for the share of foreigners being exceeded for the the largest nationality group ahead of Turks first time at the beginning of 2008. (117,607 persons; share of 1.3 %) and Bosnians (96,583 persons; share of 1.1 %). Rank four to eleven follow considerably further behind: Syria

4

(0.6 %), Afghanistan (0.5 %), Russia (0.5 %), Ko- 20 years as their number increased by 235,374. sovo, Northern Macedonia (both 0.3 %), Iran, Like total population urbanization growth as a re- China, Iraq (each 0.2 %). In total almost 4 of 5 TCNs sult of TCN immigration only takes place in a few living in Austria belong to these nationalities. Fig- regions, i.e. cities and metropolitan areas (see Fig- ure 1 gives an overview of the origin and the share ure 2 (d)). of the different nationalities in Austria. Figure 2: Population Development (Δ 2002 to 2020) Population Development (a)

The number of inhabitants in Austria has increased in the past and will continue to grow in the future. As of January 1, 2020, there were 8,901,064 per- sons registered in Austria. This corresponds to an increase of 837,424 citizens (+10.4 %) within about 20 years since 2002 (before eastward en- largement of EU in 2004). The main cause of pop- ulation growth is immigration to Austria (EU-28, (b) EFTA & TCN), as the birth rate is only slightly higher than the death rate of the domestic population (see Figure 2 (a)).

Analyzing regional units (districts), the strong dis- persion is striking. The median Austrian popula- tion development is 0,0 %, i.e. half of the districts has grown, and the other half has shrunk. The Aus- trian population is relocating to the cities and met- (c) ropolitan areas of Vienna, Linz, Salzburg, Inns- bruck and Bregenz along the east-west axis, whereas in the south there is an internal migration from rural areas to cities. The cities and metropol- itan areas of Graz and Klagenfurt, where growth and stagnation have occurred, are notable excep- tions to this trend. The total number of Austrians living in Austria slightly increased by 81.462 inhab- (d) itants (+1.1 %; see Figure 2 (b)).

EU (incl. UK) and EFTA citizens are the main factor behind population growth in Austria (+520,588 inh., +201,9 %). From a regional per- spective the range is +51,8 % (minimum) to +396,4 % (maximum). Similarly to the Austrians, this group of population lives in or close to cities. Germans (2.2 %), Romanians (1.4 %) and Hungari- Source: Statistics Austria (2020a), own illustration ans (1.0 % of total population) account for the largest share (see Figure 2 (c)). Summing up, both, domestic and foreign popula- tion groups tend to (re-)locate in cities and metro- TCN account for approx. 8 % at the beginning of politan regions. This is at the expense of rural ar- 2020. Immigrants from Serbia (1.4 %), Turkey eas which are experiencing an overall decline in (1.3 %) and Bosnia (1.1 %) are the largest national population. In particular, this is the case for the groups. TCNs became more relevant in the past southernmost province of Carinthia, which will

5

shrink in the future (up to current population fore- which spans to northern Burgenland, the popula- casts). tion will also increase in the regions of the provin- cial capitals of Graz, Salzburg, Innsbruck and Bre- From a regional perspective, the areas with the genz, as well as in the central Upper Austrian re- strongest decline in population are located in the gion of Linz-Wels and in the Carinthian cities of federal states of Lower Austria (Waldviertel), Klagenfurt and Villach. In these regions, the popu- Styria (Mur-Mürzfurche) and Carinthia (except lation will grow steadily until 2040. The main rea- metropolitan area of Klagenfurt-Villach). These re- son for this is the strong external immigration, as gions belong to the periphery and have a poor eco- well as mostly positive balances of internal migra- nomic structure. They suffer from higher depopu- tion and birth surpluses. Besides urbanization lation and birth rate deficits. In a total of four dis- tourist regions in the west, i.e. Tyrol and Vorarl- tricts, the population is projected to decline by berg, also tend to grow (see Figure 3 (a)). 10 % or more by 2040, namely in Wolfsberg (- 11.6 %), Spittal an der Drau (-11.4 %), Hermagor (- Vienna and its surrounding areas expect the high- 13.4 %) and Murau (-14.6 %; see Figure 3 (a)). est growth rate of domestic-born population whereas projected shrinking regions correspond Figure 3: Projections of population growth (Δ 2020 to 2040) (a) to those which experienced a decline in the past (see Figure 2 (b)). Again Murau has the greatest decline of 14.6 % until 2040, followed by the pe- ripheral regions of Carinthia, Salzburg and Styria (see Figure 3 (b))

Similar to the domestic-born population, the for- eign-born population will settle primarily in or near the cities of Vienna, Linz, Graz, Salzburg and Innsbruck, whereas it performs below-average but (b) positive in rural areas (except for Waidhofen an der Thaya; see Figure 3 (c)).

Population Structure

While the debate about changing demographic structures in most cases turns around population ageing and the sustainability of pension systems, urbanization etc. a focus must be taken on the rel- evance of labor supply, i.e. the population aged (c) 15 to 64 years, in industrialized economies and nations. From an economic perspective, labor (done by human beings) is an essential element in the production of goods and services. The term ‘la- bor force’ comprises all of those who work for gain, whether as employees, employers, or as self- employed, and it includes the unemployed who are seeking work.

Source: ÖROK (2019), own illustration Starting from 5.924.377 people in this age group as of Jan. 1, 2020, the maximum will be reached at Up to 2040 predominantly metropolitan areas will the beginning of 2021 with 5.932.642 per- grow. In addition to the Vienna metropolitan area, sons (+0.14 %). Thereafter, the labor force poten- tial will decline, as more people will move from

6

working age to retirement age in the 2020s than of domestic born labor force. There are great dis- will be added at younger ages or through immigra- parities between the federal states as there is a tion. As a result, the number of 15-64 year-olds in tendency of Austrians relocating to Vienna (only - 2040 will already be -4.7 % lower than in 2020, at 6.8 %) and its metropolitan areas in Lower Austria 5.644.119 inhabitants (-280.258 inh.). (only -9.6 %). The decline in the other provinces ranges from -11.9 % (Vorarlberg) to -21.6 % Further analysis on a regional level reveals that ex- (Carinthia; see Figure 4 (b)). The situation is differ- cept for Vienna (+2.9 %) the potential labor force ent for the population born abroad as it is pro- will decline in each federal state till 2040, if the jected to rise by +263.648 inh. (+18.6 %) all over projections for the population born in Austria and Austria. Again there are striking regional differ- abroad are considered (see Figure 4 (a)-(c)). The ences as abroad born population rises by almost southern provinces Carinthia (-15.7 %) and Styria one third (+29.2 %) in Burgenland followed by (-9.3 %) suffer the highest losses followed by Bur- Lower Austria, Upper Austria, Tyrol, Styria (range genland (-7.3 %). All other states, Upper Austria, from +22.8 % to +20.1 %) and Carinthia (17.9 %) Tyrol (both -5.4 %), Lower Austria (-4.5 %) and and Salzburg (+17.8 %). Vienna (+15.4 %) and Vor- Vorarlberg (-5.2 %) operate at the same level. arlberg (+15.1 %) are expecting the lowest growth Figure 4: Projections of Working Age Population rates (see Figure 4 (b)). (15 to 64 years; Δ 2020 to 2040) (a) These figures underline the high relevance of mi- gration, especially for rural regions, concerning the development of the working age population. Without immigration, its population between 15 and 64 years would shrink even more dramati- cally, being a challenge from an economic point of view (see Figure 4 (a)).

The total-age dependency ratio is a measure of the age structure of the population. It relates the num- (b) ber of individuals who are likely to be “dependent” on the support of others for their daily living – the young (up to 14 years old) and the elderly (65 plus years old) – to the number of those indi- viduals who are, being working age from 15 to 64 years old, capable of providing this support.

The total-age dependency ratio in is 50,2 %. The dependency rate is usually lower in (c) metropolitan areas and cities than in rural regions. The key factors here are prospering economy and low child ratios. Most metropolitan regions in Aus- tria are also characterized by relatively low rates (e.g. Vienna: 45,1 %). High dependency ratios are found in regions where the share of children and the elderly (65 years and older) is above average. These regions are mainly found in Carinthia and Burgenland (both 54.7 %). As a result of ageing so- Source: Statistics Austria (2020b), own illustration cieties, the total-age dependency ratios will rise in the future by 17 percentage points (pp) to 67,2 % Differentiated by the country of birth, the projec- in 2040. tions show total decline of 543.906 inh. (-12.1 %)

7

Figure 5: Total-Age Dependency Ratio (Δ 2020 to 2040) Education and (Un)Employment (a) The tertiary education rate of Austria's domestic population is 31.2 %, which is only slightly above that of the total population (31.1 %). Compared to the EU-27 (27.9 %) Austria has an above-average tertiary education rate. Austria ranks 14th within the EU-27, ahead of its neighboring countries Slo- venia (29.3 %), Germany (26.0 %) and Hungary (22.5 %). Luxembourg (41.0 %), Ireland (40.7 %) (b) and Cyprus (40.0 %) rank at the top end, while the Czech Republic (21.6 %), Italy (17.4 %) and Roma- nia (16.0 %) rank at the bottom.

Across the world, educational attainment is signif- icantly higher in cities1 than in towns and suburbs2, which in turn is higher than in rural areas3. In rural areas in Austria, only 25.6 % of individuals have university degrees, i.e. tertiary education

(c) (ISCED2011 levels 5 to 8), compared to 29.8 % in towns & suburbs and 38.8 % in cities. In contrast, secondary degrees are more common in less densely populated areas. While less than 40.9 % of city residents have secondary, i.e. upper second- ary and post-secondary non-tertiary education (ISCED2011 levels 3 to 4) attainment level, 51,0 % and 57.4 % of residents in towns & suburbs and rural areas do. Primary educational attainment

Source: Statistics Austria (2020b), own illustration level has a comparable low relevance as in cities approx. only 1 out of 5 residents in Austria belongs Considering the nine Austrian regions, the in- to this group, in towns and suburbs and rural areas crease varies from 8.8 pp (Vienna) to almost the share is even lower. 25 pp, with highest ratios in Burgenland (23.7 pp) and Carinthia (24.6 pp) in 2040; see Figure 5 (a)). In cities, where most of the Non EU-27 citizens Figure 5 (b) and (c) differentiate between Austrian live, more than 40.6 % have only primary educa- and abroad born total-age dependency ratios. tion, i.e. less than primary, primary and lower sec- Without immigration the dependency ratios ondary education (ISCED levels 0 to 2, see Table would evolve worse than with because of the high 1). This share is even higher in towns & sub- share of domestic age group 65 years or older urbs (48.5 %) and rural areas (48.1 %). On con- (range from 13.4 pp to 30.8 pp vs. 5.1 pp (Tyrol) to trary, only 25.6 % of Non EU-27 citizens have a de- 9.9 pp (Carinthia)). gree from university, but this value is still much higher than in suburban (14.8 %) and rural

1 A city is a local administrative unit (LAU) where at least 50 density of at least 1 500 inhabitants per km² and collectively % of the population lives in one or more urban centres (Eu- a minimum population of 50 000 inhabitants after gap-filling rostat 2020b). (Eurostat 2020c; Eurostat 2020d). 2 Towns & suburbs are areas where less than 50 % of the 3 A rural area is an area where more than 50 % of its popula- population lives in rural grid cells and less than 50 % of the tion lives in ‘rural grid cells’, i.e. grid cells that are not identi- population lives in urban centres, i.e. a cluster of contiguous fied as urban centres or as urban clusters (Eurostat 2020e; grid cells of 1 km² (excluding diagonals) with a population Eurostat 2020f).

8

(17.3 %) regions. Differences of educational at- likely to be unemployed than people without a vo- tainment level between Austrians and EU-27 resi- cational qualification. This correlation can also be dents range within a narrow band from 1.0 pp seen for Austria: Those with primary education are to 4.9 pp and are therefore comparable. ). In to- most affected by unemployment (10.0 %) and this tal, EU-27 citizens (38.3 %) residing in Austria have situation has remained constant for years. Con- a higher tertiary education rate than domestic versely, people with tertiary education have the population (31.2 %) whereas Non EU-27 rank last lowest unemployment rate (2.8 %), those with (21.7 %). secondary a value of 3.6 %. Similar results are ob- served at the European level (EU-27; see Figure 7). Table 1: Working Age Population by citizenship, degree of urbanization and educational attainment level 2019 Figure 6: Unemployment rate 2019 (20 to 64 years)

Educational Attainment Level (ISCED Urbanisation Citizenship Citizenship Primary Secondary Tertiary Austrians 16.6% 42.9% 40.6% EU-27 14.7% 39.8% 45.5%

Cities Cities Non EU-27 40.6% 33.8% 25.6% Total 20.2% 40.9% 38.8% Austrians 16.5% 52.6% 31.0% EU-27 18.8% 49.2% 32.0% Source: Eurostat (2020g), own illustration Non EU-27 48.5% 36.7% 14.8% suburbs

Towns & Towns Figure 7: Unemployment rate by educational attainment Total 19.2% 51.0% 29.8% level 2019 (25 to 64 years; ISCED 2011) Austrians 16.5% 58.0% 25.5% EU-27 13.9% 55.6% 30.4% 14% Rural areas areas Non EU-27 48.1% 34.6% 17.3% 12.5% Total 17.0% 57.4% 25.6% 12% Source: Eurostat (2020a), own illustration 10.0% 10% The high relevance of education for employment and income, is also valid for Austria and its regions. 8% In total, about 188,000 (45.7 % female) persons 6% 5.2% were unemployed in Austria in 2019 according to 4.0% 4% 3.6% the labor force survey (LFS). Compared to the pre- 2.8% vious year, this corresponds to a reduction of - 2% 6.9 %. In relation to the labor force potential, this 0% results in an unemployment rate of 4.3 % (female: Primary Secondary Tertiary 4.2 %; male: 4.4%). The comparative value for the EU-27 is 6.6 % (female: 6.9 %; male: 6.3 %; age EU-27 Austria group 20 to 64 years). Vienna has by far the high- est unemployment rate at 9.0 % (female: 9.1 %; Source: Eurostat (2020g), own illustration male: 9.9 %), followed by Lower Austria (3.8 %; fe- In contrast, the higher the educational attainment male: 3.7 %; male: 3.9 %), Carinthia (3.7 %; fe- the higher is the employment rate, i.e. 84.7 % for male: 4.4 %; male: 3.0 %) and Burgenland (3.6 %). tertiary, 76.1 % for secondary and 48.2 % for pri- Tyrol (1.9 %; 2.0 % female; 1.7 % male) and Salz- mary level. From an immigrational point of view, burg (2.2 %) are the provinces with the lowest un- this observation holds for other citizenships as employment rates (see Figure 6). well, as the risk of unemployment decreases by In general, there is a correlation between the level the highest educational attainment. But Austrians of education and unemployment, according to have the highest employment rate for each educa- which the well-educated are significantly less tional attainment level (86.5 %; 76.7 %; 48.3 %),

9

followed by EU-27 citizens (81,9 %; 75.1 %; and that of Non EU-27 is almost five times as high 52.9 %) and Non EU-27 citizens (65,5 %; 69.5 %; (24.4 %). In total, almost one half (48.7 %) of early 45.8 %); the only exception is “less than primary, school leavers, is jobless, but no relying state- primary and lower secondary education” of Aus- ments can be made in regard to immigrational sta- trian and EU-27 citizens (48,3 % vs. 52.9 %; see Fig- tus because of missing data (Eurostat 2020i). ure 8). Income and Gross Domestic Product Figure 8: Employment rate by educational attainment level and citizenship 2019 (25 to 64 years; ISCED 2011) As for the employment rate, the educational at- tainment level has a positive impact on mean and 100% median4 equivalized net income, i.e. the higher 86.5% 81.9% the educational attainment level the higher the in- 80% 75.1%76.7% come. 69.5% 65.5% This explains why the mean equivalized income is 60% 52.9% always slightly higher for each educational level. 48.3%45.8% Working persons in Austria at primary level earn 40% less (mean: € 22,412 or median: € 21.190) than secondary (€ 27,750 or € 26,358) and tertiary lev- 20% els (€ 34.051 or € 30.743); see Figure 9).

0% Figure 9: Mean and median equivalized net income by edu- Primary Secondary Tertiary cational attainment level 2018 (18 to 64 years; ISCED 2011) 40,000 EU-27 Austrians Non EU-27 35,000 34,051 30,743 Source: Eurostat (2020h), own illustration 30,000 27,750 26,358 The abovementioned findings foster the theoreti- 25,000 22,412 21,290 cal conclusions by Becker (1964) that the educa- 20,000 tional level hast a significant influence on the pro- 15,000 fessional career and the risk of unemployment of human beings. 10,000 5,000 This theory and the empirical evidence can also be applied to the “Early Leavers from Education and 0 Primary Secondary Tertiary Training”. This population group refers to persons aged 18 to 24 who have completed at lower sec- Mean equivalised net income ondary education and are not involved in further Median equivalised net income education or training. From a fiscal point of view, this specific group is of particular relevance as they Source: Eurostat (2020j), own illustration are more likely to be (long-term) unemployed. As working Austrians have a higher educational at- tainment level than EU-27 and Non EU-27 this In 2019, 10.2 % of the 18-24 year olds in the EU-27 could be one reason (amongst others) that their were part of this group (male: 11.9 %; female: mean (€ 30,361 vs. € 25,778 and € 21,064) and 8.4 %). Austria ranges with a value of 7,8% below median (€ 27,997 vs. € 21,600 and € 19,474) . The share of EU-27 citizens the European average equivalized net income is higher (see Figure 10). residing in Austria is two times as high (12,3 %)

4The median is more robust, i.e. less sensitive to outliers than the mean.

10

Figure 10: Mean and median equivalized net income by factors like seniority, promotions to better-paid broad group of citizenship 2018 (18 to 64 years) occupations as well as job stability also have to be 35,000 taken into account. These factors also cause an im- 30,361 provement in an individual's firm-specific human 30,000 27,997 25,778 capital (training on the job) as the median age of 25,000 foreign workers and therefore the length of time 21,600 21,064 19,474 20,000 within the company is lower than domestic. How- ever, these assumptions cannot be verified in the 15,000 context of this analysis due to a lack of available 10,000 data.

5,000 The indicator ‘people at risk of poverty or social exclusion’ corresponds to the sum of persons who 0 are: at risk of poverty after social transfers, se- Austria EU-27 Non EU-27 verely materially deprived or living in households Mean equivalised net income with very low work intensity5. As Non-EU 28 citi- Median equivalised net income zens have a higher risk of unemployment and they earn less than domestic residents and EU-28 al- Source: Eurostat (2020k), own illustration most 2 of 5 persons (38,4 %) belong to this group A further analysis and decomposition of wage dif- though the value of EU-27 is quite high as well ferentials between domestic and foreign workers (29.9 %). Austrians face the lowest risk of social ex- done by Lehmer and Ludsteck (2013) reveals that clusion and poverty (13.1 %; see Table 2).

Table 2: Population aged 18 and over by risk of poverty, material deprivation, work intensity 2019 Dimensions of Poverty and Social Exclusion Citizenship Material Non Risk of Poverty Work Intensity Austria EU-28 Deprivation EU-28 At risk Severe Very low 0.3% 0.6% 2.9% At risk Severe Not very low 0.4% 1.4% 3.1% At risk Non severe Very low 1.6% 4.7% 7.4% At risk Non severe Not very low 7.9% 20.5% 16.1% Not at risk Severe Very low 0.3% 0.4% 1.0% Not at risk Severe Not very low 0.5% 0.4% 4.4% Not at risk Non severe Very low 2.1% 1.9% 3.5% People at Risk of Poverty or Social Exclusion 13.1% 29.9% 38.4% Not at risk Non severe Not very low 86.8% 69.9% 61.6% People at no Risk of Poverty or Social Exclusion 86.8% 69.9% 61.6% Total 100.0% 100.0% 100.0% Source: Eurostat (2020l), own illustration

A way to calculate gross domestic product (GDP) the rent earned by those who lease their land or is, to sum up, all the income earned by factors of structures to firms; and dividends, the profits paid production from firms in the economy—the wages to the shareholders, the owners of the firms’ phys- earned by labor; the interest paid to those who ical capital. This is a valid measure because the lend their savings to firms and the government; money firms earn by selling goods and services

5 Persons are considered to be at risk of poverty after threshold, which is set at 60 % of the national median social transfers, if they have an equivalized disposable equivalized disposable income. People living in house- income below the risk-of-poverty threshold, which is holds with very low work intensity are those aged 0-59 set at 60 % of the national median equivalized dispos- living in households where the adults (aged 18-59) able income. Persons are considered to be at risk of work 20% or less of their total work potential during poverty after social transfers, if they have an equival- the past year (Eurostat, 2020m). ized disposable income below the risk-of-poverty

11

must go somewhere; whatever is not paid as Austria is Salzburg (154 %); it is over 64/47/44 pp wages, interest or rent is profit. Ultimately, profits richer than Burgenland/Lower Austria/Carinthia. are paid out to shareholders as dividends Vienna ranks second (150 %). The western regions (Krugman and Wells 2017). Tyrol (136 %) and Vorarlberg (143 %) are also lo- cated at the top whereas Upper Austria (131 %) In 2019, compensation of employees (wages and and Styria (118 %) are in the midfield (Eurostat salaries plus employers’ social contributions) was 2020n). the largest income component of EU-27 GDP ac- counting for 47.5 % and 48.1 % in the euro area. Table 3: Nominal regional GDP (2019) and Austria lies above European (both EU-27 and Euro real growth rate (Δ 2002 to 2019) area) average. Taxes on production and imports (less subsidies) accounted for 12.3 % (EU 11.9 %; Euro area 11.5 %). Gross operating surplus and prices prices prices capita capita Province Province mixed income accounted for 39.2 % of GDP in Aus- €) GDP (in GDP (in €) €) per GDP (in (Δ 2002 to 2019) 2002 to (Δ 2019) 2002 to (Δ Real growth rate Real growth rate Real tria, 40.6 % of GDP for the EU-27 and 40.4 % in the marketat current marketat current Euro area. Austrian GDP at current prices Burgenland 9,273 28.6% 31,600 22.1% amounted to approx. € 397.6 billion (+3.2%) in Carinthia 21,506 26.5% 38,300 26.7% Lower 44,780 61,706 30.7% 36,700 21.5% 2019 and GDP per inhabitant equaled € Austria (Eurostat 2020m). Upper 68,380 32.8% 46,000 24.6% Austria Within 20 years Austria’s gross domestic product Salzburg 29,852 34.4% 53,600 26.2% (GDP) increased by 29.2 %. In terms of regional Styria 50,831 31.1% 40,800 26.0% GDP per capita, all provinces also recorded posi- Tyrol 36,383 35.2% 48,100 23.1% tive real growth, ranging from 1.1 % in Vienna to Vienna 100,348 21.3% 52,700 1.1% Vorarlberg 19,162 37.8% 48,400 24.9% 26.7 % in Carinthia since 2002 (Austria: 18.9%). Extra-Regio 135 -12.8% . . The highest regional GDP per capita at current Total 397,575 29.2% 44,800 18.9% prices occurred in Salzburg (approx. € 53,600), Source: Statistics Austria (2020c) and Statistics Austria (2020d), own illustration succeeded by Vienna (approx. € 52,700). As in pre- vious years, the eastern and southern provinces Economic structure and entrepreneurship were below the national level of € 44,800 (Burgen- In general, a distinction is made between three dif- land: € 31,600; Lower Austria: € 36,700; Carinthia: ferent sectors of the economy, the primary (agri- € 38,300, Styria: € 40,800; see Table 3). culture and forestry), secondary (manufacturing) Provinces with a high share of TCNs have a high and tertiary (services) sectors. In recent decades, GDP per capita and vice versa. On the one hand, the Austrian economy has undergone a structural prosperous regions are more attractive to mi- transformation, which is why another important grants; on the other hand, these regions benefit sector has developed alongside the established from a better availability of labor supply, i.e. sectors - the information and communications greater number of working age people. In this con- technology sector (ICT sector), which is composed text, one speaks of a so-called cross-fertilization, which results from the interaction of these two factors.

Expressing GDP in PPS (purchasing power stand- ards) eliminates differences in price levels be- tween countries or regions. This approach is suita- ble in cases where there is a significant difference in levels of prosperity, as can be observed among Austria provinces. The most prosperous region in

12

of subsectors of the production6 and services7 sec- an increase of +13.3 % compared to 2008. Almost tors. three quarters (72.6 %) are employed in the ser- vice sector (3.4 million persons), where employ- In total, approx. 4.6 million people were employed ment has increased by +17.6 % since 2008. Em- (employees, marginal employed, freelance, self- ployment in manufacturing evolved below aver- employed) in . This corresponds to age (+8.3 %) and though primary sector increased

Table 4: Employment by economic activity 2019 Employed EU-28 Economic Activity persons in Share Δ 08-19 Austrians & TCN 2019 EFTA Primary Sector 34.9 0.8% 35.7% 55.9% 36.6% 7.5% Agriculture, forestry and fishing 34.9 0.8% 35.7% 55.9% 36.6% 7.5% Secondary Sector 1064.1 22.9% 8.3% 78.4% 12.9% 8.7% Mining and Quarrying 6.3 0.1% -0.9% 88.5% 7.7% 3.8% Manufacturing 685.3 14.8% 7.1% 80.7% 11.5% 7.9% Electricity, gas, steam and air conditioning sup- 26.9 0.6% -2.9% 95.4% 2.9% 1.7% ply Water supply, sewerage, waste management 18.6 0.4% 23.9% 80.9% 10.6% 8.5% Construction 327.0 7.0% 11.2% 71.8% 17.1% 11.1% ICT manufacturing 13,328 0.4% 0.2% 82.5% 10.5% 7.0% Tertiary Sector 3370.7 72.6% 17.6% 77.7% 14.0% 8.3% Wholesale and retail trade, repair of motor ve- 652.9 14.1% 7.0% 80.4% 10.9% 8.7% hicles Transportation and storage 226.0 4.9% 5.5% 73.2% 15.1% 11.7% Accommodation and food service activities 341.9 7.4% 28.3% 58.4% 25.7% 16.0% Information and communication 120.8 2.6% 33.8% 83.4% 11.3% 5.4% Financial and insurance activities 125.3 2.7% -7.7% 90.2% 6.2% 3.7% Real estate activities 57.3 1.2% 1.9% 78.1% 11.3% 10.6% Professional, scientific and technical activities 261.6 5.6% 33.0% 83.5% 11.4% 5.1% Administrative and support service activities 279.0 6.0% 28.1% 57.0% 23.8% 19.2% Public administration and defense 593.4 12.8% 11.2% 94.8% 3.2% 2.0% Education 134.4 2.9% 34.6% 77.8% 14.8% 7.4% Human health and social work activities 380.1 8.2% 49.3% 71.1% 23.9% 5.1% Arts, entertainment and recreation 63.4 1.4% 33.1% 74.9% 16.9% 8.2% Other service activities 126.2 2.7% 2.5% 81.7% 10.4% 7.9% Activities of households as employers, undiffer- 7.6 0.2% -58.8% 59.6% 26.1% 14.3% entiated goods- and services Activities of extraterritorial organizations and 0.8 0.0% 23.4% 48.2% 31.9% 20.0% bodies ICT services 84,869 2.3% 40.6% 83.7% 10.7% 6.9% Others 174.5 3.8% -22.3% 97.6% 1.7% 0.7% Military and civil servants 4.7 0.1% -37.4% 100.0% 0.0% 0.0% Parents karenz with working relationship 72.6 1.6% -26.0% 100.0% 0.0% 0.0% Others 97.2 2.1% -18.3% 95.7% 3.1% 1.2% Total 4644.2 100.0% 13.3% 78.4% 13.5% 8.1% ICT total 98,197 2.6% 33.3% 83.5% 10.6% 6.9% Source: BMAFJ (2021)

6 Manufacture of electronic components and boards, Manu- 7 Wholesale of information and communication equipment, facture of computers and peripheral equipment, Manufac- Software publishing, Telecommunications, Computer pro- ture of communication equipment, Manufacture of con- gramming, consultancy and related activities, Data pro- sumer electronics, Manufacture of magnetic and optical me- cessing, hosting and related activities; web portals, Repair of dia (Eurostat 2020o). computers and communication equipment (Eurostat 2020o).

13

by 35.7 % it is of very low importance, i.e. only “Activities of extraterritorial organizations and 0.8 % of total employment. Approx. 4 % belong to bodies” (31.9 %) has very low absolute im- the category ‘Others’ (“military and civil servants”, portance (only 770 persons in total or 0.02 %). In “parents karenz” with “working relationship”, “Electricity, gas, steam and air conditioning sup- “others”). 2.6 % work in the ICT, which performed ply” (2.9 %), “Public administration and de- very well (+33,3 %) in the past years mainly caused fense” (3.2 %), “Financial and insurance activi- by services (+40.6 %) rather than manufacturing ties” (6.2 %), “Other service activities” (10.4 %), (+0.2 %; see Table 4). “Water supply, sewerage, waste manage- ment” (10.6 %), “Wholesale and retail trade, re- A more detailed examination of the different sec- pair of motor vehicles” (10.9 %), “Real estate ac- tors of the economy reveals that there have also tivities”, “Information and communication” (both been significant shifts within the sectors: the only 11.3 %), “Professional, scientific and technical ac- declines in employment were recorded in the sec- tivities” (11.4 %) and “Manufacturing” (11.5 %)” tors "Activities of households as employers, undif- EU-28 & EFTA workers have minor relevance (be- ferentiated goods- and services" (-58.8 %), " Fi- low 13.5 %). Summing up EU-28 & EFTA citizens nancial and insurance activities" (-7.7 %), "Electric- are underrepresented in the secondary sector (ex- ity, gas, steam and air conditioning supply" (- cept “Construction”) whereas their skill and 2.9 -0.9 %). "Hu- %) and “Mining and Quarrying” ( knowledge is more relevant in the tertiary sector. man health and social work activities" (+49.3 %), "Education" (+34.6 %), “Professional, scientific TCNs preferably work in “Administrative and sup- and technical activities” (+33.0 %), "Administra- port service activities” (19,2 %), “Accommodation tive and support service activities" (+28.1 %), "In- and food service activities” (16.0 %)”, “Activities of formation and communication" (+33.8 %) as well households as employers, undifferentiated goods- as " Accommodation and food service activi- and services” (14.3 %)), “Transportation and Stor- ties" (+28.3 %), on the other hand, are among age” (11.7 %), “Construction” (11.1 %)”, “Real es- those economic sections with the largest absolute tate activities” (10.6 %) and “Water supply, sew- growth. age, waste management” (8.5 %). From a sectoral point of view, TCNs have almost the same rele- Further analysis by citizenship shows that almost vance in the secondary (8.7 %) and tertiary sector 4 out of 5 blue- and white-collar workers of the (8.3 %; difference of only 0.4 pp). Overall, the ac- secondary (78,4 %) and tertiary sector (77.7 %) are count for 8.1 % of all working people as they are Austrians. The same holds for EU-27 & EFTA slightly underrepresented in the primary sector (12.9 % and 14.0 %) and TCN (8.7) % and 8.3 %) (7.5 %) workers. Almost the same ratio applies to the ICT sector (83.5 %/10.6 %/6.9 %). These findings correlate with EU-28 & EFTA citi- zens except for high-tech industry and knowledge- EU-28 & EFTA citizens are overrepresented (above intensive services8 where a higher educational at- 13.5 %) in “Activities of households as employers, tainment level is necessary. The greatest variation undifferentiated goods- and services” (26.1 %), is between “Human health and social work activi- “Accommodation and food service activi- ties” (18.8 pp), “Arts, entertainment and recrea- ties” (25.7 %), “Human health and social work ac- tion” (8.7 pp) and “Information and communica- tivities” (23.9 %), “Administrative and support ser- tion” (5.9 pp). As a large number of “Information vice activities” (23.8 %), “Arts, entertainment and and Communication” belongs to the ICT sector (16.9 (17.1 %), recreation” %), “Construction” they are also less engaged there (6.9 % vs. 10.6 %). “Transportation and Storage” (15.1 %) and “Edu- cation” (14.8 %).

8 High-tech industry and knowledge-intensive services products traded broken down by technological inten- describe manufacturing and services industries or sity (Eurostat 2020p).

14

Lower Austria (84,5 %), Tyrol (83.5 %) and Salz- 0.9 pp (Vienna) whereas the greatest gap is in burg (81.3 %) have the highest share of Austrians neighboring Burgenland (20.6 pp; see Figure working in the secondary sector whereas Vi- 11 (b)). enna (64.1 %), Carinthia (70.3 %) and Vorarl- Research and Innovation berg (73.0 %) have the lowest share. The number of TCNs exceeds the number of EU-28 &EFTA in Following the neoclassical and endogenous Vorarlberg (14.3 % vs. 12.7 %) and almost in Salz- growth theories, technological advance is believed burg (10.3 % vs. 10.5 %); in the remaining federal to be one of the major drivers of economic provinces EU-28 &EFTA outweighs TCN (see Figure growth. From this perspective there is a growing 11 (a)). interest to investigate the link between re-

Figure 11: Employment by economic sector and citizenship search & development (R&D), innovation, entre- 2018 preneurship and economic growth achieved by (a) Secondary Sector human capital from abroad (EU-27 and Non EU- 0% 50% 100% 27).

Burgenland 80.3% 25.6% 4.1% A well-known indicator provided to measure

Carinthia 70.3% 10.2% 5.3% achievements of countries or regions in R&D is GERD, i.e. regional/national gross domestic ex- Lower… 84.5% 12.7% 7.7% penditure on R&D as a percentage of GDP. GERD Upper… 79.6% 10.8% 7.8% is estimated at 3.18 % for Austria (2019) which is a Salzburg 81.3% 10.5% 10.3% slight increase of 0.13 percentage points to 2017. The total amount of research expenditures was Styria 79.2% 12.0% 4.5% € 12.7 billion, the largest share (47.6 % or 6.04 bil- Tyrol 83.5% 11.3% 8.4% lion) was financed by Austrian companies. 24.6 % Vorarlberg 73.0% 12.7% 14.3% (or € 3.12 billion) were financed by the federal Vienna 64.1% 19.6% 16.4% government. More than € 0,75 billion (or 6.0 %) were funded indirectly via research grants. 4.3 % Austria 78.4% 12.9% 8.7% (or € 0,55 billion) was provided by the local gov- Austria EU-28 & EFTA TCN ernments, 15.9 % (or € 2.02 billion) by foreign (b) Tertiary Sector countries and 1.6 % (or € 0,2 billion) was financed 0% 50% 100% by other sources. The majority of the funding from Burgenland 74.2% 23.2% 2.6% abroad came from foreign companies whose sub- sidiary companies conduct research in Austria and Carinthia 84.1% 11.4% 4.5% includes returns from EU research programs and Lower… 80.7% 13.3% 6.0% initiatives as well (Eurostat 2020q). Upper… 82.6% 10.8% 6.6% Data on regional level (NUTS 2 or federal states) is Salzburg 76.2% 15.1% 8.7% published with a two year time lag. Styria was the Styria 82.3% 12.9% 4.8% province with the highest ratio in 2017 (4.87 %).

Tyrol 76.2% 16.6% 7.2% Vienna followed in second place with 3.60 % just ahead of Upper Austria with 3.46 %. All other Vorarlberg 74.7% 16.4% 9.0% provinces were well behind and below the all-Aus- Vienna 71.0% 14.9% 14.0% trian ratio of 3.05 %. Burgenland ranked last

Austria 77.7% 14.0% 8.3% (0.85 %), with Salzburg (1.59 %), Vorarlberg Austria EU-28 & EFTA TCN (1.75 %) and Lower Austria (1.80 %) only slightly

Source: BMFAJ (2021), own illustration better. Tyrol’s ratio is 2.88 %, just behind Carinthia (2.94 %, see Table 5). An analysis of the tertiary sector shows that EU- 28 & EFTA citizens have a higher share of at least

15

Table 5: Intramural R&D expenditure as percentage of In accordance to the research & development gross domestic product (GDP) 2017 (R&D) funding structure almost two thirds (63 0% Expenditures Province % of GDP or 29,948 full time equivalent (FTE)) of the re- (€ mio.) Burgenland 74.39 0.85% searchers work in the business enterprise sector. Carinthia 584.18 2.94% More than one of four researchers (28.4 % or Lower Austria 1,047.41 1.80% 13,513 FTE) works in the higher education sector, Upper Austria 2,191.16 3.46% 7.7 % (or 3,665 FTE) in the government sector and Salzburg 443.02 1.59% 0.8 % (or 395 FTE) in the private non-profit sector. Styria 2,320.33 4.87% This makes a total of 47,521 FTE or 131,032 per- Tyrol 968.24 2.88% sons. Further analysis on educational attainment Vorarlberg 317.88 1.75% Vienna 3,343.17 3.60% level reveals that most researchers (65.6 %) have Total 11,289.78 3.05% tertiary education but not doctoral or equivalent Source: Statistics Austria (2020f) and Eurostat (2020r), own level; only 27.0 % belong to the latter whereas illustration 7.4 % have less than primary, primary, secondary and post-secondary non-tertiary education (see Figure 12).

Figure 12: R&D researchers (full-time equivalent (FTE)) by sectors of performance and educational attainment level 2017 (ISCED2011) 25,000

21,820

20,000

15,000

10,000

7,089 6,250 4,852 5,000 3,276 2,087 1,538 174 31 182 182 0 40 Business enterprise sector Government sector Higher education sector Private non-profit sector Less than primary, primary, secondary and post-secondary non-tertiary education (levels 0-4) Tertiary education excluding doctoral or equivalent level (levels 5-7) Doctoral or equivalent level

Source: Eurostat (2020s), own illustration Within the period from 2016 to 2018, 62.6 3% of 2018, around 14.9 % of total turnover concen- enterprises in Austria developed and/or intro- trated on these product innovations. 6.3 % of total duced new or improved goods or services, imple- turnover accounted for market novelties, 8.6 % mented new or improved business processes or fell upon product innovations which were not new conducted innovation activities. 34.6 % of enter- to the market, but only new to the enterprise. Al- prises introduced new or improved goods or ser- together, 23% of the enterprises introduced at vices onto the market in the reference period. In least one market novelty onto their market be- tween 2016 and 2018. In 55 % of the enterprises

16

new or improved business processes were intro- populated by more than two million people in the duced.9 second half of the 2020s. In general, urban and suburban regions are expected to grow in Austria 9.8 billion were spent in total on innovation ac- € at the expense of rural regions. tivities. 83 % of those expenditures fell upon re- search and development (R&D) and 17 % onto fur- 1.49 million immigrants live in Austria, or 16.7 % of ther innovation activities, such as the acquisition the population. According to forecasts, their num- of highly developed machinery or external ber will rise to 2.22 million by 2040 (+49.0 %). The knowledge respectively training or design activi- share of foreign-born inhabitants thus increases to ties for innovations. 22 % by 2040. The majority will come from EU & EFTA member states, but TCNs became more 18 % of all enterprises cooperated with other en- relevant in the recent past. terprises or institutions on their innovation activi- ties. Universities or other higher education institu- On January 1, 2020, a total of 1,486,223 persons tions were the cooperation partners reported with non-Austrian citizenship were living in Aus- prominently; 61 % of all innovators with innova- tria. This corresponds to a share of around 16.7 % tion cooperation collaborated with this type of in- of Austria's total population. Among non-Austrian stitution. nationals, slightly more than half came from EU- 28 (incl. UK) and EFTA countries. 8.0 % of total The “lack of skilled employees” as well as the set- population were third country nationals (TCN), be- rities within the own enter- ting of “different prio ing individuals who are neither Austrians, EU citi- prise” were those hampering factors for innova- zens nor EFTA citizens. By comparison, the share tions which were quoted most often by enter- of TCN in EU-27 (excl. UK) and EU-28 (incl. UK) is (Statis- prises with a degree of “high importance” 5.7 % and 3.8 %, therefore Austria lying above the tics Austria 2020g). average. As the group of migrants (EU-27 or Non-EU-27) liv- In general, the pattern of education of the EU-28 ing in Austria has a lower share of tertiary educa- and non-EU countries differs substantially from tion it is therefore likely that they are underrepre- that of Austrians: The foreign population of third sented in Austrian R&D personnel. countries is disproportionately represented in pri- Conclusion mary and secondary education, whereas they are underrepresented in secondary and tertiary edu- This Statistical Briefing examines the economic cation levels (ISCED2011). Austria and the EU-27 and spatial impact of immigration with special re- do not differ much in terms of educational levels. gard to third-country nationals (TCN) in Austria. To a certain degree, this explains why inclusion in By 2040, Austria will have 9.4 million inhabitants. the labor market is lower among migrants than Without immigration from abroad, the population among nationals. In 2019, the employment rate would then be 7.2 million, i.e. less than today. Aus- for domestic workers (15 to 64 years old) was tria's population will grow as a result of immigra- 76.7 %, for people from the EU-28 countries 75,1% tion solely. According to the forecasts, Austria's and third countries 69.5%. population will grow at a slightly higher rate in the future than in the recent past. The immigration will have the most impact in Vienna, which will be

9 These business process innovations included; New or im- other administrative operations; new or improved business proved methods for producing goods or providing services; practices for organizing procedures or external relations; new new or improved logistics, delivery or distribution methods; or improved methods of organizing work responsibility, deci- new or improved methods for information processing or com- sion making or human resource management; new or im- munication; new or improved methods for accounting or proved marketing methods.

17

While domestic workers in 2019 were mostly em- stagnating, accompanied by declining population ployed in manufacturing and trade (both ap- figures. This development is also typical for rural prox. 14 %), health care and social work (ap- regions of other provinces (Salzkammergut, prox. 10 %) and construction (approx. 8 %), the Lungau, etc.). same applies to migrant workers, with a majority Provinces with a high share of TCNs mostly have a of migrant workers working in manufacturing (ap- high GDP per capita and vice versa. This finding prox. 16 %) and trade (approx. 15 %), closely fol- correlates with the population density. The higher lowed by accommodation and food services (ap- it is, the more economically attractive a regional prox. 12 %) and construction (approx. 11 %). entity is for both groups, domestic (Austrian) and As working Austrians and EU-28 citizens have a foreign population (EU-28 & TCN). higher educational attainment level than Non EU- Governments need to invest in education on the 28, this could be one reason that their income is one hand and in research and development on the higher. As the TCNs immigrants are on average other, to enable long-term prosperity and growth. younger than Austrians, immigration increases the Currently the national gross domestic expendi- ratio of workers to retirees, therewith shielding tures on R&D (as a percentage of GDP) are esti- the economy against the long-term financing chal- mated at 3.18 % (2019) which is a slight increase lenge of social security. of 0.13 percentage points to 2017. But there is The below average regional GDP growth rate, es- again a striking regional variation from densely pecially in rural provinces like Burgenland and populated areas with a high number of innovative Carinthia, is indirectly linked to immigration from enterprises (e.g. Vienna and Styria) to less densely abroad but preferably to internal migration from populated provinces (e.g. Lower Austria and Bur- rural areas to cities (especially to the cities and genland). metropolitan areas of Vienna and Graz). These Summing up, immigrants play a pivotal role in re- emigrations tend to reinforce the erosion of the viving and growing many rural areas communities, economic basis, as they have a negative impact on and with the appropriate policies, they could play the potential workforce and the economic attrac- an even more significant role in sustaining them. tiveness of the region. Economic performance is

18

Bibliography and source references

Becker, Gary Stanley (1964): Human Capital: A theoretical Eurostat (2020m): GDP and main components (output, ex- and empirical analysis, with special reference to education. penditure and income), https://appsso.eurostat.ec.eu- 3rd edition. New York. ropa.eu/nui/show.do?dataset=nama_10_gdp&lang=en

BMAFJ (Federal Ministry of Labour) (2021): NACE classes (4th Eurostat (2020n): Regional gross domestic product (PPS per level), https://www.dnet.at/bali/Nace4EN.aspx inhabitant in % of the EU27 (from 2020) average) by NUTS 2 regions, https://ec.europa.eu/eurostat/data- Eurostat (2020a): Population by educational attainment browser/view/tgs00006/default/table?lang=en level, sex, age, citizenship and degree of urbanisation (%), https://ec.europa.eu/eurostat/data- Eurostat (2020o): ICT sector, https://ec.europa.eu/euro- browser/view/edat_lfs_9916/default/table?lang=en stat/cache/metadata/en/isoc_se_esms.htm

Eurostat (2020b): City, Eurostat (2020p): High-tech industry and knowledge-inten- https://ec.europa.eu/eurostat/statistics- sive services, https://ec.europa.eu/euro- explained/index.php?title=Glossary:City stat/cache/metadata/en/htec_esms.htm

Eurostat (2020c): Town or suburb, https://ec.europa.eu/eu- Eurostat (2020q): GERD by sector of performance and source rostat/statistics-explained/index.php?title=Glos- of funds, http://appsso.eurostat.ec.eu- sary:Town_or_suburb ropa.eu/nui/show.do?dataset=rd_e_gerdfund&lang

Eurostat (2020d): Urban centre, Eurostat (2020r): Gross domestic product (GDP) at current https://ec.europa.eu/eurostat/statistics- market prices by NUTS 2 regions, https://appsso.euro- explained/index.php?title=Glossary:Urban_centre stat.ec.europa.eu/nui/show.do?da- taset=nama_10r_2gdp&lang=en Eurostat (2020e): Rural area, https://ec.europa.eu/euro- stat/statistics-explained/index.php/Glossary:Rural_area Eurostat (2020s): Total R&D personnel and researchers by sectors of performance, educational attainment level Eurostat (2020f): Rural grid cell, https://ec.europa.eu/euro- (ISCED2011) and sex, https://ec.europa.eu/eurostat/data- stat/statistics-explained/index.php?title=Glossary:Ru- browser/view/rd_p_persqual11/default/table?lang=en ral_grid_cell Krugman, Paul, Wells, Robin (2017): Economics, Fifth Edition, Eurostat (2020g): Unemployment rates by sex, age, educa- New York. tional attainment level and NUTS 2 regions (%), https://appsso.eurostat.ec.europa.eu/nui/show.do?da- Lehmer, Florian; Ludsteck, Johannes (2013): Lohnanpassung taset=lfst_r_lfu3rt&lang=en von Ausländern am deutschen Arbeitsmarkt: Das Herkunftsland ist von hoher Bedeutung, IAB-Kurzbericht, Eurostat (2020h): Employment rates by sex, age, educational 01/2013), Nuremberg. attainment level, citizenship and NUTS 2 regions, https://ec.europa.eu/eurostat/data- ÖROK (Austrian Conference on Spatial Planning) (2019): browser/view/tepsr_wc140/default/table?lang=en Excel-Tabellen mit den Ergebnissen für die Bundesländer, NUTS3-Regionen und Bezirke (Prognoseregionen). Eurostat (2020i): Early leavers from education and training https://www.oerok.gv.at/fileadmin/user_upload/Bilder/2.Re by sex and citizenship, http://appsso.eurostat.ec.eu- iter- ropa.eu/nui/show.do?dataset=edat_lfse_01&lang=en Raum_u._Region/2.Daten_und_Grundlagen/Bevoelkerungsp rognosen/Prognose_2018/BevPrognose_2018_Ergebnisse.xl Eurostat (2020j): Mean and median income by educational sx attainment level - EU-SILC survey, http://appsso.euro- stat.ec.europa.eu/nui/show.do?lang=en&dataset=ilc_di08 Statistics Austria (2020a): Population at the beginning of the year 2002 to 2020 (harmonised territorial boundaries Eurostat (2020k): Mean and median income by broad group 1.1.2020), http://statcube.at/statcube/opendata- of citizenship (population aged 18 and over), base?id=debevstandjbab2002 https://appsso.eurostat.ec.europa.eu/nui/show.do?da- taset=ilc_di15&lang=en Statistics Austria (2020b): Beginning-of-the-year population by native country from 2020 to 2101, http://stat- Eurostat (2020l): Persons aged 18 and over by risk of pov- cube.at/statcube/opendatabase?id=debevjahresanfgebland erty, material deprivation, work intensity of the household and most frequent activity status of the person - intersec- Statistics Austria (2020c): Gross regional product 2019 by tions of Europe 2020 poverty target indicators, Länder (NUTS 2): overview, https://www.statis- https://appsso.eurostat.ec.europa.eu/nui/show.do?da- tik.at/wcm/idc/idcplg?IdcService=GET_NATIVE_FILE&Revi- taset=ilc_pees02&lang=en sionSelectionMethod=LatestReleased&dDocName=029465

19

Statistics Austria (2020d): Gross regional product in volume terms by Länder, https://www.statis- tik.at/wcm/idc/idcplg?IdcService=GET_NATIVE_FILE&Revi- sionSelectionMethod=LatestReleased&dDocName=115603

Statistics Austria (2020e): Disposable income of private households 2012-2019 by Länder (NUTS 2): per capita, https://www.statis- tik.at/wcm/idc/idcplg?IdcService=GET_NATIVE_FILE&Revi- sionSelectionMethod=LatestReleased&dDocName=029543

Statistics Austria (2020f): Gross Regional Product (GRP), Gross domestic expenditure for R&D and regional research intensities 2017, https://www.statis- tik.at/wcm/idc/idcplg?IdcService=GET_NATIVE_FILE&Revi- sionSelectionMethod=LatestReleased&dDocName=103648

Statistics Austria (2020g): 63% der österreichischen Unternehmen sind innovationsaktiv https://www.statistik.at/web_de/statistiken/energie_umwel t_innovation_mobilitaet/forschung_und_innovation/123638 .html

20

ECONOMIC IMPACT OF MIGRATION

Statistical Briefings Bulgaria

Call: H2020-SC6-MIGRATION-2019

Work Programmes:

H2020-EU.3.6.1.1. The mechanisms to promote smart, sustainable and inclusive growth H2020-EU.3.6.1.2. Trusted organisations, practices, services and policies that are necessary to build resilient, inclusive, participatory, open and creative societies in Europe, in particular taking into account migration, integration and demographic change

Deliverable 4.2 – Statistical briefing for Bulgaria on economic impact of TCNs in MATILDE re- gions

Authors: Birgit Aigner-Walder, Albert Luger, Rahel M. Schomaker with contributions from Anna Krasteva and Evelina Staikova-Mileva

Approved by Work Package Manager of WP4: Simone Baglioni, UNIPR Approved by Scientific Head: Andrea Membretti, UEF Approved by Project Coordinator: Jussi Laine, UEF

Version: 28.05.2021

DOI: 10.5281/zenodo.4817376

This document was produced under the terms and conditions of Grant Agreement No. 870831 for the European Commission. It does not necessary reflect the view of the European Union and in no way anticipates the Commission’s future policy in this area.

2

Introduction

This document presents the results of an assessment of the impact of migration and a measure- ment of the contribution provided by ‘third country nationals’ (TCNs) to the economic systems in the receiving contexts. In total, 10 statistical briefings for the MATILDE regions have been compiled, including the following countries: Austria, Bulgaria, Finland, Germany, Italy, Norway, Spain, Sweden, Turkey and the United Kingdom. Each document focuses on the following dimen- sions of economic development: economic growth, labour markets, innovation, and entrepre- neurship. Moreover, a short introduction on the relevance of TCNs in the respective country as well as the population development and structure are considered. The results of a literature review and a comparative analysis of the 10 statistical briefings are published separately.

Within the Matilde project there is a special focus on so-called ‘third country nationals’ (TCNs). MATILDE considers TCNs as Non EU citizens, who reside legally in the European Union (EU) and who are the target of EU integration policies. A TCN is “any person who is not a citizen of the European Union within the meaning of Art. 20(1) of TFEU and who is not a person enjoying the European Union right to free movement, as defined in Art. 2(5) of the Regulation (EU) 2016/399 (Schengen Borders Code).“ (European Commission 2020a). According to this definition, nationals of European Free Trade Association (EFTA) member countries Norway, Island, Liechtenstein and Switzerland are not considered to be TCNs. TCNs cover economic migrants, family migrants, stu- dents and researchers, highly skilled migrants, and forced migrants.

3

BULGARIA Figure 1: Origin and share of total population of different nationalities in bulgaria

Source: Eurostat (2020a), own illustration Third Country Nationals (TCNs) in Bulgaria to these nationalities; there is no data about citi- zenship available for 0.16 % of the population. Fig- On January 1, 2020, a total of 104,642 persons ure 1 gives an overview of the origin and the share with non-Bulgarian citizenship were living in Bul- of the different nationalities in Bulgaria. garia. This corresponds to a share of 1.5 % of Bul- garia's total population. Among non-Bulgarian na- Population Development & Population Structure tionals, only 13.7 % (14,342 persons; 0.2 % of total The number of inhabitants in Bulgaria has already population) came from EU-28 (incl. UK) and EFTA decreased in the past and this development is pro- countries (190 persons; 0.003 %). A total of jected to continue in the future. As of January 1, 90,110 persons (or 1.3 % of total population) were 2020, there were 6,951,482 persons registered in third country nationals (TCN), being individuals Bulgaria. This corresponds to a decrease of who are neither Bulgarians, EU-28 citizens nor 1,283,355 citizens (-15.1 %) within the last EFTA citizens. By comparison, the share of TCN in 20 years since 2000 (before eastward enlarge- EU-27 (excl. UK) and EU-28 (incl. UK) is 5.7 % and ment of EU in 2004). The main causes of popula- 3.8 %, therefore Bulgaria lying way below the av- tion decline are emigration of Bulgarian inhabit- erage. Considering the nationalities of TCN, Rus- ants to prosperous countries and low fertility sians (0,38 %) are above Turks (0.28 %), Syrians rates. (0.20 %) and Ukrainians (0.11 %). In total, almost 9 out of 10 (87.2 %) TCN living in Bulgaria belong

4

Figure 2: Population development (Δ 2000 to 2020), NUTS 3

Source: Eurostat (2020b), own illustration

A long-term analysis of population development the sustainability of pension systems, urbanization since the turn of the new millennium reveals clear etc., one relevant aspect is labor supply, i.e. the regional differences. Varna (+6.7 %) and Sofia population aged 15 to 64 years or 20 to 64 years in (+9.7 %) are the only Bulgarian NUTS 3 regions ex- industrialized economies and nations. From an periencing population growth since 2000. Sofia as economic perspective, labor (done by human be- the capital has the highest growth rate. Con- ings) is an essential element in the production of versely, all other regions recorded a sharp decline goods and services. The term ‘labor force’ com- in population. Relative reductions were highest for prises all of those who work for gain, whether as the northwestern regions Vidin (-40.3 %), Vratsa (- employees, employers, or as self-employed, and it 37.6 %) and Montana (-33.6 %; see Figure 2). includes the unemployed who are seeking for work. There are three main causes that explain foregone developments: (1) a migratory movement from As Figure 3 (a)-(c) indicate, the decrease in the economically less developed Eastern European working-age population within the regions is countries to prosperous regions of European Un- mainly caused by the decrease in the Bulgarian ion and North America (2) Ongoing rural-urban mi- population (-600,300 inh.). As the share of foreign- gration increasingly taking place within Bulgaria ers is very small and decreased in the past as well (Varna and Sofia) (3) a sharp decline in birth rates (-3,400 inh.) immigration has no impact to allevi- after 1990. ate or worsen the problem. Summing up, both, do- mestic population groups are (re-)locating in cities While most discussions about changing demo- and metropolitan regions (especially Sofia and graphic structures focus on population ageing and

5

Varna). This is at the expense of rural areas which (82.69 %) and Pleven (81.72 %). As a result of age- are experiencing an overall decline in population. ing societies, the total-age dependency ratio will rise in the future by 13.8 pp to 82.1 % in 2040 re- From a regional perspective the areas with the sulting in implications for publicly funded social se- highest working-age population declines are lo- curity schemes (e.g., fiscal policy). cated in the north, i.e. northwestern (-30.3 %), northern central (-23.8 %), and east, i.e. south- Figure 3: Working-age population eastern (-13.5 %) and northeastern (-10.3 %). (20 to 64 years; Δ 2003 to 2019), NUTS 2 (a) In other words, this is the case for all regions ex- cept for the southwestern part of Bulgaria where Sofia is located with a stagnating population in working-age (0.0 %). These past developments un- derline the high relevance of migration concerning the development of the working-age population especially for rural regions. Immigration of non- Bulgarians was of negligible relevance in the past as the share of foreigners is quite low. Due to the out migration of Bulgarians to North America and Central Europe the labor force potential, i.e. pop- ulations between 20 and 64 years shrunk dramat- ically in the past, being a challenge from an eco- (b) nomic point of view (see Figure 3 (a) and (b)). Also, the future projections up to 2040 (Figure 4) give reason for concern. Except of a projected stagnat- ing development in Sofia, all Bulgarian regions show a decline in working-age population – up to even almost 50 %.

The total-age dependency ratio is a measure of the age structure of the population. It relates the num- ber of individuals who are likely to be “dependent” on the support of others for their daily living – the young (up to 19 years old) and the elderly

(65 plus years old) – to the number of those indi- (c) viduals who are, being in working age from 20 to 64 years, capable of providing this support. The to- tal-age dependency ratio in is 68.2 %. The dependency ratio is usually lower in metropolitan areas and cities than in rural regions. The key factors here are a more prosperous econ- omy, i.e. a high ratio of working people, and low child ratios. Most metropolitan regions in Bulgaria are also characterized by relatively low rates (e.g., Sofia: 57.50 %). High dependency ratios are found in regions where the share of children and the el- derly (65 years and older) is above average. These regions are mainly Vidin (86.57 %), Lovech Source: Eurostat (2020c), own illustration

6

Figure 4: Projections of working-age population (20 to 64 years; Δ 2020 to 2040), NUTS 3

Source: Eurostat (2020d), own illustration Education and (Un)Employment which in turn is higher than in rural areas3. In rural areas in Bulgaria, only 8.3 % of individuals have The tertiary education rate of Bulgaria's domestic university degrees, i.e. tertiary education population is 24.6 %, which is only slightly below (ISCED2011 levels 5 to 8), compared to 20.5 % in that of the total population (24.7 %). Compared to towns & suburbs and 36.1 % in cities. the EU-28 (29.5 %) Bulgaria has a below-average tertiary education rate. Bulgaria ranks 21th within In contrast, secondary degrees are more common the EU-28, ahead of Portugal (23.8 %), Slovakia in less densely populated areas. While less than (23.1 %) and Hungary (22.5 %). Luxembourg 50.6 % of city residents have secondary, i.e. upper (41.0 %), Ireland (40.7 %) and United Kingdom secondary and post-secondary non-tertiary edu- (40.6 %) rank at the top end, while the Czech Re- cation (ISCED2011 levels 3 to 4) attainment level, public (21.6 %), Italy (17.4 %) and Romania 59.2 % and 53.6 % of residents in towns & suburbs (16.0 %) rank at the bottom. and rural areas do so. Due to lack of data no state- ments can be made in regard to immigrants Across the world, educational attainment is signif- and/or foreigners living in Bulgaria (see Table 1). icantly higher in cities1 than in towns and suburbs2,

1 A city is a local administrative unit (LAU) where at least 50 density of at least 1 500 inhabitants per km² and collectively % of the population lives in one or more urban centres (Eu- a minimum population of 50 000 inhabitants after gap-filling rostat 2020e). (Eurostat 2020f; Eurostat 2020g). 2 Towns & suburbs are areas where less than 50 % of the 3 A rural area is an area where more than 50 % of its popula- population lives in rural grid cells and less than 50 % of the tion lives in ‘rural grid cells’, i.e. grid cells that are not identi- population lives in urban centres, i.e. a cluster of contiguous fied as urban centres or as urban clusters (Eurostat 2020h; grid cells of 1 km² (excluding diagonals) with a population Eurostat 2020i).

7

Table 1: Working-age population by citizenship, degree of seen for Bulgaria. Those with primary education urbanization and educational attainment level 2019 are most affected by unemployment (13.0 %) and this situation has remained constant for years. Educational Attainment Level Citizenship (ISCED11) Conversely, people with tertiary education have the lowest unemployment rate (1.9 %), those with

Urbanisation Primary Secondary Tertiary secondary education a value of 3.4 %. The same Bulgarians 13.4% 50.6% 36.1% trend can be observed at the European level (EU- EU-28 n.a. n.a. n.a. 28; Figure 6).

Cities Cities Non EU-28 n.a. n.a. n.a. Total 13.4% 50.6% 36.1% Figure 6: Unemployment rate by educational attainment Bulgarians 20.3% 59.2% 20.5% level 2019 (20 to 64 years; ISCED 2011) EU-28 n.a. n.a. n.a. Non EU-28 n.a. n.a. n.a. 14 13.0 suburbs Towns & Towns Total 20.3% 59.2% 20.5% 12.4 Bulgarians 38.1% 53.6% 8.3% 12 EU-28 n.a. n.a. n.a. 10

Rural Non EU-28 n.a. n.a. n.a. Total 38.1% 53.6% 8.3% 8 Source: Eurostat (2020j), own illustration 6 5.6 The unemployment rate in Bulgaria is 4.2 %, being 4.0 below the EU-28 average of 6.2 %. From a regional 4 3.4 perspective, North Western has the highest unem- 1.9 2 ployment rate at 10.6 % (female: 9.2 %; male: 11.9 %), followed by Northeastern (5.9 %; 0 female: 6.6. %; male: 5.3 %) and Northern Central Primary Secondary Tertiary (5.1 %; female: 4.8 %; male: 5.4 %). Southwestern (2.2 %; female: 1.9 %; male: 2.5%), Southeastern EU-28 Bulgaria

(2.5 %; female: 2.3 %; male: 4.5%) and Southern Source: Eurostat (2020k), own illustration Central (3.0 %; female: 2.8 %; male: 3.2 %) are the provinces with the lowest unemployment rates In contrast, the higher the educational attainment (see Figure 5). the higher the employment rate, i.e. 88.5 % for tertiary, 76.0 % for secondary and 51.2 % for pri- Figure 5: Unemployment rate 2019 (20 to 64 years) mary level. No reliable statements can be made due to a lack of data probably caused by the low number of foreigners living in Bulgaria.

The abovementioned findings foster the theoreti- cal conclusions by Becker (1964) that the educa- tional level has a significant influence on the pro- fessional career and the risk of unemployment of human beings.

This theory and the empirical evidence can also be applied to the “Early Leavers from Education and Training”. This population group refers to persons aged 18 to 24 who have completed a lower sec- Source: Eurostat (2020k), own illustration ondary education and are not involved in further In general, there is a correlation between the level education or training. From a fiscal point of view, of education and unemployment, according to this specific group is of particular relevance as they which the well-educated are significantly less are more likely to be (long-term) unemployed. In likely to be unemployed than people without a vo- 2019, 10.3 % of the 18-24 year olds in the EU-28 cational qualification. This correlation can also be were part of this group (male: 11.9 %; female:

8

8.6 %). Bulgaria ranges with a value of 13.9 % Figure 8: Mean and median equivalized net income by above the European average. In total, almost two broad group of citizenship 2018 (18 to 64 years) thirds (63.3 %) of all early leavers are jobless but 7,000 no relying statements can be made in regard to im- 6,194 migrational status because of missing data (Euro- 6,000 5,551 4,899 stat 2020l). 5,000 4,799

Income and Gross Domestic Product 4,000

As for the employment rate, the educational at- 3,000 tainment level has a positive impact on mean and 2,000 median4 equivalized net income, i.e. the higher the educational attainment level the higher the in- 1,000 come. Working persons in Bulgaria at primary 0 level earn less (mean: € 2,868; median: € 2,532) Foreign country Bulgaria than secondary (€ 5,544 or € 4,819) and tertiary levels (€ 10,030 or € 7,282; see Figure 7). The Mean equivalised net income Median equivalised net income mean equivalized net income of Bulgarians is higher than of foreigners (€ 6,194 vs. € 5,551) Source: Eurostat (2020n), own illustration though it is the opposite when analyzing the me- The indicator ‘people at risk of poverty or social dian (€ 4,899 vs. € 4.799). Therefore, no clear and exclusion’ corresponds to the sum of persons who reliable statements can be made in regard to in- are: at risk of poverty after social transfers, se- come which is probably because of the low num- verely materially deprived or living in households ber of foreigners living in Bulgaria (see Figure 8). with very low work intensity. Almost one quarter Figure 7: Mean and median equivalized net income by edu- of foreigners (24.7 %) belongs to this group cational attainment level 2018 (18 to 64 years; ISCED 2011) whereas the situation for the domestic population is slightly worse (+7.8 pp or 32.5 %) are compara- 12,000 bly low (see Table 2). Due to a lack of data no 10,030 10,000 statements can be made in regard to EU-28 citi- zens or TCNs. 8,000 7,282 In 2019, compensation of employees (wages and 6,000 5,544 4,819 salaries plus employers’ social contributions) was 4,000 the largest income component of GDP in EU-28 2,868 2,532 and the Euro area, accounting for 47.8 % and 2,000 48.0 %. Bulgaria (43.7 %) lies below European av- erage (both EU-28 and Euro area). Taxes on pro- 0 duction and imports (less subsidies) accounted for Primary Secondary Tertiary 11.5 % (EU-28: 11.9 %; Euro area: 11.5 %). Gross Mean equivalised net income operating surplus and mixed income accounted Median equivalised net income for 44.8 % of GDP in Bulgaria, 40.3 % of GDP for

Source: Eurostat (2020m), own illustration the EU-28 and 40.5 % in the Euro area. Bulgarian GDP at current prices amounted to approx. € 61.2 billion in 2019 and GDP per inhabitant equaled € 8,780 (Eurostat 2020p).

4 The median is more robust, i.e. less sensitive to outliers than the mean. This explains why the mean equivalized income is always slightly higher for each educational level.

9

Table 2: Population aged 18 and over by risk of poverty, material deprivation, work intensity 2019

Risk of poverty At risk Not at risk At risk Not at risk Work Intensity Material Deprivation Bulgaria Foreign country Very low Severe 2.6 0.6 0.8 0.0 Not very low Severe 8.6 9.3 3.9 1.2 Very low Non severe 1.4 1.1 1.8 0.8 Not very low Non severe 8.9 16.2 People at Risk of Poverty or Social Exclusion 32.5 24.7 Not at risk Not at risk Not very low Non severe 67.5 75.4: People at no Risk of Poverty or Social Exclusion 67.5 75.4 Total 100.0 100.0 Source: Eurostat (2020o), own illustration

Table 3: Nominal regional GDP 2019 (approx. € 6,800). As in previous years, all regions and real growth rate (Δ 2002 to 2019) except Southwestern were below the national Total Per Capita level of € 8,800. Provinces with a high share of foreigners have a high GDP per capita and vice versa. On the one

hand, prosperous regions are more attractive to migrants; on the other hand, these regions benefit from a better availability of labor supply, i.e. Region greater number of working age people. In this con- GDP (in €) GDP (in text, one speaks of a so-called cross-fertilization, current marketcurrent prices GDP (in €) €) per capita GDP (in

at current marketat current prices at which results from the interaction of these two 2015=100; Δ 2002 to 2019) 2002 to Δ 2015=100;

of the EU27 (from 2020) average EU27 the 2020) of (from factors. added (GVA) at basic prices (Index, prices at basic (GVA) added Purchasing power standard (PPS, EU27 (PPS, standard power Purchasing Real growth rate of regional gross value gross value growth rate regional Real of from 2020), per inhabitant in percentage per inhabitant from 2020), Expressing GDP in PPS (purchasing power stand- Northern and ards) eliminates differences in price levels be- 21,623 33.4 6,200 38 Eastern tween countries or regions. There are significant Bulgaria differences in levels of prosperity among Bulgarian Northwestern 3,939 11.1 5,400 32 Northern NUTS 2 regions. The most prosperous region in 4,557 31.7 5,900 35 Central Bulgaria is Southwestern (89 %); it is over Northeastern 6,308 44.3 6,800 41 57/54/52 pp richer than North Western/ Northern Southeastern 6,819 36.5 6,600 40 Central/ Southern Central. Northeastern ranks South-Western and South 39,617 69.9 11,300 68 second (41 %) and Southeastern third (40 %; see Central Bulgaria Table 3). Southwestern 30,951 75.2 14,800 89 Southern Economic structure and entrepreneurship 8,666 51.6 6,200 37 Central Bulgaria 61,240 56.2 8,800 53 In general, a distinction is made between three dif- Source: Eurostat (2020q), own illustration; ferent sectors of the economy, the primary (agri- Eurostat (2020r), own illustration culture and forestry), secondary (manufacturing) Within the last 20 years Bulgaria’s gross value and tertiary (services) sectors. added (GVA) increased by +56.2 % (real growth In total, approx. 2.3 million people were employed rate). In terms of regional GVA all NUTS 2 regions in . This corresponds to an in- recorded positive real growth, ranging from crease of +3.0 % compared to 2015. Almost two +11.1 % in North Western to +75.2 % in Southeast- thirds (65.7 %) are employed in the service sector, ern since 2000. The highest regional GDP per cap- where employment has increased by +5.2 % since ita at current prices occurred also in Southeastern 2015. Employment in secondary and primary sec- (approx. € 14,800) succeeded by Northeastern

10

tor shrunk by -0.9 % and -1.7 %. Generally speak- 1.7 %), “Manufacturing” (-1.4 %) and “Electricity, ing “Agriculture, forestry and fishing” is of very low gas, steam and air conditioning supply” (-1.4 %). " importance, i.e. only 3.0 % of total employment Information and communication" (+29.1 %), " Pro- (see Table 4). fessional, scientific and technical activities” (+11.3 %) as well as " Accommodation and food A more detailed examination of the different sec- service activities" (+10.8 %), on the other hand, tors of the economy reveals that there have also are among those economic sections with the larg- been significant shifts within the sectors: the only est absolute growth. There is no information avail- declines in employment were recorded in the sec- able about the citizenship of the workers (see Ta- tions " Mining and quarrying" (-22.3 %), "Educa- ble 4). tion” (-1.7 %), " Agriculture, forestry and fishing" (-

Table 4: Employment by economic activity 2019 Employed Economic Activity persons in Share Δ 15-19 2019 Primary Sector 69,726 3.0% -1.7% Agriculture, forestry and fishing 69,726 3.0% -1.7% Secondary Sector 726,309 31.3% -0.9% Mining and quarrying 19,137 0.8% -22.3% Manufacturing 505,119 21.7% -1.4% Electricity, gas, steam and air conditioning supply 29,834 1.3% -1.4% Water supply, sewerage, waste management 36,955 1.6% 5.8% and remediation activities Construction 135,264 5.8% 3.6% Tertiary Sector 1,526,526 65.7% 5.2% Wholesale and retail trade; repair of motor vehicles and 379,575 16.3% 1.6% motorcycles Transportation and storage 149,934 6.5% 5.9% Accommodation and food service activities 122,458 5.3% 10.8% Information and communication 102,249 4.4% 29.1% Financial and insurance activities 56,872 2.4% 2.3% Real estate activities 23,971 1.0% 5.0% Professional, scientific and technical activities 82,840 3.6% 11.3% Administrative and support service activities 116,941 5.0% 8.0% Public administration and defence; 112,322 4.8% 0.2% compulsory social security Education 162,757 7.0% -1.7% Human health and social work activities 140,448 6.0% 2.5% Arts,entertainment and recreation 36,989 1.6% 9.2% Other service activities 39,170 1.7% 7.1% Total 2,322,561 100.0% 3.0% Source: NSI (2021), own illustration

Research and Innovation A well-known indicator provided to measure achievements of countries or regions in R&D is Following the neoclassical and endogenous GERD, i.e. regional/national gross domestic ex- growth theories, technological advance is believed penditure on R&D as a percentage of GDP. GERD to be one of the major drivers of economic is estimated at 0.76 % for Bulgaria (2018) which is growth. From this perspective, there is a growing a slight increase of 0.02 pp to 2017. The total interest to investigate the link between research & amount of research expenditures was € 423.8 mil- development (R&D), innovation, entrepreneur- lion, the largest share (43.4 %) was financed by the ship and economic growth achieved by human business enterprise sector. 32.9 % were financed capital from abroad (EU-28 and non-EU-28). by foreign countries (rest of the world) and 23.7 %

11

were funded by the government sector (Eurostat reveals that most researchers (54.5 %) have ter- 2020s). tiary education but not doctoral or equivalent level; 41.1 % belong to the latter whereas 4.4 % Regional data is published on NUTS 1 and NUTS 2 have less than primary, primary, secondary and level. Southwestern was the province with the post-secondary non-tertiary education (see Figure highest ratio in 2018 (1.14 %). Northwestern fol- 9). lowed in second place with 0.49 % just ahead of Southern Central (0.44 %) and Northeastern Table 5: Intramural R&D expenditure as percentage of (0.43 %). The ratio in the remaining provinces is gross domestic product (GDP) 2017 Expenditures % of 0.32 % (Northern Central and South Eastern; see Province (€ mio.) GDP Table 5). Northern and Eastern 79.7 0.38% Bulgaria In accordance to the R&D funding structure almost Northwestern 19.1 0.49% one half (43.4 %) of the researchers work in the Northern Central 13.6 0.32% business enterprise sector. Approx. one third Northeastern 25.4 0.43% (30.6 %) works in the government sector, one Southeastern 21.5 0.32% South-Western and South 344.1 0.98% quarter (25.3 %) in the higher education sector Central Bulgaria and 0.6 % in the private non-profit sector. This Southwestern 308.8 1.14% makes a total of 15,094 full-time equivalent (FTE). Southern Central 35.3 0.44% Further analysis on educational attainment level Bulgaria 423.8 0.76% Source: Eurostat (2020t), own illustration

Figure 9: R&D researchers (full-time equivalent (FTE)) by sectors of performance and educational attainment level 2017 (ISCED2011) 3,500 3,137 3,000 2,532 2,500

2,000

1,500 1,353 1,280

1,000

506 500 136 14 3 0 Business enterprise sector Government sector Higher education sector Private non-profit sector

Less than primary, primary, secondary and post-secondary non-tertiary education (levels 0-4) Tertiary education excluding doctoral or equivalent level (levels 5-7) Doctoral or equivalent level

Source: Eurostat (2020u), own illustration

12

Conclusion

This Statistical Briefings examines the economic countries. From 2003 to 2019 almost all Bulgarian and spatial importance and impact of immigration NUTS 2 regions suffered heavily with regard to the with special regard to third-country nationals working-age population. As the share of immi- (TCN) in Bulgaria. grants was of low importance in the past, foreign- ers as TCN had almost no impact on the develop- On January 1, 2020, a total of 104,642 persons ment of the labor force potential (population aged with non-Bulgarian citizenship were living in Bul- between 20 and 64 years). garia. This corresponds to a share of 1.5 % of Bul- garia's total population. Among non-Bulgarian na- Projections regarding the working-age population tionals, only 13.7 % came from EU-28 (incl. UK) up to 2040 give reason for concern; with the ex- and EFTA. A total of 90,110 persons (or 1.3 % of ception of Sofia (with a stagnating projection) all total population) were third country nationals regions will further loose people aged between 20 (TCN), being individuals who are neither Bulgari- and 65 years. A decrease in emigration as well as ans, EU-28 citizens nor EFTA citizens. an increase in immigration of TCN could therefore play a pivotal role in reviving its (rural) areas. Within the last 20 years the population decline of Bulgaria has mainly been dominated by emigra- tion of Bulgarians to more prosperous regions and

13

Bibliography and source references

Becker, Gary Stanley (1964): Human Capital: A theoretical Eurostat (2020o): Persons aged 18 and over by risk of pov- and empirical analysis, with special reference to education. erty, material deprivation, work intensity of the household 3rd edition. New York. and most frequent activity status of the person - intersec- tions of Europe 2020 poverty target indicators, Eurostat (2020a): Population on 1 January by age group, sex https://appsso.eurostat.ec.europa.eu/nui/show.do?da- and citizenship, https://appsso.eurostat.ec.eu- taset=ilc_pees02&lang=en ropa.eu/nui/show.do?dataset=migr_pop1ctz&lang=en Eurostat (2020p): GDP and main components (output, ex- Eurostat (2020b): Population change - Demographic balance penditure and income), https://appsso.eurostat.ec.eu- and crude rates at regional level (NUTS 3), ropa.eu/nui/show.do?dataset=nama_10_gdp&lang=en https://appsso.eurostat.ec.europa.eu/nui/show.do?da- taset=demo_r_gind3&lang=en Eurostat (2020q): Gross domestic product (GDP) at current market prices by NUTS 2 regions, https://appsso.euro- Eurostat (2020c): Population by sex, age, citizenship, labour stat.ec.europa.eu/nui/show.do?da- status and NUTS 2 regions, https://appsso.eurostat.ec.eu- taset=nama_10r_2gdp&lang=en ropa.eu/nui/show.do?dataset=lfst_r_lfsd2pwn&lang=en Eurostat (2020r), Real growth rate of regional gross value Eurostat (2020d): Population on 1st January by age, sex, type added (GVA) at basic prices by NUTS 2 regions - percentage of projection and NUTS 3, https://appsso.eurostat.ec.eu- change on previous year, https://appsso.eurostat.ec.eu- ropa.eu/nui/show.do?dataset=proj_19rp3 ropa.eu/nui/show.do?dataset=nama_10r_2gvagr&lang=en Eurostat (2020e): City, Eurostat (2020s): GERD by sector of performance and source https://ec.europa.eu/eurostat/statistics- of funds, http://appsso.eurostat.ec.eu- explained/index.php?title=Glossary:City ropa.eu/nui/show.do?dataset=rd_e_gerdfund&lang Eurostat (2020f): Town or suburb, https://ec.europa.eu/eu- Eurostat (2020t): Intramural R&D expenditure (GERD) by rostat/statistics-explained/index.php?title=Glos- NUTS 2 regions, https://ec.europa.eu/eurostat/data- sary:Town_or_suburb browser/view/tgs00042/default/table?lang=en Eurostat (2020g): Urban centre, Eurostat (2020u): Total R&D personnel and researchers by https://ec.europa.eu/eurostat/statistics- sectors of performance, educational attainment level explained/index.php?title=Glossary:Urban_centre (ISCED2011) and sex, https://ec.europa.eu/eurostat/data- Eurostat (2020h): Rural area, https://ec.europa.eu/euro- browser/view/rd_p_persqual11/default/table?lang=en stat/statistics-explained/index.php/Glossary:Rural_area NSI (2021): Average annual wages and salaries of the em- Eurostat (2020i): Rural grid cell, https://ec.europa.eu/euro- ployees under labour contract by economic activity group- stat/statistics-explained/index.php?title=Glossary:Ru- ings and sectors in 2019, https://www.nsi.bg/en/con- tent/6439/total-economic-activity-groupings-kind-owner- ral_grid_cell ship -gender Eurostat (2020j): Population by educational attainment level, sex, age, citizenship and degree of urbanization, http://appsso.eurostat.ec.europa.eu/nui/show.do?da- taset=edat_lfs_9916&lang=en

Eurostat (2020k): Unemployment rates by sex, age, educa- tional attainment level and NUTS 2 regions (%), https://appsso.eurostat.ec.europa.eu/nui/show.do?da- taset=lfst_r_lfu3rt&lang=en

Eurostat (2020l): Early leavers from education and training by sex and citizenship, http://appsso.eurostat.ec.eu- ropa.eu/nui/show.do?dataset=edat_lfse_01&lang=en

Eurostat (2020m): Mean and median income by educational attainment level - EU-SILC survey, http://appsso.euro- stat.ec.europa.eu/nui/show.do?lang=en&dataset=ilc_di08

Eurostat (2020n): Mean and median income by broad group of citizenship (population aged 18 and over), https://appsso.eurostat.ec.europa.eu/nui/show.do?da- taset=ilc_di15&lang=en

14

ECONOMIC IMPACT OF MIGRATION

Statistical Briefings Finland

Call: H2020-SC6-MIGRATION-2019

Work Programmes:

H2020-EU.3.6.1.1. The mechanisms to promote smart, sustainable and inclusive growth H2020-EU.3.6.1.2. Trusted organisations, practices, services and policies that are necessary to build resilient, inclusive, participatory, open and creative societies in Europe, in particular taking into account migration, integration and demographic change

Deliverable 4.2 – Statistical briefing for Finland on economic impact of TCNs in MATILDE re- gions

Authors: Birgit Aigner-Walder, Albert Luger, Rahel M. Schomaker with contributions from Jussi Laine and Daniel Rauhut

Approved by Work Package Manager of WP4: Simone Baglioni, UNIPR Approved by Scientific Head: Andrea Membretti, UEF Approved by Project Coordinator: Jussi Laine, UEF

Version: 28.05.2021

DOI: 10.5281/zenodo.4817376

This document was produced under the terms and conditions of Grant Agreement No. 870831 for the European Commission. It does not necessary reflect the view of the European Union and in no way anticipates the Commission’s future policy in this area.

2

Introduction

This document presents the results of an assessment of the impact of migration and a measure- ment of the contribution provided by ‘third country nationals’ (TCNs) to the economic systems in the receiving contexts. In total, 10 statistical briefings for the MATILDE regions have been compiled, including the following countries: Austria, Bulgaria, Finland, Germany, Italy, Norway, Spain, Sweden, Turkey and the United Kingdom. Each document focuses on the following dimen- sions of economic development: economic growth, labour markets, innovation, and entrepre- neurship. Moreover, a short introduction on the relevance of TCNs in the respective country as well as the population development and structure are considered. The results of a literature review and a comparative analysis of the 10 statistical briefings are published separately.

Within the Matilde project there is a special focus on so-called ‘third country nationals’ (TCNs). MATILDE considers TCNs as Non EU citizens, who reside legally in the European Union (EU) and who are the target of EU integration policies. A TCN is “any person who is not a citizen of the European Union within the meaning of Art. 20(1) of TFEU and who is not a person enjoying the European Union right to free movement, as defined in Art. 2(5) of the Regulation (EU) 2016/399 (Schengen Borders Code).“ (European Commission 2020a). According to this definition, nationals of European Free Trade Association (EFTA) member countries Norway, Island, Liechtenstein and Switzerland are not considered to be TCNs. TCNs cover economic migrants, family migrants, stu- dents and researchers, highly skilled migrants, and forced migrants.

3

FINLAND Figure 1: Origin and share of total population of different nationalities in Finland

Source: Eurostat (2020a) Third Country Nationals (TCNs) in Finland overview of the origin and the share of the differ- ent nationalities in Finland. On January 1, 2020, a total of 267,629 persons with non-Finnish citizenship were living in Finland. Population Development & Population Structure This corresponds to a share of 4.8 % of Finland's The number of inhabitants in Finland increased in total population. Among non-Finnish nationals, the past but according to recent projections a de- 37.8 % (101,072 persons; 1.8 % of total popula- crease of population will take place in the future. tion) came from EU-28 (incl. UK) and EFTA coun- As of January 1, 2020, there were 5,525,292 per- tries (1,553 persons; 0.03 %). A total of sons registered in Finland. This corresponds to an 165,004 persons (or 3.0 % of total population) increase of 353,990 citizens (+ 6.8 %) within the were third country nationals (TCN), being individ- last 20 years since 2000 (before eastward enlarge- uals who are neither Finnish, EU-28 citizens nor ment of EU in 2004). EFTA citizens. By comparison, the share of TCN in EU-27 (excl. UK) and EU-28 (incl. UK) is 5.7 % and A long-term analysis of population development 3.8 %, therefore Finland lying below the average. since the turn of the new millennium reveals clear regional differences. Helsinki-Uusimaa (+19.96 %), Considering the nationalities of TCN, Russians (0.52 %) are above Iraqis (0.25 %), Chinese (incl. Hong Kong; 0.18 %), Thai (0.14 %) and Indians (0.12 %). In total, 40.5 % of TCN living in Finland belong to these nationalities. Figure 1 gives an

4

Åland (+ 14.90 %)1 and Pirkanmaa (+14.03 %) rec- downturn: Kainuu (-14.43 %), Etelä-Savo orded the highest increase of population. Approx. (-13.91 %) and Kymenlaakso (-8.32 %; see Figure half of the regions recorded population growth, 2). others registered a considerable population

Figure 2: Population development (Δ 2000 to 2020), NUTS 3

Source: Eurostat (2020b), own illustration

1 The Åland Islands is an autonomous Finnish region and (CEC, 2020). Over the last 15 year, Åland Islands is one of the holds a special legal status in the EU (Government of Åland Nordic regions displaying the highest rate of regional devel- Island, 2020). It has its own operation program under the opment. Despite being rural and remote, Åland Islands is a Cohesion Policy receiving funding from the European Re- wealthy and well-off region (Rauhut and Costa, 2021). gional Development Fund and the European Social Fund

5

While most discussions about changing demo- As Figure 3 (a)-(d) indicate, the working-age popu- graphic structures focus on population ageing and lation in Finland already decreased within the time the sustainability of pension systems, urbanization period considered (-49,800 inh.). The decline etc., another economically relevant aspect is the would have been even higher without immigration labor supply, i.e. the population aged 15 to 64 from EU-28 (except Finnish; +28,300 inh.) and years or 20 to 64 years in industrialized economies non-EU 28 (+52,900 inh.) countries, as the work- and nations. From an economic perspective, labor ing-age population of Finnish people shrank since (done by human beings) is an essential element in 2006 (-133,000 inh.). From a regional perspective, the production of goods and services. The term ‘la- Helsinki-Uusimaa (+9.5 %) was the only region bor force’ comprises all of those who work for growing, while Åland (+0.6 %) almost stagnated. gain, whether as employees, employers, or as self- Pohjois- ja Itä-Suomi (-8.3 %), Etelä-Suomi employed, and it includes the unemployed who (- 7.2 %) and Länsi-Suomi (-3.2 %) recorded the are seeking for work. highest declines.

Figure 3: Working-age population (20 to 64 years; Δ 2006 to 2019) (a) Total -49,800 inh. (c) Non EU-28 +52,900 inh.

(b) Finish -133,000 inh. (d) EU-28 +28,300 inh.

Source: Eurostat (2020c), own illustration

6

According to current forecasts the decrease of la- The total-age dependency ratio in Finland in 2020 bor force potential does not stop in the future. is 74.4 %. The dependency rate is usually lower in Working-age population is projected to decrease metropolitan areas and cities than in rural regions. by – 3.3 % until 2040. There are only 3 regions The key factors here are a more prosperous econ- which are projected to grow in the future, i.e. Hel- omy, i.e. more people in working-age, and low sinki-Uusimaa (8.13 %), Åland (+ 5.08 %) and Pir- child ratios. Most metropolitan regions in Finland kanmaa (1.03 %; see Figure 4). are also characterized by relatively low ratios. High dependency ratios are found in Etelä-Pohjanmaa The total-age dependency ratio is a measure of the (90.50 %), Etelä-Savo (93.53 %) and Keski- age structure of the population. It relates the num- Pohjanmaa (92.45%). As a result of the ageing so- ber of individuals who are likely to be “dependent” ciety, the total-age dependency ratios will rise in on the support of others for their daily living the – the future by 2.4 pp to 77.0 % in 2040, resulting in young (up to 19 years old) and the elderly implications/challenges for publicly funded social (65 plus years old) to the number of those indi- – security schemes (e.g., expenditures for pensions). viduals who are, being at working age from 20 to 64 years old, capable of providing this sup- port.

Figure 4: Projections of working-age population (20 to 64 years; Δ 2020 to 2040), NUTS 3

Source: Eurostat (2020d), own illustration

7

Education and (Un)Employment and non-EU-28 (23.7 %) citizens have a lower ter- tiary education rate than the domestic population The tertiary education rate of Finland's domestic (39.1 %). population is 39.1 %, which is only slightly above that of the total population (38.5 %). Compared to Table 1: Working-age population by citizenship, degree of the EU-28 (29.5 %) Finland ranks 5th within the urbanization and educational attainment level 2019 EU-28 next to Cyprus (40.0 %), ahead of Lithuania Educational Attainment Level (37.9 %), Sweden (37.8 %) and Estonia (36.5 %). Citizenship (ISCED11) Luxembourg (41.0 %), Ireland (40.7 %) and United

Kingdom (40.6 %) rank at the top end, while the Urbanisation Primary Secondary Tertiary Czech Republic (21.6 %), Italy (17.4 %) and Roma- Fi nnish 13.1% 40.0% 46.9% nia (16.0 %) rank at the bottom. EU-28 29.9% 35.7% 34.4%

Cities Cities Non EU-28 39.0% 35.3% 25.6% Across the world, educational attainment is signif- Total 14.6% 39.7% 45.6% icantly higher in cities than in towns and suburbs, Fi nnish 16.8% 47.1% 36.2% EU-28 33.4% 43.8% 22.9% which in turn is higher than in rural areas. In rural Non EU-28 49.5% 32.2% 18.4% suburbs

2 & Towns areas in Finland, only 30.7 % of individuals have Total 17.5% 46.8% 35.7% university degrees, i.e. tertiary education Fi nnish 19.3% 49.7% 31.0% (ISCED2011 levels 5 to 8), compared to 35.7 % in EU-28 35.3% 47.9% n.a.

Rural Non EU-28 45.2% 35.0% n.a. towns & suburbs3 and 45.6 % in cities4. While less Total 19.7% 49.6% 30.7% than 39.7 % of city residents have secondary, i.e. Source: Eurostat (2020j), own illustration upper secondary and post-secondary non-tertiary education (ISCED2011 levels 3 to 4) attainment The unemployment rate in Finland is 6.1 %, being level, 46.8 % and 49.6 % of residents in towns & close to EU-28 average of 6.2 %. From a regional suburbs and rural areas do. Primary educational perspective, Pohjois- ja Itä-Suomi has the highest attainment level has low relevance as approx. only unemployment rate at 7.3 % (female: 5.9 %; 1 out of 5 residents in Finland belongs to this male: 8.5 %). All other regions range close to each group. other, with Helsinki-Uusimaa recording the lowest rate (5.7 %; female: 5.6 %; male: 5.9 %; see Figure Concerning the educational attainment level of 5). TCN, big differences can be recorded. In cities, where most of the non-EU-28 citizens live, approx. In general, there is a correlation between the level 40 % (39.0 %) only have primary education. This of education and unemployment, according to share is even higher in towns & suburbs (49.5 %) which the well-educated are significantly less and rural areas (45.2 %). On the contrary, only likely to be unemployed than people without a vo- 25.6 % of non-EU-28 citizens living in cities have a cational qualification. This correlation can also be university degree but this value is still much higher seen for Finland. Those with primary education than in suburban (18.4 %) regions. Differences of are most affected by unemployment (12.4 %). educational attainment level between Finnish and Conversely, people with tertiary education have EU-28 residents range within a broad band from the lowest unemployment rate (4.0 %), those with 1.8 pp to 16.8 pp and are therefore scarcely com- secondary a value of 7.1 %. Similar results are ob- parable, too (see Table 1). In total, EU-28 (28.8 %) served at the European level (EU-28; see Figure 6)

2 A rural area is an area where more than 50 % of its popula- grid cells of 1 km² (excluding diagonals) with a population tion lives in ‘rural grid cells’, i.e. grid cells that are not identi- density of at least 1 500 inhabitants per km² and collectively fied as urban centres or as urban clusters (Eurostat 2020e; a minimum population of 50 000 inhabitants after gap-filling Eurostat 2020f). (Eurostat 2020g; Eurostat 2020h). 3 Towns & suburbs are areas where less than 50 % of the 4 A city is a local administrative unit (LAU) where at least 50 population lives in rural grid cells and less than 50 % of the % of the population lives in one or more urban centres (Eu- population lives in urban centres, i.e. a cluster of contiguous rostat 2020i).

8

Figure 5: Unemployment rate 2019 (20 to 64 years) as well, as the risk of unemployment decreases by the highest educational attainment.

As Figure 7 shows, Finns have the highest employ- ment rate for tertiary educational attainment level (86.7 %), but EU-28 ranks first for primary (76.6 % vs. 52.2 %) and secondary educational attainment level (80.3 % vs. 74.9 %). Non-EU-28 citizens have the lowest employment rate on each level (62.6 %; 58.1 %; 40.2 %).

Figure 6: Unemployment rate by educational attainment level 2019 (20 to 64 years; ISCED 2011)

14 12.4 12.4 12

10

8 7.1

6 5.6 4.0 4.0 4

2

0 Primary Secondary Tertiary

EU-28 Finland

Source: Eurostat (2020k), own illustration

Figure 7: Employment rate by educational attainment level and citizenship 2019 (20 to 64 years; ISCED 2011)

100 90 86.7 80.3 79.5 80 76.6 74.9 70 62.6 58.1 60 52.2 50 40.2 40 30 20 10 0 Primary Secondary Tertiary Source: Eurostat (2020k), own illustration EU-28 Non EU-28 Finland In contrast, the higher the educational attainment level, the higher the employment rate, i.e. 86.2 % Source: Eurostat (2020l), own illustration for tertiary, 74.6 % for secondary and 52.0 % for primary level. From an immigrational point of view, this observation holds for other citizenships

9

The abovementioned findings foster the theoreti- Figure 8: Mean and median equivalized net income by edu- cal conclusions by Becker (1964) that the educa- cational attainment level 2018 (18 to 64 years; ISCED 2011) tional level has a significant influence on the pro- 40,000 fessional career and the risk of unemployment of 35,692 35,000 human beings. 32,027 30,000 26,793 This theory and the empirical evidence can also be 24,941 25,000 24,150 applied to the “Early Leavers from Education and 21,093 Training”. This population group refers to persons 20,000 aged 18 to 24 who have completed a lower sec- 15,000 ondary education and are not involved in further 10,000 education or training. From a fiscal point of view, this specific group is of particular relevance as they 5,000 are more likely to be (long-term) unemployed. In 0 2019, 10.3 % of the 18-24 year olds in the EU-28 Primary Secondary Tertiary were part of this group (male: 11.9 %; female: Mean equivalised net income 8.6 %). Finland ranges with a value of 7.3 % below Median equivalised net income the European average. The share of EU-28 citizens Source: Eurostat (2020n), own illustration residing in Finland and that of non-EU-28 residents is slightly higher than that of the domestic popula- Figure 9: Mean and median equivalized net income by tion (7.1 %). In total, approx. 60 % (58.9 %) of early broad group of citizenship 2018 (18 to 64 years) school leavers are jobless, but no relying state- 35,000 ments can be made in regard to the immigrational 30,000 30,000 28,531 status due to missing data (Eurostat 2020m). 27,170 25,000 23,028 Income and Gross Domestic Product 21,449 20,000 18,882 As for the employment rate, the educational at- tainment level has a positive impact on mean and 15,000 median5 equivalized net income, i.e. the higher 10,000 the educational attainment level the higher the in- come. 5,000

Working persons in Finland at primary educational 0 EU-28 Non EU-28 Finland level earn less (mean: € 24,150; median: € 21,093) than secondary (mean: 26,793; me- € Mean equivalised net income dian: € 24,941) and tertiary levels (mean: Median equivalised net income

€ 35,692; median: € 32,027; see Figure 8). As Source: Eurostat (2020o), own illustration working Finnish people have a higher educational attainment level (in regard to tertiary education A further analysis and decomposition of wage dif- rate) than EU-28 and non-EU-28 (see Table 1) this ferentials between domestic and foreign workers could be one reason (amongst others) that their done by Lehmer and Ludsteck (2013) reveals that mean (€ 30,000 vs. € 28,531 and € 21,449) and me- factors like seniority, promotions to better-paid dian (€ 27,170 vs. € 23,028 and € 18,882) equival- occupations as well as job stability also have to be ized net income is higher (see Figure 9). taken into account. These factors cause an im- provement in an individual's firm-specific

5 The median is more robust, i.e. less sensitive to outliers than the mean. This explains why the mean equivalized income is always slightly higher for each educational level.

10

Table 2: Population aged 18 and over by risk of poverty, material deprivation, work intensity 2019 Risk of poverty At risk Not at risk At risk Not at risk At risk Not at risk Work Intensity Material Deprivation Finland EU-28 Non EU-28 Very low Severe 0.4 0.4 2.0 0.6 4.7 2.7 Not very low Severe 0.4 1.0 0.7 1.0 0.0 3.2 Very low Non severe 3.3 2.4 0.0 5.2 8.8 5.2 Not very low Non severe 7.6 8.2 9.9 People at Risk of Poverty or Social Exclusion 15.5 17.7 34.5 Not at risk Not at risk Not at risk Not very low Non severe 84.5 82.3 65.4 People at no Risk of Poverty or Social Exclusion 84.5 82.3 65.4 Total 100.0 100.0 100.0 Source: Eurostat (2020p), own illustration human capital (training on the job) as the median pears Finland’s gross value added (GVA) increased age of foreign workers and therefore the length of by 23,4 % (real growth rate). In terms of regional time within the company is lower than for domes- Table 3: Nominal regional GDP (2019) tic ones. However, these assumptions cannot be and real growth rate (Δ 2002) verified in the context of this analysis due to a lack Total Per Capita of available data.

The indicator ‘people at risk of poverty or social exclusion’ corresponds to the sum of persons who are: at risk of poverty after social transfers, se- verely materially deprived or living in households with very low work intensity. As non-EU 28 citizens Region have a higher risk of unemployment and they earn €) GDP (in

less than domestic and EU-28 residents, more €) per capita GDP (in at current marketat current prices marketat current prices than one third (34.5 %) belongs to this group. EU- 2019) 2002 to Δ 2015=100; of the EU27 (from 2020) average EU27 the 2020) of (from

28 citizens (17.7 %) and Finnish (15.5 %) face the (Index, at prices basic (GVA) added Purchasing power standard (PPS, EU27 (PPS, standard power Purchasing Real growth rate of regional gross value gross value growth rate regional Real of lowest risk of social exclusion and poverty (see Ta- in percentage per inhabitant from 2020), Manner-Suomi 239,126 23.5 43,500 111 ble 2). Länsi-Suomi 52,967 20.1 38,400 98 Helsinki- In 2019, compensation of employees (wages and 94,961 29.9 56,500 144 Uusimaa salaries plus employers’ social contributions) was Etelä-Suomi 44,756 16.6 38,900 99 the largest income component of GDP in EU-28 Pohjois- 46,442 21.5 36,200 93 and the Euro area, accounting for 47.8 % and ja Itä-Suomi 48.0 %. Finland (46.6 %) lies below the European Åland 1,359 12.2 45,500 116 Finland 240,561 23.4 43,600 111 average (both EU-28 and Euro area). Taxes on pro- Source: Eurostat (2020r), own illustration; duction and imports (less subsidies) accounted for Eurostat (2020s), own illustration 12.8 % (EU-28: 11.9 %; Euro area: 11.5 %). Gross GVA all provinces also recorded positive real operating surplus and mixed income accounted growth, ranging from 12.2 % in Åland to 29.9 % in for 40.6 % of GDP in Austria, 40 3 % of GDP for the Helsinki-Uusimaa since 2000. The highest regional EU-28 and 40.5 % in the Euro area. Finnish GDP at GDP per capita at current prices occurred also in current prices amounted to approx. € 240.3 billion Helsinki-Uusimaa (approx. € 56,500) succeeded by in 2019 and GDP per inhabitant equaled € 43,510 45,500). As in previous years, all (Eurostat 2020tq). Åland (approx. € other provinces were below the national level of Table 3 provides information about nominal re- € 43,600 (Pohjois- ja Itä-Suomi: € 36,200; Länsi- gional GDP and real growth rates, whereas the Suomi: € 38,400; Etelä-Suomi: € 38,900; see Table federal states (NUTS 1) are grey-shaded. Within 20 3).

11

Table 4: Employment by economic activity 2019 Employed EU-28 Occupational Groups persons in Share Δ 10-18 Finnish & TCN 201 8 EFTA Armed forces 8,015 0.3% -21.9% 99.6% 0.1% 0.3% Managers 95,823 4.0% 14.0% 96.0% 1.7% 2.3% Professionals 483,791 20.4% 14.3% 94.4% 2.0% 3.6% Technicians and associate professionals 430,756 18.1% 2.3% 96.5% 1.2% 2.3% Clerical support workers 130,288 5.5% -23.6% 95.9% 1.3% 2.8% Service and sales workers 479,199 20.2% 1.9% 92.5% 2.0% 5.5% Skilled agricultural, forestry and fishery workers 53,533 2.3% -28.0% 94.1% 1.8% 4.1% Craft and related trades workers 245,237 10.3% -2.4% 91.4% 4.7% 3.9% Plant and machine operators, and assemblers 202,202 8.5% -5.2% 93.2% 2.4% 4.4% Elementary occupations 155,083 6.5% -1.6% 82.9% 5.0% 12.1% Unknown 89,741 3.8% 81.4% 88.2% 4.2% 7.7% Total 2,373,668 100.0% 2.1% 93.2% 2.4% 4.5% Source: StatFin (2019), own calculations and illustration

Provinces with a high share of TCN have a high in “Skilled agricultural, forestry and fishery work- GDP per capita and vice versa. On the one hand, ers” (-28.0 %) and “Clerical support workers” (- prosperous regions are more attractive to mi- 23.6 %; see Table 3). grants; on the other hand, these regions benefit Further analysis by citizenship shows that 2.4 % from a better availability of labor supply, i.e. belong to the group of EU-28 & EFTA (2.4 %) or greater number of working age people. In this con- TCN (4.5 %). EU-28 & EFTA citizens are overrepre- text, one speaks of a so-called cross-fertilization, sented Craft and related trades workers which results from the interaction of these two in “ ” (4.7 Elementary occupations %). factors. %) and “ ” (5.0 TCN tend to work in the same categories (3.9 % Expressing GDP in PPS (purchasing power stand- and 12.1 %) but also in “Service and sales workers” ards) eliminates differences in price levels be- (5.5 %), “Plant and machine operators, and assem- tween countries or regions. There are significant blers” (4.4 %) and “Skilled agricultural, forestry differences in levels of prosperity among Finnish and fishery workers” (4.1 %). Due to a lack of in- provinces. The most prosperous region in Finland formation the category “Unknown” is excluded is Helsinki-Uusimaa (144 %); it is over 51/46/45/28 from the analysis. pp richer than Pohjois- ja Itä-Suomi/Länsi-Su- Research and Innovation omi/Etelä-Suomi/Åland; see Table 3). Following the neoclassical and endogenous Economic structure and entrepreneurship growth theories, technological advance is believed In total, approx. 2.4 million people were employed to be one of the major drivers of economic in 2018 in Finland. This corresponds to an increase growth. From this perspective, there is a growing of +2.1 % compared to 2010. Up to occupational interest to investigate the link between research & groups more than two thirds (69.0 %) work as development (R&D), innovation, entrepreneur- ”Professionals” (20.4 %), “Service and sales work- ship and economic growth achieved by human ers” (20.2 %), “Technicians and associate profes- capital from abroad (EU-28 and non-EU-28). sionals” (18.1 %) and “Craft and related trades A well-known indicator provided to measure wrkers” (10.3 %). The highest growth rates are in achievements of countries or regions in R&D is the groups of “Professionals” (14.3 %) and “Man- GERD, i.e. regional/national gross domestic ex- agers” (14.0 %) whereas “Technicians and associ- penditures on R&D as a percentage of GDP. GERD ate professionals” (2.3 %) and “Service and sales is estimated at 2.79 % for Finland (2019) which is workers” (1.9 %) evolved at average level. All re- a slight increase of 0.06 percentage points to main groups diminished with the highest declines 2017. The total amount of research expenditures was € 6.72 billion, the largest share (54,5 %) was

12

financed by the business enterprise sector. 28.0 % Despite the relevance of migration for labor sup- were financed by the government sector. 15.4 % ply, the employment rate of TCN is much lower were financed by foreign countries (rest of the compared to Finnish and EU-28 citizens. One rele- world) and 1.8 % were funded by the private non- vant factor to be considered here is the lower ed- profit sector (Eurostat (2020u)). ucational attainment level of migrants, with the consecutive effects of lower income levels and a Data on regional level (NUTS 2) is published with a higher risk of poverty. one year time lag. Helsinki-Uusimaa is the prov- ince with the highest ratio in 2018 (3.48 %) suc- The pattern of education of TCN differs substan- ceeded by Pohjois- ja Itä-Suomi (2.56 %) and Länsi- tially from that of Finnish nationals: The foreign Suomi (2.54 %). Etelä-Suomi (1.72 %) takes fifth population of third countries is disproportionately and Åland (only 0.41 %) last place, see Table 5). represented in primary education, whereas they are underrepresented in secondary and tertiary Table 5: Intramural R&D expenditure as percentage of gross domestic product (GDP) 2017 educational levels. Expenditures % of Province From a spatial point of view, the concentration of (€ mio.) GDP Manner-Suomi 6,432.3 2.76% better-educated migrants is highest in Finnish cit- Länsi-Suomi 1,320.8 2.54% ies. Provinces with a higher ratio of TCN are also Helsinki-Uusimaa 3,177.2 3.48% characterized by higher economic growth, being Etelä-Suomi 756.1 1.72% more attractive for TCN on the one hand and of- Pohjois- ja Itä-Suomi 1,178.2 2.56% fering more labor supply on the other hand. Åland 5.6 0.41% Finland 6,437.7 2.76% The below average regional GVA growth rate, es- Source: Source: Eurostat (2020t), own illustration pecially in rural provinces like Etelä-Suomi is indi- Conclusion rectly linked to immigration from abroad but pref- erably to internal migration from rural areas to cit- This Statistical Briefing examines the economic ies and metropolitan areas. These emigration and spatial impact of immigration with special re- flows tend to reinforce the erosion of the eco- gard to third-country nationals (TCN) in Finland. nomic basis, because they have a negative impact On January 1, 2020, a total of 267,629 persons on the potential workforce and the economic at- with non-Finnish citizenship were living in Finland. tractiveness of the region. Economic performance This corresponds to a share of 4.8 % of Finland's is stagnating, accompanied by declining popula- total population. Among non-Finnish nationals, tion figures. 37.8 % (1.8 % of total population) came from EU- Governments need to invest in education and in 28 (incl. UK) and EFTA countries (0.03 %). 3.0 % of research and development (R&D) to enable long- total population were third country nationals, be- term prosperity and growth. Currently the na- ing individuals who are neither Finnish, EU-28 citi- tional gross domestic expenditures on R&D (as a zens nor EFTA citizens. By comparison, the share percentage of GDP) are estimated at 2.79 % for of TCN in EU-27 (excl. UK) and EU-28 (incl. UK) is Finland (2019) which is a slight increase of 5.7 % and 3.8 %, therefore Finland lying below the 0.06 percentage points to 2017. But there is again average. a striking regional variation from densely popu- Within the last 20 years the working-age popula- lated areas with a high number of innovative en- tion in Finland did already decrease. Although all terprises (e.g. Helsinki-Uusimaa) to less densely regions recorded an immigration of EU-28 and populated provinces. Non-EU-28 residents, this development could not Summing up, immigrants in Finland play a signifi- be stopped due to the comparable high decrease cant role in reviving rural areas, and with the ap- of Finnish people in working-age. propriate policies, they could play an even more significant role in sustaining them.

13

Bibliography and source references

Becker, Gary Stanley (1964): Human Capital: A theoretical Eurostat (2020n): Mean and median income by educational and empirical analysis, with special reference to education. attainment level - EU-SILC survey, http://appsso.euro- 3rd edition. New York. stat.ec.europa.eu/nui/show.do?lang=en&dataset=ilc_di08

CEC (2020). Cohesion Policy and Finland, https://ec.eu- Eurostat (2020o): Mean and median income by broad group ropa.eu/regional_policy/en/information/publications/fact- of citizenship (population aged 18 and over), sheets/2014/cohesion-policy-and-finland https://appsso.eurostat.ec.europa.eu/nui/show.do?da- taset=ilc_di15&lang=en Eurostat (2020a): Population on 1 January by age group, sex and citizenship, https://appsso.eurostat.ec.eu- Eurostat (2020p): Persons aged 18 and over by risk of pov- ropa.eu/nui/show.do?dataset=migr_pop1ctz&lang=en erty, material deprivation, work intensity of the household and most frequent activity status of the person - intersec- Eurostat (2020b): Population change - Demographic balance tions of Europe 2020 poverty target indicators, and crude rates at regional level (NUTS 3), https://appsso.eurostat.ec.europa.eu/nui/show.do?da- https://appsso.eurostat.ec.europa.eu/nui/show.do?da- taset=ilc_pees02&lang=en taset=demo_r_gind3&lang=en Eurostat (2020q): GDP and main components (output, ex- Eurostat (2020c): Population by sex, age, citizenship, labour penditure and income), https://appsso.eurostat.ec.eu- status and NUTS 2 regions, https://appsso.eurostat.ec.eu- ropa.eu/nui/show.do?dataset=nama_10_gdp&lang=en ropa.eu/nui/show.do?dataset=lfst_r_lfsd2pwn&lang=en Eurostat (2020r): Gross domestic product (GDP) at current Eurostat (2020d): Population on 1st January by age, sex, type market prices by NUTS 2 regions, https://appsso.euro- of projection and NUTS 3, https://appsso.eurostat.ec.eu- stat.ec.europa.eu/nui/show.do?da- ropa.eu/nui/show.do?dataset=proj_19rp3 taset=nama_10r_2gdp&lang=en Eurostat (2020e): City, Eurostat (2020s), Real growth rate of regional gross value https://ec.europa.eu/eurostat/statistics- added (GVA) at basic prices by NUTS 2 regions - percentage explained/index.php?title=Glossary:City change on previous year, https://appsso.eurostat.ec.eu- Eurostat (2020f): Town or suburb, https://ec.europa.eu/eu- ropa.eu/nui/show.do?dataset=nama_10r_2gvagr&lang=en rostat/statistics-explained/index.php?title=Glos- Eurostat (2020t): GERD by sector of performance and source sary:Town_or_suburb of funds, http://appsso.eurostat.ec.eu- Eurostat (2020g): Urban centre, ropa.eu/nui/show.do?dataset=rd_e_gerdfund&lang https://ec.europa.eu/eurostat/statistics- Eurostat (2020u): Intramural R&D expenditure (GERD) by explained/index.php?title=Glossary:Urban_centre NUTS 2 regions, https://ec.europa.eu/eurostat/data- Eurostat (2020h): Rural area, https://ec.europa.eu/euro- browser/view/tgs00042/default/table?lang=en stat/statistics-explained/index.php/Glossary:Rural_area Government of Åland Islands (2020). Åland och EU, Eurostat (2020i): Rural grid cell, https://ec.europa.eu/euro- https://www.regeringen.ax/aland-omvarlden/aland-eu stat/statistics-explained/index.php?title=Glossary:Ru- Lehmer, Florian; Ludsteck, Johannes (2013): Lohnanpassung ral_grid_cell von Ausländern am deutschen Arbeitsmarkt: Das Eurostat (2020j): Population by educational attainment level, Herkunftsland ist von hoher Bedeutung, IAB-Kurzbericht, sex, age, citizenship and degree of urbanization, 01/2013), Nuremberg. http://appsso.eurostat.ec.europa.eu/nui/show.do?da- Rauhut, D. and Costa, N. (2021) Territorial Cohesion in Den- taset=edat_lfs_9916&lang=en mark, Finland, Norway and Sweden: Measuring the Impact Eurostat (2020k): Unemployment rates by sex, age, educa- 2007 and 2017. Danish Journal of Geography Doi: tional attainment level and NUTS 2 regions (%), 10.1080/00167223.2021.1920444 https://appsso.eurostat.ec.europa.eu/nui/show.do?da- StatFin (2019): 115t -- Employed persons by occupational taset=lfst_r_lfu3rt&lang=en group (Classification of Occupations 2010, levels 1 to 2), Eurostat (2020l): Employment rates by sex, age, educational background country, sex and year, 2010-2018, attainment level, citizenship and NUTS 2 regions, https://pxnet2.stat.fi/PXWeb/pxweb/en/StatFin/Stat- https://ec.europa.eu/eurostat/data- Fin__vrm__tyokay/statfin_tyokay_pxt_115t.px/ browser/view/tepsr_wc140/default/table?lang=en

Eurostat (2020m): Early leavers from education and training by sex and citizenship, http://appsso.eurostat.ec.eu- ropa.eu/nui/show.do?dataset=edat_lfse_01&lang=en

14

ECONOMIC IMPACT OF MIGRATION

Statistical Briefings Germany

Call: H2020-SC6-MIGRATION-2019

Work Programmes:

H2020-EU.3.6.1.1. The mechanisms to promote smart, sustainable and inclusive growth H2020-EU.3.6.1.2. Trusted organisations, practices, services and policies that are necessary to build resilient, inclusive, participatory, open and creative societies in Europe, in particular taking into account migration, integration and demographic change

Deliverable 4.2 – Statistical briefing for Germany on economic impact of TCNs in MATILDE re- gions

Authors: Birgit Aigner-Walder, Albert Luger, Rahel M. Schomaker with contributions from Tobias Weidinger, Stefan Kordel and Lukas Schorner

Approved by Work Package Manager of WP4: Simone Baglioni, UNIPR Approved by Scientific Head: Andrea Membretti, UEF Approved by Project Coordinator: Jussi Laine, UEF

Version: 28.05.2021

DOI: 10.5281/zenodo.4817376

This document was produced under the terms and conditions of Grant Agreement No. 870831 for the European Commission. It does not necessary reflect the view of the European Union and in no way anticipates the Commission’s future policy in this area.

2

Introduction

This document presents the results of an assessment of the impact of migration and a measure- ment of the contribution provided by ‘third country nationals’ (TCNs) to the economic systems in the receiving contexts. In total, 10 statistical briefings for the MATILDE regions have been compiled, including the following countries: Austria, Bulgaria, Finland, Germany, Italy, Norway, Spain, Sweden, Turkey and the United Kingdom. Each document focuses on the following dimen- sions of economic development: economic growth, labour markets, innovation, and entrepre- neurship. Moreover, a short introduction on the relevance of TCNs in the respective country as well as the population development and structure are considered. The results of a literature review and a comparative analysis of the 10 statistical briefings are published separately.

Within the Matilde project there is a special focus on so-called ‘third country nationals’ (TCNs). MATILDE considers TCNs as Non EU citizens, who reside legally in the European Union (EU) and who are the target of EU integration policies. A TCN is “any person who is not a citizen of the European Union within the meaning of Art. 20(1) of TFEU and who is not a person enjoying the European Union right to free movement, as defined in Art. 2(5) of the Regulation (EU) 2016/399 (Schengen Borders Code).“ (European Commission 2020a). According to this definition, nationals of European Free Trade Association (EFTA) member countries Norway, Island, Liechtenstein and Switzerland are not considered to be TCNs. TCNs cover economic migrants, family migrants, stu- dents and researchers, highly skilled migrants, and forced migrants.

3

GERMANY Figure 1: Origin and share of total population of different nationalities in Germany

Source: Eurostat (2020a), own illustration

Third Country Nationals (TCNs) in Germany Bosnians and Herzegovinians (0.24 %). In total, al- most more than one half of TCNs living in Germany On January 1, 2020, a total of 10,398,022 persons belong to these nationalities. Figure 1 gives an with non-German citizenship were living in Ger- overview of the origin and the share of the differ- many. This corresponded to a share of 12.5 % of ent nationalities in Germany. Germany's total population. Among non-German nationals, almost one half (4,454,418 persons; Population Development & Population Structure 6.1 % of total population) came from EU- While the number of inhabitants in Germany has 28 (incl. UK) and EFTA countries (51,411 persons; increased in the past, it is projected to decrease in 0.07 %). A total of 5,892,193 persons (or 7.1 % of the future. As of January 1, 2020, there were total population) were third country nationals 83,166,711 persons registered in Germany. This (TCNs), being individuals who are neither Ger- corresponds to an increase of 1,003,236 citizens mans, EU-28 citizens nor EFTA citizens. By compar- (+1.22 %) within the last 20 years since 2000 (be- ison, the share of TCNs in EU-27 (excl. UK) and EU- fore the eastward enlargement of the EU in 2004). 28 (incl. UK) was 5.7 % and 3.8 %, therefore Ger- many was above the average. The main cause of population growth is immigra- tion to Germany (EU-28, EFTA & third countries) as Considering the nationalities of TCNs, Turks the birth rate is only slightly higher than the death (1.59 %) are above Syrians (0.91 %), Russians rate of the domestic population. (0.28 %), Afghans (0.27 %), Iraqis (0.27 %) and

4

A long-term analysis of population development other half has shrunk. As can be seen in Figure 3 since the turn of the new millennium reveals clear (a) the working-age population increased in the regional differences. 54.9 % of German NUTS 3-re- southern states and Berlin in particular, while a gions have more inhabitants than in 2000. Pots- sharp decrease occurred in the Eastern States. In dam located near the capital Berlin has the highest general, the German working-age population is re- growth rate (29.1 %). Regions surrounding Munich locating to the regions, cities and metropolitan ar- as well as the former German Democratic Republic eas of Swabia, Upper Bavaria, Darmstadt, (GDR) cities of Dresden and Leipzig followed a sim- Stuttgart, and Trier etc. (NUTS 2). ilar trajectory. Conversely, most of the eastern As Figure 3 (b)-(d) indicate, the increase in the German regions recorded a sharp decline in popu- working-age population within the regions is lation. Relative reductions were highest for Ober- mainly caused by the increase in the non-German spreewald-Lausitz in Brandenburg (26.2%); also population. Without immigration, all German nearby NUTS 3-regions shrunk by more than 20 %. states would have suffered from a downturn in There are three main causes that explain foregone working-age population, with the exception of Ba- developments: (1) a migratory movement from varia with a stagnating development. Summing Eastern to Western Germany took place in the up, many rural areas are experiencing an overall past for the benefit of many regions in the former decline in population, often mainly triggered by German Länder (federal states, NUTS 1). It has be- (re)location trends to cities and metropolitan re- come less significant as net migration has been gions. In particular, this is the case for Eastern Ger- balanced in recent years. (2) Ongoing rural-urban many, which is expected to shrink in the future es- migration increasingly taking place within Eastern pecially in terms of working-age population. Germany and (3) a sharp decline in birth rates af- From a regional perspective, the areas with the ter 1990. highest declines of working-age population are lo- Rural and socio-economically weak regions in cated in Brandenburg, i.e. Frankfurt (Oder) (- Western Germany also experienced a loss in the 22.64 %), Oberspreewald-Lausitz (-21.32 %), number of inhabitants. On the one hand, these ar- Spree-Neiße (-19.06 %), Uckermark (-18.72 %) and eas are located far away from prosperous regions Cottbus (-17.14 %). These regions, often consid- and their centers or, on the other hand, popula- ered peripheral, are characterised by the declining tion shrank as a result of structural change (e.g. coal industry. They suffer from higher depopula- Ruhr region; see Figure 2). tion and birth rate deficits.

While most discussions about changing demo- Analyzing working-age population developments graphic structures focus on population ageing and up to 2040, solely the metropolitan areas are ex- the sustainability of pension systems, urbanization pected to grow, though there are differences be- etc., one prominent field lies on the relevance of tween the eastern and western part of Germany. labour supply, i.e. the population aged While the labour force potential will rapidly grow 15 to 64 years or 20 to 64 years in industrialized in the West, e.g. Cologne, Dortmund, Düsseldorf economies and nations. From an economic per- or Frankfurt, and the South, e.g. Augsburg, Mu- spective, labour (done by human beings) is an es- nich, Nuremberg, Stuttgart, etc. the growth will be sential element in the production of goods and restrained in the East, e.g. Berlin, Dresden, Leipzig services. The term ‘labour force’ comprises all of or Magdeburg (see Figure 4). those who work for gain, whether as employees, These past and future developments underline the employers, or as self-employed, and it includes the high relevance of migration, especially for rural re- unemployed who are seeking for work. gions, concerning the development of the work- Analyzing the working-age population at NUTS 1 ing- age population. Without immigration, its pop- level, the strong dispersion is striking. The median ulation between 20 and 64 years would shrink, German working-age population development is even more dramatically, representing a challenge 0.0 %, i.e. half of the Länder has grown, and the from an economic point of view.

5

The total-age dependency ratio is a measure of the low child ratios, most metropolitan areas and cit- age structure of the population. It relates the num- ies in Germany are characterized by relatively low ber of individuals who are likely to be ‘dependent’ rates (e.g., Münster: 53.4 %). High dependency ra- on the support of others for their daily living – the tios, in contrast, are found in rural regions and re- young (up to 19 years old) and the elderly gions where the share of children and the elderly (65 plus years old) – to the number of those indi- (65 years and older) is above average. These re- viduals who are, being working age from 20 to 64 gions are mainly found in Saxony (e.g., Görlitz years old, capable of providing this support. 87.1 %). As a result of ageing societies, the total- age dependency ratios are expected to rise in the The total-age dependency ratio in 2020 in Ger- future by 20.2 pp to 87.5 % in 2040 resulting in im- many is 67.3 %. Due to prospering economy and plications for publicly funded social security schemes (e.g., fiscal policy).

Figure 2: Population Development (Δ 2000 to 2020), NUTS 3

Source: Eurostat (2020b), own illustration

6

Figure 3: Working-age Population (20 to 64 years; Δ 2006 to 2019) (a) (c)

(b) (d)

Source: Eurostat (2020c), own illustration

7

Figure 4: Projections of Working-age Population (20 to 64 years; Δ 2020 to 2040) NUTS 3

Source: Eurostat (2020d), own illustration Education and (Un)Employment bourg (41.0 %), Ireland (40.7 %) and United King- dom (40.6 %) rank at the top end, while the Czech The tertiary education rate of Germany's domestic Republic (21.6 %), Italy (17.4 %) and Romania population is 26.7 %, which is only slightly above (16.0 %) rank at the bottom. that of the total population (26.0 %). Compared to the EU-28 average (29.5 %), Germany has a below- Across the world, educational attainment is signif- average tertiary education rate. Germany ranks icantly higher in cities1 than in towns and suburbs2, 20th within the EU-28, ahead of Bulgaria (24.7 %), which in turn is higher than in rural areas. In rural Portugal (23.8 %) and Slovakia (23.1 %). Luxem-

1 A city is a local administrative unit (LAU) where at least 50 population lives in urban centres, i.e. a cluster of contiguous % of the population lives in one or more urban centres (Eu- grid cells of 1 km² (excluding diagonals) with a population rostat 2020e). density of at least 1 500 inhabitants per km² and collectively 2 Towns & suburbs are areas where less than 50 % of the a minimum population of 50 000 inhabitants after gap-filling population lives in rural grid cells and less than 50 % of the (Eurostat 2020f; Eurostat 2020g).

8

areas3 in Germany, only 21.2 % of individuals have From a regional perspective, Berlin has the highest university degrees, i.e. tertiary education unemployment rate at 5.2 % (female: 4.8 %; (ISCED2011 levels 5 to 8), compared to 23.3 % in male: 5.6 %), followed by Bremen (4.9 %; fe- towns & suburbs and 31.7 % in cities. In contrast, male: n.a. %; male: 6.4 %) and Saxony-Anhalt secondary degrees are more common in less (4.6 %; female: 4.2 %; male: 5.1 %). Bavaria (2.0 %; densely populated areas. While less than 48.2 % of female: 1.9 %; male: 2.1 %), Baden-Wuerttemberg city residents have secondary, i.e. upper second- (2.3 %; female: 2.1 %; male: 2.5 %) and Rhineland- ary and post-secondary non-tertiary education Palatinate (2.6 %; female: 2.3 %; male: 2.9%) are (ISCED2011 levels 3 to 4) attainment level, 56.2 % the provinces with the lowest unemployment and 62.1 % of residents in towns & suburbs and rates. For Trier, there is no data available (see Fig- rural areas do so. ure 5). In total, the unemployment rate of Ger- many is 3.1 % (female: 2.7 %; male: 3.5%). In cities, where most of the non-EU-28 citizens live, more than 40 % have only primary education, Figure 5: Unemployment rate 2019 (20 to 64 years) i.e. less than primary, primary and lower second- ary education (ISCED levels 0 to 2, see Table 1). This share is even higher in towns & suburbs (51.2 %) and rural areas (52.4 %). On contrary, only 24.6 % of non-EU-28 citizens have a degree from university but this value is still much higher than in suburban (15.7 %) and rural (15.4 %) re- gions. Differences of educational attainment level between Germans and EU-28 citizens range within a broad band from 2.0 pp to 17.3 pp and are there- fore scarcely comparable (see Table 1). In total, EU-28 (23.2 %) and non-EU-28 (20.7 %) citizens have a lower tertiary education rate than domestic population (26.7 %).

Table 1: Working-age population by citizenship, degree of urbanization and educational attainment level 2019 Educational Attainment Level Citizenship (ISCED11) isation isation Urban- Urban- Primary Secondary Tertiary Germans 15.4% 51.5% 33.1% EU-28 30.6% 41.2% 28.2%

Cities Cities Non EU-28 43.5% 31.8% 24.6% Source: Eurostat (2020k), own illustration Total 20.1% 48.2% 31.7% Germans 16.9% 58.8% 24.3% In general, there is a correlation between the level EU-28 34.2% 47.3% 18.5% of education and unemployment, according to Non EU-28 51.2% 33.0% 15.7% which the well-educated are significantly less suburbs Towns & Towns Total 20.5% 56.2% 23.3% likely to be unemployed than people without a vo- Germans 15.0% 63.5% 21.5% EU-28 29.8% 50.8% 19.5% cational qualification. This correlation can also be seen for Germany. Those with primary education Rural Non EU-28 52.4% 32.2% 15.4% Total 16.7% 62.1% 21.2% are most affected by unemployment (8.3 %) and Source: Eurostat (2020j), own illustration this situation has remained constant for years. Conversely, people with tertiary education have

3 A rural area is an area where more than 50 % of its popula- tion lives in ‘rural grid cells’, i.e. grid cells that are not identi- fied as urban centres or as urban clusters (Eurostat 2020h; Eurostat 2020i).

9

the lowest unemployment rate (1.9 %), those with Figure 7: Employment rate by educational attainment level secondary a value of 2.7 %. The same trend can be and citizenship (20 to 64 years; ISCED 2011) observed at the European level (EU-28 average; 100 90.7 see Figure 6), even though with higher unemploy- 90 82.9 82.0 85.9 ment rates. 80 73.4 70 63.9 66.6 67.9 Figure 6: Unemployment rate by educational attainment 60 51.1 level (20 to 64 years; ISCED 2011) 50 40 14 30 12.4 20 12 10 0 10 Primary Secondary Tertiary 8.3 8 EU-28 Non EU-28 Germany 5.6 6 Source: Eurostat (2020l), own illustration 4.0 4 2.7 This theory and the empirical evidence can also be 1.9 2 applied to the ‘Early Leavers from Education and Training’. This population group refers to persons 0 aged 18 to 24 who have completed a lower sec- Primary Secondary Tertiary ondary education and are not involved in further education or training. From a fiscal point of view, EU-28 Germany this specific group is of particular relevance as they

Source: Eurostat (2020k), own illustration are more likely to be (long-term) unemployed. In 2019, 10.3 % of the 18-24 year olds in the EU-28 In contrast, the higher the educational attainment were part of this group (male: 11.9 %; female: the higher the employment rate, i.e. 89.0 % for 8.6 %). With a value of 10.3 %, Germany ranges at tertiary, 81.3 % for secondary and 61.8 % for pri- the European average. The share of EU-28 citizens mary level. This observation holds for other citi- residing in Germany is more than double (24.9 %) zenships, too, as the risk of unemployment de- and that of non-EU-28 citizens is also more than creases by the level of educational attainment. But double (25.7 %) than that of the domestic popula- Germans have the highest employment rate for tion (7.6 %). In total, approx. 50 % (47.6 %) of all tertiary educational attainment level (90.7 %) early leavers are jobless but because of missing while EU-28 citizens rank first for primary (73.4 % data no relying statements can be made with re- vs. 63.9 %) and secondary educational attainment gard to the immigration status (Eurostat 2020m). level (82.9 % vs. 82.0 %). Non-EU-28 citizens have the lowest employment rate on each level (67.9 %; Income and Gross Domestic Product 66.6 %; 51.1 %; see Figure 7). As for the employment rate, the educational at- The abovementioned findings foster the theoreti- tainment level has a positive impact on mean and cal conclusions by Becker (1964) that the educa- median4 equivalized net income5, i.e. the higher tional level has a significant influence on the pro- the educational attainment level the higher the in- fessional career and the risk of unemployment of come. human beings though intersectional aspects (e.g., Working persons in Germany at primary level earn gender, race, region) have to be taken into ac- less (mean: € 19,521; median: € 18,136) than sec- count as well. ondary (€ 25,326 or € 23,699) and tertiary levels (€

4 The median is more robust, i.e. less sensitive to outliers than 5 Geis-Thöne (2019) criticises using net income instead of the mean. This explains why the mean equivalized income is gross income in this context, as taxes and social security con- always slightly higher for each educational level. tributions heavily depend on family structures.

10

32,907 or € 29,350; see Figure 8). Working Ger- capital (training on the job) as the median age of mans earn less than EU-28 but more than non-EU- foreign workers and therefore the length of time 28 citizens (mean: € 27,645 vs. € 29,841 and within the company is lower than for domestic € 21,777; median: € 24,972 vs. € 26,729 and ones. However, these assumptions cannot be ver- € 19,355; see Figure 9). ified in the context of this analysis due to a lack of available data. Figure 8: Mean and median equivalized net income by edu- cational attainment level 2018 (18 to 64 years; ISCED 2011) The indicator ‘people at risk of poverty or social

35,000 32,907 exclusion’ corresponds to the sum of persons who are: at risk of poverty after social transfers, se- 29,350 30,000 verely materially deprived or living in households 25,326 25,000 23,699 with very low work intensity. As non-EU-28 citi- 19,521 zens have a higher risk of unemployment and they 20,000 18,136 earn less than Germans and EU-28 citizens, more 15,000 than one quarter (28.4 %) belongs to this group whereas the values of EU-28 (14.2 %) and domes- 10,000 tic population (17.8 %) are comparably low (see 5,000 Table 2; cf. Weidinger & Kordel 2021). 0 Primary Secondary Tertiary In 2019, compensation of employees (wages and salaries plus employers’ social contributions) was Mean equivalised net income the largest income component of GDP in EU-28 Median equivalised net income and the Euro area, accounting for 47.8 % and Source: Eurostat (2020n), own illustration 48.0 %. Germany (53.5 %) lies above European av- Figure 9: Mean and median equivalized net income by erage (both EU-28 and Euro area). Taxes on pro- broad group of citizenship 2018 (18 to 64 years) duction and imports (less subsidies) accounted for 35,000 9.9 % (EU-28: 11.9 %; Euro area: 11.5 %). Gross 29,841 operating surplus and mixed income accounted 30,000 26,729 27,645 for 36.6 % of GDP in Germany, 40.3 % of GDP for 24,972 25,000 the EU-28 and 40.5 % in the Euro area. German 21,777 19,355 GDP at current prices amounted to approx. 20,000 € 3,449.0 billion in 2019 and GDP per inhabitant 15,000 equaled € 41,510 (Eurostat 2020q).

10,000 Table 2 provides information about nominal re- gional GDP and real growth rates, whereas the 5,000 federal states (NUTS 1) are grey-shaded. Within 20 0 years Germany’s gross value added (GVA) in- EU-28 Non EU-28 Germany creased by 23.5 % (real growth rate). In terms of regional GVA all NUTS 2 regions also recorded pos- Mean equivalised net income itive real growth, ranging from 11.0 % in Saarland Median equivalised net income to 34.8 % in Leipzig since 2000. The highest re- Source: Eurostat (2020o), own illustration gional GDP per capita at current prices occurred in A further analysis and decomposition of wage dif- the northern provinces like Hamburg (approx. ferentials between domestic and foreign workers € 67,300) succeeded by regions in the south (Up- done by Lehmer and Ludsteck (2013) reveals that per Bavaria; approx. € 59,700). The capital of Ber- factors like seniority, promotions to better-paid lin ranks also at the top (approx. € 42,300). As in occupations as well as job stability also have to be previous years, the eastern provinces of the for- taken into account. These factors cause an im- mer German Democratic Republic (GDR) were far provement in an individual's firm-specific human below the national level of € 41,500 (Saxony-An-

11

Table 2: Population aged 18 and over by risk of poverty, material deprivation, work intensity 2019 Risk of poverty At risk Not at risk At risk Not at risk At risk Not at risk Work Intensity Material Deprivation Germany EU-28 Non EU-28 Very low Severe 0.8 0.1 1.1 0.2 0.5 2.7 Not very low Severe 1.1 0.7 0.3 0.8 1.8 1.9 Very low Non severe 2.8 1.6 2.5 1.5 5.2 1.1 Not very low Non severe 10.7 7.8 15.2 People at Risk of Poverty or Social Exclusion 17.8 14.2 28.4 Not at risk Not at risk Not at risk Not very low Non severe 82.2 85.8 71.6 People at no Risk of Poverty or Social Exclusion 82.2 85.8 71.6 Total 100.0 100.0 100.0 Source: Eurostat (2020p), own illustration

Table 3: Nominal regional GDP (2019) Lower Saxony 307,510 25.8 38,500 112 and real growth rate (Δ 2000) Brunswick 80,135 34.1 50,200 146 Total Per Capita Hanover 85,382 19.6 39,700 115 Lüneburg 49,594 22.0 28,900 84 Weser-Ems 92,399 26.9 36,500 106

North Rhine- 715,854 17.8 39,900 116

Westphalia Düsseldorf 221,024 15.3 42,500 123

Cologne 197,110 22.3 44,100 128 Münster 89,552 18.2 34,100 99

Region Detmold 80,219 18.8 39,000 113

GDP (in €) GDP (in Arnsberg 127,949 13.9 35,700 104 Rhineland- current marketcurrent prices

GDP (in €) €) per capita GDP (in 145,329 18.1 35,500 103

at current marketat current prices at Palatinate 2015=100; Δ 2002 to 2019) 2002 to Δ 2015=100; Koblenz 51,080 16.6 34,100 99 of the EU27 (from 2020) average EU27 the 2020) of (from

added (GVA) at basic prices (Index, prices at basic (GVA) added Trier 16,494 19.8 31,000 90 Purchasing power standard (PPS, EU27 (PPS, standard power Purchasing Real growth rate of regional gross value gross value growth rate regional Real of from 2020), per inhabitant in percentage per inhabitant from 2020), Rhine-Hesse- Baden- 77,755 18.6 37,700 110 525,198 27.2 47,400 137 Palatinate W uerttemberg Saarland 36,383 11.0 36,800 107 Stuttgart 224,062 28.2 54,000 157 Sa xony 128,625 25.8 31,600 92 Karlsruhe 128,375 24.8 45,700 133 Dresden 51,383 26.4 32,200 93 Freiburg 89,751 24.9 39,600 115 Chemnitz 42,197 17.9 29,500 86 Tübingen 83,009 30.9 44,600 129 Leipzig 35,045 34.8 33,500 97 Ba varia 634,609 31.0 48,400 141 Sa xony-Anhalt 63,725 12.7 28,900 84 Upper Bavaria 280,479 33.6 59,700 173 Schleswig- 98,261 20.1 33,900 98 Lower Bavaria 49,463 30.4 39,800 116 Holstein Upper 48,206 31.6 43,400 126 Th uringia 63,947 21.2 29,900 87 Palatinate Germany 3,449,050 23.5 41,500 120 Upper 41,896 23.9 39,300 114 Source: Eurostat (2020r), own illustration; Franconia Eurostat (2020s), own illustration Middle 82,247 30.0 46,400 135 Franconia halt: € 28,900; Mecklenburg-Hither Pomerania: Lower 55,274 25.1 42,000 122 Franconia € 29,000; Brandenburg: € 29,700; Thuringia: Swabia 77,043 29.9 40,700 118 € 29,900; Saxony: € 31,600). Except for Hamburg Berlin 154,673 32.7 42,300 123 (67,300), Bremen (€ 49,500), Bavaria (€ 48.400) Brandenburg 74,838 22.6 29,700 86 Baden- Wuerttemberg (€ 47,400), Hesse Bremen 33,779 16.0 49,500 144 (€ 47,100) and Berlin (€ 42,300) all western prov- Hamburg 124,076 23.8 67,300 195 Hesse 295,533 16.5 47,100 137 inces rank below German average. Darmstadt 212,495 16.4 53,000 154 Provinces with a high share of TCNs have a high Giessen 37,151 17.0 35,500 103 Kassel 45,887 16.4 37,600 109 GDP per capita and vice versa. On the one hand, Mecklenburg- prosperous regions are more attractive to mi- Hither 46,711 18.2 29,000 84 grants; on the other hand, these regions benefit Pomerania from a better availability of labour supply, i.e. a

12

greater number of working-age people. In this Economic structure and entrepreneurship context, one speaks of a so-called ‘cross-fertiliza- In general, a distinction is made between three dif- tion’, which results from the interaction of these ferent sectors of the economy, the primary (agri- two factors. culture and forestry), secondary (manufacturing) Expressing GDP in PPS (purchasing power stand- and tertiary (services) sectors. ards) eliminates differences in price levels be- In total, approx. 33.7 million people were em- tween countries or regions. There are significant ployed (employees are subject to social insurance differences in levels of prosperity among German contributions) in . This corre- regions. The most prosperous region in Germany sponds to an increase of +12.9 % compared to is Hamburg (195 %); it is over 111/109/108 pp 2013. Almost three quarters (71.2 %) are em- richer than Mecklenburg-Hither Pomerania and ployed in the service sector (24.0 million persons), Saxony-Anhalt/ Brandenburg/ Thuringia. Upper where employment has increased by +15.1 % Bavaria ranks second (173 %) and Stuttgart third since 2013. Employment in manufacturing evolved (157 %; see Table 3). below average (+7.8 %) and though primary sector increased by +8.2 %, it is of very low importance, i.e. only 0.7 % of total employment. Approx. 0.005 % belong to the category ‘Others’ (see Table 4).

Table 4: Employment by economic activity 2019 Employed Economic Activity persons in Share Δ 13-19 Germans Foreigners 2019 Primary Sector 225.67 0.7% 8.2% 82.9% 17.1% Agriculture, forestry and fishing 225.67 0.7% 8.2% 82.9% 17.1% Secondary Sector 9,474.13 28.1% 7.8% 87.9% 12.1% Mining and quarrying 64.07 0.2% -19.0% 93.7% 6.3% Manufacturing 7,017.34 20.8% 6.3% 89.0% 10.9% Electricity, gas, steam and air conditioning supply 239.61 0.7% 1.3% 96.7% 3.3% Water supply, sewerage, waste management 260.35 0.8% 12.6% 90.9% 9.1% Construction 1,892.78 5.6% 15.5% 81.8% 18.1% Tertiary Sector 24,038.63 71.2% 15.1% 87.3% 12.6% Wholesale and retail trade, repair of motor vehicles 4,581.30 13.6% 8.6% 89.4% 10.6% Transportation and storage 1,868.85 5.5% 22.1% 79.3% 20.6% Accommodation and food service activities 1,087.15 3.2% 23.2% 65.2% 34.6% Information and communication 1,162.38 3.4% 27.9% 89.1% 10.8% Financial and insurance activities 972.78 2.9% -4.0% 95.6% 4.4% Real estate activities 280.41 0.8% 21.0% 91.6% 8.4% Professional, scientific and technical activities 2,323.63 6.9% 24.1% 91.0% 9.0% Administrative and support service activities 2,276.60 6.7% 15.1% 70.7% 29.2% Public administration and defense 1,872.72 5.6% 9.3% 97.1% 2.9% Education 1,347.39 4.0% 15.9% 92.4% 7.6% Human health and social work activities 5,044.47 15.0% 18.4% 91.8% 8.2% Arts, entertainment and recreation 300.87 0.9% 21.3% 86.6% 13.4% Other service activities 853.29 2.5% 6.2% 88.8% 11.2% Activities of households as employers, undifferentiated 49.06 0.1% 12.9% 73.3% 26.6% goods- and services Activities of extraterritorial organizations and bodies 17.73 0.1% -17.9% 80.3% 19.5% Others 1.69 0.0% -67.4% 76.9% 22.8% Others 1.69 0.0% -67.4% 76.9% 22.8% Total 33,740.12 100.0% 12.9% 87.4% 12.5% Source: Federal Employment Agency (Bundesagentur für Arbeit Statistik) (2014, 2020a), own calculations and illustration

13

A more detailed examination of the different sec- A further analysis by citizenship shows that almost tors of the economy reveals that there have also 9 out of 10 employees of the secondary (87.9 %) been significant shifts within the sectors: the only and tertiary sector (87.3 %) are Germans. Con- declines in employment were recorded in the sec- versely, non-Germans account for 12.1 % (second- tors "Mining and quarrying" (-19.0 %), “Activities ary) and 12.6 % (tertiary); due to a lack of data no of extraterritorial organizations and bodies" (- statements can be made in regard to TCNs. 17.9 %) and “Financial and insurance activities” (- Foreigners, i.e. all persons with non-German citi- 4.0 %). " Information and communica- zenship, are overrepresented (above 12.5 %) in tion" (+27.9 %), " Professional, scientific and tech- “Accommodation and food service activities” nical activities" (+24.1 %), “Accommodation and (34.6 %), “Administrative and support service ac- food service activities” (+23.2 %), " Transportation tivities (29.2 Activities of households as em- and storage" (+22.1 %), "Arts, entertainment and ” %), “ ployers, undifferentiated goods- and services recreation" (+21.3 %) as well as "Real estate activ- ” (26.6 Transportation and storage (20.6 %), ities" (+21.0 %), on the other hand, are among %), “ ” Activities of extraterritorial organizations and those economic sections with the largest absolute “ bodies (19.5 (18.1 Agri- growth. ” %), “Construction” %), “ culture, forestry and fishing” (17.1 %) and “Arts, entertainment and recreation” (13.4 %).

Figure 10: Employment by economic sector and citizenship 2019 (a) Secondary Sector (b) Tertiary Sector 0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100%

Baden- Baden- 83.4% 9.4% 7.2% 83.4% 9.3% 7.4% Wuerttemberg Wuerttemberg Bavaria 86.5% 7.9% 5.6% Bavaria 83.7% 9.3% 7.0%

Berlin 85.4% 7.3% 7.3% Berlin 84.0% 6.9% 9.1%

Brandenburg 94.6% 3.9%1.5% Brandenburg 92.8% 4.5%2.7%

Bremen 89.3% 5.2% 5.5% Bremen 88.2% 4.8% 7.0%

Hamburg 88.9% 5.6% 5.6% Hamburg 86.7% 5.7% 7.5%

Hesse 84.3% 8.4% 7.3% Hesse 83.4% 8.1% 8.5%

Lower Saxony 90.8% 5.4%3.8% Lower Saxony 90.7% 4.7%4.6% Mecklenburg- Mecklenburg- 95.8% 3.0%1.2% 95.9% 2.3%1.9% Hither Pomerania Hither Pomerania North Rhine- North Rhine- 87.7% 6.0% 6.3% 88.1% 5.2% 6.7% Westphalia Westphalia Rhineland- Rhineland- 88.4% 6.2% 5.3% 88.3% 6.1% 5.6% Palatinate Palatinate Saarland 86.5% 9.8% 3.8% Saarland 87.8% 7.2% 5.0%

Saxony 95.2% 3.5%1.3% Saxony 94.3% 3.2%2.5%

Saxony-Anhalt 95.8% 2.6%1.6% Saxony-Anhalt 95.6% 2.1%2.3% Schleswig- Schleswig- 93.1% 3.2%3.7% 92.4% 3.6%4.0% Holstein Holstein Thuringia 95.1% 3.3%1.6% Thuringia 93.9% 3.3%2.8%

Germany 87.9% 6.7% 5.4% Germany 87.4% 6.4% 6.2%

Germany EU-28 Non EU-28 Germany EU-28 Non EU-28

Source: Federal Employment Agency (Bundesagentur für Arbeit Statistik) (2020b), own illustration

14

Mecklenburg-Hither Pomerania, Saxony-Anhalt Table 5: Intramural R&D expenditure as percentage of (both 95,8 %), Saxony (95.2 %), Thuringia (95.1 %) gross domestic product (GDP) 2017 Expenditures % of and Brandenburg (94.6 %) have the highest share Province (€ mio.) GDP of Germans working in the secondary sector Baden-Wuerttemberg 27,895.3 5.70% whereas Baden-Wuerttemberg (83.4 %), Hesse Stuttgart 15,918.8 7.69% (84.3 %) and Berlin (85.4 %) have the lowest share. Karlsruhe 5,923.1 4.95% The number of non-EU-28 exceeds the number of Freiburg 2,313.3 2.75% Tübingen 3,740.2 4.78% EU-28 workers in Schleswig-Holstein (3.7 % vs. Bavaria 18,684.3 3.12% 3.2 %), North Rhine-Westphalia (6.3 % vs. 6.0 %), Upper Bavaria 10,695.7 4.03% Bremen (5.5 % vs. 5.2 %), whereas the share is the Lower Bavaria 631.6 1.34% same in Berlin (both 7.3 %) and Hamburg (both Upper Palatinate 1,134.9 2.47% 5.6 %); in the remaining federal provinces EU-28 Upper Franconia 830.8 2.10% Middle Franconia 2,745.7 3.60% outweighs non-EU-28 employees (see Figure 10 Lower Franconia 1,479.3 2.84% (a)). An analysis of the tertiary sector shows that Swabia 1,166.2 1.61% non-EU-28 employees are of more importance as Berlin 4,742.5 3.41% the gap between them and EU-28 is only 0.2 pp Brandenburg 1,192.4 1.69% (vs. 1.3 pp in the secondary sector). The highest Bremen 907.4 2.79% Hamburg 2,495.8 2.16% gap is in Bavaria (2.3 pp), Saarland (2.2 pp), Baden- Hesse 8,174.5 2.93% Wuerttemberg (1.9 pp) and Brandenburg (1.8 pp; Darmstadt 6,561.1 3.27% see Figure 10 (b)). Giessen 901.1 2.61% Kassel 712.3 1.63% Research and Innovation Mecklenburg-Hither Pomera- 784.0 1.81% nia Following the neoclassical growth and endoge- Lower Saxony 8,921.3 3.13% nous growth theories, technological advance is be- Brunswick 5,876.0 8.52% lieved to be one of the major drivers of economic Hanover 1,812.2 2.25% Lüneburg 425.9 0.90% growth. From this perspective, there is a growing Weser-Ems 807.2 0.91% interest to investigate the link between research North Rhine-Westphalia 14,319.2 2.11% and development (R&D), innovation, entrepre- Düsseldorf 4,121.6 1.96% neurship and economic growth achieved by hu- Cologne 5,523.8 2.97% man capital from abroad (EU-28 and non-EU-28 Münster 1,043.6 1.22% Detmold 1,495.8 1.98% citizens). Arnsberg 2,134.4 1.76% A well-known indicator provided to measure Rhineland-Palatinate 3,495.1 2.46% Koblenz 343.6 0.69% achievements of countries or regions in R&D is Trier 142.4 0.88% GERD, i.e. regional/national gross domestic ex- Rhine-Hesse-Palatinate 3,009.0 3.97% penditure on R&D as a percentage of GDP. GERD Saarland 618.6 1.76% is estimated at 3.12 % for Germany (2018) which Saxony 3,393.7 2.81% is a slight increase of 0.07 percentage points to Dresden 1,990.8 4.13% Chemnitz 750.5 1.86% 2017. The total amount of research expenditures Leipzig 652.5 2.01% was € 100.67 billion, the largest share was fi- Sa xony-Anhalt 916.2 1.50% nanced by the business enterprise sector (66.0 %). Schleswig-Holstein 1,447.0 1.56% 27.9 % were financed by the government sector. Th uringia 1,358.8 2.21% Germany 99,553.6 3.05% 5.8 % were financed by foreign countries (rest of Source: Eurostat (2020u), own illustration the world) and 0.3 % were funded by the private non-profit sector (Eurostat 2020t). second place with 3.41 % just ahead of Lower Sax- ony (3.13 %) and Bavaria (3.12 %). All other fed- Data on regional level (NUTS 2) and federal states eral states were well behind and below the aver- level (NUTS 1) is published with a one-year time age German ratio of 3.05 %. Saxony-Anhalt ranked lag. Baden-Wuerttemberg was the province with last (1.50 %) with Schleswig-Holstein (1.56 %), the highest ratio in 2017 (5.7 %). Berlin followed in Brandenburg (1.69 %), Saarland (1.76 %) and

15

Mecklenburg-Hither Pomerania (1.81 %) only educational attainment level of migrants, with the slightly better. consecutive effects of lower income levels and a higher risk of poverty. The pattern of education of A deeper analysis (on NUTS 2 level) reveals that re- the EU-28 and non-EU-28 citizens differs substan- search and development (R&D) takes predomi- tially from that of Germans: The foreign popula- nantly place in metropolitan areas of large cities as tion of third countries is disproportionately repre- Braunschweig (8.52 %), Stuttgart (7.69 %), Karls- sented in primary and secondary education, ruhe (4.95 %), Tübingen (4.78 %) and Dres- whereas they are underrepresented in tertiary ed- den (4.13 %). In absolute terms of R&D spending, ucation levels. Baden-Wuerttemberg also tops the list: € 27.90 billion of Germany's total research spend- From a spatial point of view, the concentration of ing of € 99.55 billion in 2017 was spent in Baden- better-educated migrants is highest in German cit- Wuerttemberg. It is followed by Bavaria (€ 18.68 ies. Provinces with a higher ratio of TCNs are also billion) and North Rhine-Westphalia (€ 14.32 bil- characterized by higher economic growth, being lion; see Table 5). more attractive for TCNs on the one hand and of- fering more labour supply on the other hand. Available data for other countries has shown that employees in R&D have an above average share of The below average regional GVA growth rate, es- foreigners compared to the total number of em- pecially in rural provinces like Saxony-Anhalt, ployed persons by economic activity. Schleswig-Holstein or Thuringia is indirectly linked to immigration from abroad as well as to out-mi- Conclusion gration from rural areas to cities and metropolitan This Statistical Briefings tried to give an overview areas. These emigration flows tend to reinforce on the economic and spatial impact of immigra- the erosion of the economic basis, because they tion with special regard to third-country nationals have a negative impact on the potential workforce (TCNs) in Germany. and the economic attractiveness of the region. Economic performance is stagnating, accompa- On January 1, 2020, a total of 10,398,022 persons nied by declining population figures. This develop- (12.5 %) with non-German citizenship were living ment does not take place in regions of former GDR in Germany. Among non-German nationals, al- solely but also in certain regions of Western Ger- most one half came from EU-28 (incl. UK) and many. EFTA countries. A total of 5,892,193 persons (or 7.1 % of total population) were third country na- As a result, governments need to invest in educa- tionals (TCNs), being individuals who are neither tion on the one hand and in research and develop- Germans, EU-28 citizens nor EFTA citizens. ment (R&D) on the other, to enable long-term prosperity and growth impulses. Currently, the na- Within the last 20 years the population increase of tional gross domestic expenditures on R&D (as a Germany has mainly be dominated by immigra- percentage of GDP) are estimated at 3.12 % (2018) tion. From 2006 to 2019 all German regions prof- which is a slight increase of 0.07 percentage points ited heavily with regard to its working-age popula- compared to 2017. But there is again a striking re- tion. Without immigration, all German regions gional variation from densely populated areas would have suffered from a decrease in its popu- with high investments and innovative enterprises lation aged between 20 and 65 years (with the ex- (e.g. Baden-Wuerttemberg and Berlin) to less ception of Bavaria with a stagnating develop- densely populated provinces (e.g. Saxony-Anhalt ment). and Schleswig-Holstein).

Despite the relevance of migration for labour sup- Summing up, immigrants in Germany play a piv- ply, the employment rate of TCNs is much lower otal role in reviving areas with a natural popula- compared to Germans and EU-28 citizens, while tion downturn, and with the ‘right’ policies, they we find a much higher unemployment rate. One could play an even more significant role in sustain- relevant factor to be considered here is the lower ing them.

16

Bibliography and source references

Becker, Gary Stanley (1964): Human Capital: A theoretical Eurostat (2020o): Mean and median income by broad group and empirical analysis, with special reference to education. of citizenship (population aged 18 and over), 3rd edition. New York. https://appsso.eurostat.ec.europa.eu/nui/show.do?da- taset=ilc_di15&lang=en Eurostat (2020a): Population on 1 January by age group, sex and citizenship, https://appsso.eurostat.ec.eu- Eurostat (2020p): Persons aged 18 and over by risk of pov- ropa.eu/nui/show.do?dataset=migr_pop1ctz&lang=en erty, material deprivation, work intensity of the household and most frequent activity status of the person - intersec- Eurostat (2020b): Population change - Demographic balance tions of Europe 2020 poverty target indicators, and crude rates at regional level (NUTS 3), https://appsso.eurostat.ec.europa.eu/nui/show.do?da- https://appsso.eurostat.ec.europa.eu/nui/show.do?da- taset=ilc_pees02&lang=en taset=demo_r_gind3&lang=en Eurostat (2020q): GDP and main components (output, ex- Eurostat (2020c): Population by sex, age, citizenship, labour penditure and income), https://appsso.eurostat.ec.eu- status and NUTS 2 regions, https://appsso.eurostat.ec.eu- ropa.eu/nui/show.do?dataset=nama_10_gdp&lang=en ropa.eu/nui/show.do?dataset=lfst_r_lfsd2pwn&lang=en Eurostat (2020r): Gross domestic product (GDP) at current Eurostat (2020d): Population on 1st January by age, sex, type market prices by NUTS 2 regions, https://appsso.euro- of projection and NUTS 3, https://appsso.eurostat.ec.eu- stat.ec.europa.eu/nui/show.do?da- ropa.eu/nui/show.do?dataset=proj_19rp3 taset=nama_10r_2gdp&lang=en Eurostat (2020e): City, Eurostat (2020s), Real growth rate of regional gross value https://ec.europa.eu/eurostat/statistics- added (GVA) at basic prices by NUTS 2 regions - percentage explained/index.php?title=Glossary:City change on previous year, https://appsso.eurostat.ec.eu- Eurostat (2020f): Town or suburb, https://ec.europa.eu/eu- ropa.eu/nui/show.do?dataset=nama_10r_2gvagr&lang=en rostat/statistics-explained/index.php?title=Glos- Eurostat (2020t): GERD by sector of performance and source sary:Town_or_suburb of funds, http://appsso.eurostat.ec.eu- Eurostat (2020g): Urban centre, ropa.eu/nui/show.do?dataset=rd_e_gerdfund&lang https://ec.europa.eu/eurostat/statistics- Eurostat (2020u): Intramural R&D expenditure (GERD) by explained/index.php?title=Glossary:Urban_centre NUTS 2 regions, https://ec.europa.eu/eurostat/data- Eurostat (2020h): Rural area, https://ec.europa.eu/euro- browser/view/tgs00042/default/table?lang=en stat/statistics-explained/index.php/Glossary:Rural_area Federal Employment Agency (Bundesagentur für Arbeit Eurostat (2020i): Rural grid cell, https://ec.europa.eu/euro- Statistik) (2014): Sozialversicherungspflichtig und geringfügig stat/statistics-explained/index.php?title=Glossary:Ru- Beschäftigte nach Wirtschaftszweigen der WZ 2008 und ral_grid_cell ausgewählten Merkmalen. Stichtag 31. Dezember 2013. Bundesagentur für Arbeit Statistik: Nürnberg. Eurostat (2020j): Population by educational attainment level, https://statistik.arbeitsagentur.de/Statistikdaten/Detail/201 sex, age, citizenship and degree of urbanization, 312/iiia6/beschaeftigung-sozbe-wz-heft/wz-heft-d-0- http://appsso.eurostat.ec.europa.eu/nui/show.do?da- 201312 -xls.xls?__blob=publicationFile&v=1 (accessed last, taset=edat_lfs_9916&lang=en 17.05.2021)

Eurostat (2020k): Unemployment rates by sex, age, educa- Federal Employment Agency (Bundesagentur für Arbeit tional attainment level and NUTS 2 regions (%), Statistik (2020a): Sozialversicherungspflichtig und https://appsso.eurostat.ec.europa.eu/nui/show.do?da- geringfügig Beschäftigte nach Wirtschaftszweigen der WZ taset=lfst_r_lfu3rt&lang=en 2008 und ausgewählten Merkmalen. Stichtag 31. Dezember 2019. Bundesagentur für Arbeit Statistik: Nürnberg. Eurostat (2020l): Employment rates by sex, age, educational https://statistik.arbeitsagentur.de/Statistikdaten/Detail/201 attainment level, citizenship and NUTS 2 regions, 912/iiia6/beschaeftigung-sozbe-wz-heft/wz-heft-d-0- https://ec.europa.eu/eurostat/data- 201912-xlsx.xlsx?__blob=publicationFile&v=1 (accessed last, browser/view/tepsr_wc140/default/table?lang=en 17.05.2021) Eurostat (2020m): Early leavers from education and training Federal Employment Agency (Bundesagentur für Arbeit by sex and citizenship, http://appsso.eurostat.ec.eu- Statistik 2020b): Beschäftigte nach Staatsangehörigkeiten ropa.eu/nui/show.do?dataset=edat_lfse_01&lang=en (Quartalszahlen). Deutschland, Länder und die Kreise, 31. Eurostat (2020n): Mean and median income by educational Dezember 2019. Bundesagentur für Arbeit Statistik: attainment level - EU-SILC survey, http://appsso.euro- Nürnberg. https://statistik.arbeitsagentur.de/Statis- stat.ec.europa.eu/nui/show.do?lang=en&dataset=ilc_di08 tikdaten/Detail/201912/iiia6/beschaeftigung-eu-heft-eu-

17

heft/eu-heft-d-0-201912-xlsx.xlsx?__blob=publication- File&v=3 (accessed last, 17.05.2021)

Geis-Thöne, Wido (2019): Sprachkenntnisse entscheidend für die Arbeitsmarktintegration (= IW-Trends 3/2019), Köln. https://www.iwkoeln.de/fileadmin/user_up- load/Studien/IW-Trends/PDF/2019/IW-Trends_2019-03- 05_Sprachkenntnisse_fuer_Arbeitsmarktintegration.pdf (ac- cessed last, 18.03.2021)

Lehmer, Florian; Ludsteck, Johannes (2013): Lohnanpassung von Ausländern am deutschen Arbeitsmarkt: Das Herkunftsland ist von hoher Bedeutung, IAB-Kurzbericht, 01/2013), Nuremberg.

Weidinger, Tobias; Kordel, Stefan (2021): Statistical briefing for Germany on social impact of TCNs in MATILDE regions (= MATILDE Deliverable 3.2).

18

ECONOMIC IMPACT OF MIGRATION

Statistical Briefings Italy

Call: H2020-SC6-MIGRATION-2019

Work Programmes:

H2020-EU.3.6.1.1. The mechanisms to promote smart, sustainable and inclusive growth H2020-EU.3.6.1.2. Trusted organisations, practices, services and policies that are necessary to build resilient, inclusive, participatory, open and creative societies in Europe, in particular taking into account migration, integration and demographic change

Deliverable 4.2 – Statistical briefing for Italy on economic impact of TCNs in MATILDE regions

Authors: Birgit Aigner-Walder, Albert Luger, Rahel M. Schomaker with contributions from Andrea Membretti

Approved by Work Package Manager of WP4: Simone Baglioni, UNIPR Approved by Scientific Head: Andrea Membretti, UEF Approved by Project Coordinator: Jussi Laine, UEF

Version: 28.05.2021

DOI: 10.5281/zenodo.4817376

This document was produced under the terms and conditions of Grant Agreement No. 870831 for the European Commission. It does not necessary reflect the view of the European Union and in no way anticipates the Commission’s future policy in this area.

2

Introduction

This document presents the results of an assessment of the impact of migration and a measure- ment of the contribution provided by ‘third country nationals’ (TCNs) to the economic systems in the receiving contexts. In total, 10 statistical briefings for the MATILDE regions have been compiled, including the following countries: Austria, Bulgaria, Finland, Germany, Italy, Norway, Spain, Sweden, Turkey and the United Kingdom. Each document focuses on the following dimen- sions of economic development: economic growth, labour markets, innovation, and entrepre- neurship. Moreover, a short introduction on the relevance of TCNs in the respective country as well as the population development and structure are considered. The results of a literature review and a comparative analysis of the 10 statistical briefings are published separately.

Within the Matilde project there is a special focus on so-called ‘third country nationals’ (TCNs). MATILDE considers TCNs as Non EU citizens, who reside legally in the European Union (EU) and who are the target of EU integration policies. A TCN is “any person who is not a citizen of the European Union within the meaning of Art. 20(1) of TFEU and who is not a person enjoying the European Union right to free movement, as defined in Art. 2(5) of the Regulation (EU) 2016/399 (Schengen Borders Code).“ (European Commission 2020a). According to this definition, nationals of European Free Trade Association (EFTA) member countries Norway, Island, Liechtenstein and Switzerland are not considered to be TCNs. TCNs cover economic migrants, family migrants, stu- dents and researchers, highly skilled migrants, and forced migrants.

3

ITALY Figure 1: Origin and share of total population of different nationalities in Italy

Source: Eurostat (2020a), own illustration Third Country Nationals (TCN) in Italy Considering the nationalities of TCN, Albanians (0.71 %) are above Moroccans (0.69 %), Chinese On January 1, 2020, a total of 5,039,218 persons (incl. Hong Kong; 0.48 %), Ukrainians (0.38 %) and with non-Italian citizenship were living in Italy.1 Philippines (0.26 %). In total, almost one half of This corresponds to a share of 8.4 % of Italy's total TCN (42.9 %) living in Italy belong to these nation- population. Among non-Italian nationals, almost alities. Figure 1 gives an overview of the origin and one third (1,504,521 persons; 2.5 % of total popu- the share of the different nationalities in Italy. lation) came from EU-28 (incl. UK) and EFTA coun- tries (8,931 persons; 0.01 %). A total of Population Development & Population Structure 3,525,766 persons (or 5.9 % of total population) While the number of inhabitants in Italy has in- were third country nationals (TCN), being individ- creased in the past, it is projected to decrease in uals who are neither Italians, EU-28 citizens nor the future. As of January 1, 2020, there were EFTA citizens. By comparison, the share of TCN in 59,641,488 persons registered in Italy. This corre- EU-27 (excl. UK) and EU-28 (incl. UK) is 5.7 % and sponds to an increase of 2,717,964 citizens 3.8 %, therefore Italy lying above the average. (+4.8 %) within the last 20 years since 2000 (be- fore eastward enlargement of EU in 2004).

1 Moreover, according to recent estimations, in 2020 approx. another 650.000 immigrants were living in It- aly irregularly (ISPI, 2020).

2

The main cause of population growth is immigra- will shrink in the future especially in terms of tion to Italy (EU-28, EFTA & TCN), as the birth rate working age population. is only slightly higher than the death rate of the Analyzing working age population developments domestic population. up to 2040 solely Bolzano (+2.73 %) will grow in A long-term analysis of population development the future. In all other regions the labor force po- since the turn of the new millennium reveals clear tential will shrink. The highest decline in popula- regional differences. 38.2 % of Italian NUTS 3 re- tion will take place in the southern regions Car- gions have less inhabitants in 2020 than in 2001. bonia-Iglesias (-31.91 %), Medio Campidano (- The southern regions Medio Campidano (-10.5 %), 30.40 %), Oristano (-28.76 %), Nuoro (-27.92 %) Enna (-10.1 %), Vibo Valentia (-10.0 %), Potenza (- and Ogliastra (-25.89 %; see Figure 4). 9.2 %) and Carbonia-Iglesias (-8.9 %) recorded the These past and future developments underline the highest decline. The more prosperous regions in high relevance of migration, especially for rural re- the north seem to be more attractive as the high- gions, concerning the development of the working est population growth rates can be found in Reg- age population. Without immigration, their popu- gio nell'Emilia (+17.8 %), Rimini (+16.9 %), lation between 20 and 64 years would shrink even Lodi (+16,0 %), Parma (+15.9 %) and Bolzano more dramatically, being a challenge from an eco- (+15.5 %; see Figure 2). nomic point of view. While most discussions about changing demo- The total-age dependency ratio is a measure of the graphic structures focus on population ageing and age structure of the population. It relates the num- the sustainability of pension systems, urbanization ber of individuals who are li etc., one relevant aspect is labor supply, i.e. the kely to be “dependent” on the support of others for their daily living the population aged 15 to 64 years or 20 to 64 years in – young (up to 19 years old) and the elderly industrialized economies and nations. From an (65 plus years old) to the number of those indi- economic perspective, labor (done by human be- – viduals who are, being working age from 20 to 64 ings) is an essential element in the production of years old, capable of providing this support. goods and services. The term ‘labor force’ com- prises all of those who work for gain, whether as The total-age dependency ratio in 2020 in Italy is employees, employers, or as self-employed, and it 69,3 %. The dependency rate is usually lower in includes the unemployed who are seeking for metropolitan areas and cities than in rural regions. work. The key factors here are a prosperous economy, i.e. more working people, and low child ratios. As Figure 3 (a)-(d) indicate, the increase in the Most metropolitan regions in Italy are also charac- working-age population within the regions is terized by relatively low rates (e.g., Roma mainly caused by the increase in the non-Italian 65.65 %). High dependency ratios are found in re- population. Without immigration, all Italian gions where the share of children and the elderly NUTS 2 regions would have suffered from a down- (65 years and older) is above average. These re- turn in working-age population, with the excep- gions are mainly found in the North (e.g., Trieste tion of Bolzano (+1.3 % since 2007), one of the Ital- 77.13 %). As a result of ageing societies, the total- ian case study regions of the project MATILDE. age dependency ratios will rise in the future by Both, domestic and foreign population groups, are 20.6 pp to 89.9 % in 2040, resulting in implications (re-)locating to cities and metropolitan regions. for publicly funded social security schemes (e.g., This is at the expense of rural areas which are ex- pension system). periencing an overall decline in population. In par- ticular, this is the case for Northern Italy, which

3

Figure 2: Population development (Δ 2001 to 2020), NUTS 3

Source: Eurostat (2020b), own illustration

4

Figure 3: Working-age population (20 to 64 years; Δ 2007 to 2019) (a) Total -11,800 inh. (c) Non EU-28 +1,081,100 inh.

(b) Italians –1,824,100 inh. (d) EU-28 +731,300 inh.

Source: Eurostat (2020c), own illustration

5

Figure 4: Projections of working-age population (20 to 64 years; Δ 2020 to 2040) NUTS 3

Source: Eurostat (2020d), own illustration

6

Education and (Un)Employment The unemployment rate in Italy is 9.9 %, being above the EU-28 average of 6.2 %. From a regional The tertiary education rate of the population with perspective, Calabria has the highest unemploy- Italian citizenship is 18.2 %, which is only slightly ment rate at 20.9 % (female: 22.6 %; above that of the total population (17.4 %). Com- male: 20.0 %), followed by Sicilia (19.9 %; fe- pared to the EU-28 (29.5 %) Italy has a way below male: 22.5 %; male: 18.3 %) and Campania average tertiary education rate. Italy ranks 27th (19.9 %; female: 22.4 %; male: 18.3 %). Provincia within the EU-28 next to Czech Republic (21.6 %) Autonoma di Bolzano/Bozen (2.8 %; 3.1 % female; and Romania (16.0 %), which ranks last. Luxem- 2.5 % male) and Trento (4.9 %; 6.0 % female; bourg (41.0 %), Ireland (40.7 %) and United King- 4.0 % male) are the provinces with the lowest level dom (40.6 %) rank at the top end. of unemployment (see Figure 5). Across the world, educational attainment is signif- In general, there is a correlation between the level icantly higher in cities2 than in towns and suburbs3, of education and unemployment, according to which in turn is higher than in rural areas4. In rural which the well-educated are significantly less areas in Italy only 12.3 % of individuals have uni- likely to be unemployed than people without a vo- versity degrees, i.e. tertiary education (ISCED2011 cational qualification. This correlation can also be levels 5 to 8), compared to 15.2 % in towns & sub- seen for Italy. Those with primary education are urbs and 24.1 % in cities. Secondary degrees are most affected by unemployment (13.7 %). Con- almost evenly spread over cities (40.6 %), towns versely, people with tertiary education have the and suburbs (43.7 %) as well as rural areas (44.0 lowest unemployment rate (5.9 %), those with %). But the regional differences in primary educa- secondary a value of 9.2 %. The same trend can be tional levels are striking. While less than 35.2 % of observed at the European level (EU-28; see Figure city residents have primary, i.e. less than primary, 6), even though with lower unemployment rates. primary and lower secondary education (ISCED2011 levels 0 to 2) attainment level, 41.1 % Table 1: Working-age Population by citizenship, degree of and 43.8 % of residents in towns & suburbs and urbanization and educational attainment level 2019 rural areas do. Educational Attainment Level Citizenship (ISCED11) isation isation In cities, where most of the non-EU-28 citizens Urban- Primary Secondary Tertiary live, more than 35.2 % have only primary educa- Italian 33.0% 41.5% 25.5% tion. This share is even higher in towns & suburbs EU-28 33.4% 49.6% 17.0%

Cities Cities Non EU-28 57.3% 29.5% 13.2% (41.1 %) and rural areas (43.8 %). On contrary, only Total 35.2% 40.6% 24.1% 13.2 % of non-EU-28 citizens have a degree from Italian 39.3% 44.8% 15.9% university but this value is still much higher than in EU-28 39.6% 51.1% 9.3% suburban (8.5 %) and rural (8.4 %) regions. Differ- Non EU-28 65.6% 25.9% 8.5% suburbs Towns & Towns ences of educational attainment level between Total 41.1% 43.7% 15.2% Italian 42.6% 44.8% 12.6% Italian and EU-28 residents range within a band EU-28 43.1% 48.4% 8.5% from 0.3 pp to 8.5 pp with the highest variations Rural Non EU-28 67.1% 24.5% 8.4% between secondary and tertiary levels of educa- Total 43.8% 44.0% 12.3% tion (see Table 1). In total, EU-28 (11.3 %) and non- Source: Eurostat (2020j), own illustration EU-28 (10.5 %) citizens have a lower tertiary edu- cation rate than the domestic population (18.2 %).

2 A city is a local administrative unit (LAU) where at least 50 density of at least 1 500 inhabitants per km² and collectively % of the population lives in one or more urban centres (Eu- a minimum population of 50 000 inhabitants after gap-filling rostat 2020e). (Eurostat 2020f; Eurostat 2020g). 3 Towns & suburbs are areas where less than 50 % of the 4 A rural area is an area where more than 50 % of its popula- population lives in rural grid cells and less than 50 % of the tion lives in ‘rural grid cells’, i.e. grid cells that are not identi- population lives in urban centres, i.e. a cluster of contiguous fied as urban centres or as urban clusters (Eurostat, 2020h; grid cells of 1 km² (excluding diagonals) with a population Eurostat, 2020i).

7

Figure 5: Unemployment rate 2019 (20 to 64 years) Figure 6: Unemployment rate by educational attainment level 2019 (20 to 64 years; ISCED 2011)

16 13.7 14 12.4 12

10 9.2

8 5.9 6 5.6 4.0 4

2

0 Primary Secondary Tertiary

EU-28 Italy

Source: Eurostat (2020k), own illustration

Figure 7: Employment rate by educational attainment level and citizenship 2019 (20 to 64 years; ISCED 2011) Source: Eurostat (2020k), own illustration 90 In contrast, the higher the educational attainment 79.7 80 the higher the employment rate, i.e. 78.9 % for 72.8 70 67.464.7 66.4 tertiary, 66.3 % for secondary and 52.1 % for pri- 61.762.9 64.7 60 mary level. From an immigrational point of view, 50.1 this observation holds for other citizenships as 50 well, as the risk of unemployment decreases by 40 the level of educational attainment. But Italians 30 have the highest employment rate for tertiary ed- 20 ucational attainment level (79.7 %) while EU- 10 28/non-EU-28 rank first for secondary/primary ed- 0 ucational attainment level (67.4 %/62.9 %). Non- Primary Secondary Tertiary EU-28 citizens have the lowest employment rate on secondary and tertiary level (both 64.7 %; see EU-28 Non EU-28 Italy Figure 7). Source: Eurostat (2020l), own illustration The abovementioned findings foster the theoreti- cal conclusions by Becker (1964) that the educa- education or training. From a fiscal point of view, tional level has a significant influence on the pro- this specific group is of particular relevance as they fessional career and the risk of unemployment of are more likely to be (long-term) unemployed. In human beings. Considering migrants the recogni- 2019, 10.3 % of the 18-24 year olds in the EU-28 tion of educational degrees might be an additional were part of this group (male: 11.9 %; female: problem. 8.6 %). Italy ranks with a value of 13.5 % above the European average. The share of EU-28 citizens re- This theory and the empirical evidence can also be siding in Italy is almost three times as high (30.8 %) applied to the “Early Leavers from Education and and that of non-EU-28 is more than three times as Training”. This population group refers to persons high (38.5 %) than that of the domestic population aged 18 to 24 who have completed a lower sec- (11.3 %). In total, almost two thirds (64.4 %) of all ondary education and are not involved in further early leavers are jobless. An analysis by citizenship

8

shows low variation among foreigners (EU-28: As non-EU 28 citizens have a higher risk of unem- 56.8 %; Non-EU-28: 55.6 %) but Italians record the ployment and they earn less than domestic resi- highest value (67.3 %, Eurostat 2020m). dents and EU-28, more than 2 out of 5 persons (42.6 %) belong to this group, though the value of Income and Gross Domestic Product EU-28 is quite high as well (29.9 %). Italians face As for the employment rate, the educational at- the lowest risk of social exclusion and poverty tainment level has a positive impact on mean and (24.1 %; see Table 2). 5 median equivalized net income, i.e. the higher Figure 9: Mean and median equivalized net income by the educational attainment level the higher the in- broad group of citizenship 2018 (18 to 64 years) come. Working persons in Italy at primary level earn less (mean: € 15,682; median: € 14,377) than 25,000 secondary (€ 20,013 or € 18,227) and tertiary lev- 20,472 els (€ 26,854 or € 23,491; see Figure 8). As working 20,000 18,423 Italians have a higher educational attainment level 16,232 than EU-28 and non-EU-28 (see Figure 7) this could 15,000 13,398 13,414 12,387 be one reason (amongst others) that their mean (€ 20,472 vs. € 16,232 and € 13,414) and median (€ 10,000 18,423 vs. € 13,398 and € 12,387) equivalized net income is higher. 5,000

Figure 8: Mean and median equivalized net income by edu- cational attainment level 2018 (18 to 64 years; ISCED 2011) 0 EU-28 Non EU-28 Italy 30,000 26,854 Mean equivalised net income Median equivalised net income 25,000 23,491 Source: Eurostat (2020o), own illustration 20,013 20,000 18,227 15,682 In 2019, compensation of employees (wages and 15,000 14,377 salaries plus employers’ social contributions) was the largest income component of GDP in EU-28 10,000 and the Euro area, accounting for 47.8 % and 48.0 %. Italy (40.2 %) lies below European average 5,000 (both EU-28 and Euro area). Taxes on production 0 and imports (less subsidies) accounted for 12.7 % Primary Secondary Tertiary (EU-28: 11.9 %; Euro area: 11.5 %). Gross operat- ing surplus and mixed income accounted for Mean equivalised net income 47.1 % of GDP in Italy, 40.3 % of GDP for the EU-28 Median equivalised net income and 40.5 % in the Euro area. Italian GDP at current Source: Eurostat (2020n), own illustration prices amounted to approx. € 1,790.9 billion in The indicator ‘people at risk of poverty or social 2019 and GDP per inhabitant equaled € 29,680 exclusion’ corresponds to the sum of persons who (Eurostat 2020q). are: at risk of poverty after social transfers, se- verely materially deprived or living in households with very low work intensity.

5 The median is more robust, i.e. less sensitive to outliers than the mean. This explains why the mean equivalized income is always slightly higher for each educational level.

9

Table 2: Population aged 18 and over by risk of poverty, material deprivation, work intensity 2019 Risk of poverty At risk Not at risk At risk Not at risk At risk Not at risk Work Intensity Material Deprivation Italians EU-28 Non EU-28 Very low Severe 1.8 0.3 2.1 0.1 1.6 0.6 Not very low Severe 2.0 3.0 4.5 3.5 5.9 6.8 Very low Non severe 2.7 2. 6 2.9 1.4 2.4 0.5 Not very low Non severe 11.7 15.4 24.8 People at Risk of Poverty or Social Exclusion 24.1 29.9 42.6 Not at risk Not at risk Not at risk Not very low Non severe 76.0 70.1 57.3 People at no Risk of Poverty or Social Exclusion 76.0 70.1 5 7.3 Total 100.0 100.0 100.0 Source: Eurostat (2020p), own illustration

Within 20 years Italy’s gross value added (GVA) in- Table 3: Nominal regional GDP (2019) and real growth rate creased by only +5.2 % (real growth rate). In terms (Δ 2002 to 2019) of regional GVA the majority of NUTS 2 regions Total Per Capita

recorded positive real growth rates, ranging from +1.0 % in Friuli-Venezia Giulia to +24.1 % Bol-

zano/Bozen since 2000; but with exception of the 2002 to 2019)

Northeast there are regions all over Italy facing a Region

decline in economic growth. Especially the South €) GDP (in (-5.6 %) faced severe difficulties as Basilicata current marketcurrent prices the EU27 (from 2020) average (fromEU27 2020) the GDP (in €) €) per capita GDP (in at current marketat current prices at Purchasing power standard (PPS, EU27 standardpower Purchasing(PPS,EU27 value added(GVA) at basic prices Real rate growth of regional gross

(+2.2 %) was the only among 5 regions growing (Index, 2015=100; Δ from 2020), per inhabitant in percentage ofpercentage in inhabitant per from2020), within the past 20 years with Molise (-15.7 %) and Nord-Ovest 591,170 9.1 36,700 118 Calabria (-10.6 %) having the greatest decline (see Piemonte 137,782 2.3 31,700 102 Table 3). Valle d'Aosta 4,868 -3.1 38,700 125 Liguria 49,741 -6.8 32,200 104 The highest regional GDP per capita at current Lombardia 398,779 13.6 39,500 127 prices occurred also in Bolzano/Bozen (approx. Nord-Est 413,866 10.1 35,500 114 Bolzano 25,516 24.1 48,000 155 € 48,000) succeeded by Lombardia (approx. Trento 20,967 10.8 38,700 125 € 39.500) and another northeastern region, Veneto 164,860 8.0 33,600 108 Friuli-Venezia Trento (approx. € 38,700). As in previous years, 38,772 1.0 32,000 103 the southern provinces were far below the na- Giulia Emilia- 163,751 12.1 36,700 118 tional level of € 29,700 (e.g. Calabria: € 17,400; Si- Romagna cilia: € 17,900; Campania: € 18,900) but also re- Centro 385,226 5.8 32,100 103 gions in the center (Umbria: € 26.400; Marche: Toscana 118,727 5.9 31,900 103 € 27,800) performed below average. Except for Umbria 23,267 -6.4 26,400 85 Marche 42,392 2.5 27,800 90 Umbria and Marche all central, northwestern and Lazio 200,840 7.7 34,200 110 northeastern regions rank above average (see Ta- Sud 273,436 -5.6 19,600 63 ble 3). Abruzzo 33,131 -1.6 25,300 82 Molise 6,490 -15.7 21,300 69 Provinces with a high share of TCN have a high Campania 109,631 -6.5 18,900 61 GDP per capita and vice versa. On the one hand, Puglia 77,475 -4.3 19,300 62 prosperous regions are more attractive to mi- Basilicata 13,090 2.2 23,400 75 grants; on the other hand, these regions benefit Calabria 33,619 -10.6 17,400 56 Isole 124,621 -6.7 18,800 61 from a better availability of labor supply, i.e. Sicilia 89,365 -8.8 17,900 58 greater number of working age people. In this con- Sardegna 35,256 -1.2 21,600 69 text, one speaks of a so-called cross-fertilization, Italy 1,789,747 5.2 29,700 96 which results from the interaction of these two Source: Eurostat (2020r), own illustration; Eurostat (2020s), own illustration factors.

10

Expressing GDP in PPS (purchasing power stand- tor in Mezzogiorno (5.7 %) the dispersion across It- ards) eliminates differences in price levels be- aly is at a very low level (0.8 pp; see Figure 10). tween countries or regions. There are significant Due to a lack of data no reliable statements can be differences in levels of prosperity among Italian re- made in regard to TCNs. gions. The most prosperous region in Italy is also Figure 10: Employment by economic sector and citizenship Bolzano/Bozen (155 %); it is over 99/97/94 pp 2019 richer than Calabria/Sicilia/Campania which are (a) Primary Sector: 908,779 inh.; 3.9 % Employment the poorest regions from an economic point of 0% 20% 40% 60% 80% 100% view. Lombardia ranks second (127 %) and Trento Nord 83.4% 16.6% third (125 %; see Table 3).

Economic structure and entrepreneurship Nord-ovest 82.7% 17.3%

In total, approx. 23.46 million people were em- Nord-est 83.9% 16.1% ployed in . This corresponds to an in- crease of +1.2 % compared to 2008. Centro 70.8% 29.2%

In general, a distinction is made between three dif- Mezzogiorno 83.9% 16.1% ferent sectors of the economy, the primary (agri- culture and forestry), secondary (manufacturing) Italy 81.7% 18.3% and tertiary (services) sectors. Almost three quar- Italian Foreign ters (70.2 %) are employed in the service sector, (b) Secondary Sector: 6,042,459 inh.; 25.9 % Employment where employment has increased by +3.7 % since 0% 20% 40% 60% 80% 100% 2008. Employment in manufacturing shrunk by - 3.9 % and has a share of 25.9 %. Though primary Nord 87.1% 12.9% sector increased by +0.2 % it is of low importance, i.e. only 3.9 % of total employment (see Figure 11 Nord-ovest 86.9% 13.1% (a)-(c) and Figure 11). Nord-est 87.3% 12.7% Further analysis by territories and citizenship shows that more than 1 of 10 workers (10.7 %) are Centro 85.0% 15.0% foreigners. The absolute number is the highest in the northern part (1,465,667 inh.; 12.0 %.) Mezzogiorno 95.4% 4.6% whereas the share is highest in the center Italy 88.4% 11.6% (656,322 inh.; 13.2 %). Except for the northeast- ern part where the number of Italians almost stag- Italian Foreign (c) Tertiary Sector: 16,408,628 inh.; 70.2 % Employment nates (0.1 %) all remaining territories show a neg- 0% 20% 40% 60% 80% 100% ative tendency especially in the southern part (Mezzogiorno: -7.0 %). In contrast, the number of Nord 88.6% 11.4% foreigners evolves at a very high level ranging from 28.9 % (Nord-est) to 98.5 % (Mezzogiorno) within Nord-ovest 88.4% 11.6% more than 10 years since 2008 (see Figure 11). From a sectoral point of view foreigners preferably Nord-est 88.8% 11.2% work in the primary sector (18.3 %). Only 11.6 % work in the secondary sector and 10.0 % in the ter- Centro 87.9% 12.1% tiary sector where their absolute number is high- Mezzogiorno 94.3% 5.7% est (1,637,551 inh.). Especially in the central part of Italy almost 3 out of 10 workers (29.2 %) in the Italy 90.0% 10.0% primary sector are foreigners. Except for the com- Italian Foreign parably low share of foreigners in the tertiary sec- Source: Istat (2021), own calculations and illustration

11

Figure 11: Employment by nationalities and territories, 2008 and 2019 12,000,000 100.0% 98.5%

10,724,347

10,000,000 80.0%

61.6% 8,000,000 60.0%

6,132,983 38.7% 6,000,000 5,799,409 40.0% 34.3%

4,591,363 28.9% 4,330,925 4,000,000 20.0%

-0.7% 0.1% -0.6% 2,000,000 -1.4% 0.0% -7.0%

1,465,667 846,716 618,951 656,322 383,197 0 -20.0% Nord Nord-ovest Nord-est Centro Mezzogiorno

Italian Foreign ∆ 08-19 ∆ 08-19 Italian Foreign

Source: Istat (2021), own calculations and illustration

Research and Innovation The total amount of research expenditures was € 25.23 billion, the largest share (54.9 %) being fi- Following the neoclassical and endogenous nanced by the business enterprise sector. 33.1 % growth theories, technological advance is believed were financed by the government sector. 10.6 % to be one of the major drivers of economic by foreign countries (rest of the world), 1.4 % and growth. From this perspective, there is a growing 0.7 % by private non-profit and higher education interest to investigate the link between research & sector (Eurostat 2020t). development (R&D), innovation, entrepreneur- ship and economic growth achieved by human Regional data is published on NUTS 1 and NUTS 2 capital from abroad (EU-28 and non-EU-28). level. Piemonte has the highest ratio in 2018 (2.18 %). Emilia-Romagna (2.03 %) and Lazio A well-known indicator provided to measure (1.75 %) take second and third place, ahead of Fri- achievements of countries or regions in R&D is uli-Venezia Giulia (1.67 %), Trento (1.57 %) and GERD, i.e. regional/national gross domestic ex- Toscana (1.55 %). All other provinces are well be- penditure on R&D as a percentage of GDP. GERD hind and below the all-Italian ratio of 1.42 %. Sar- is estimated at 1.42 % for Italy (2018) which is a degna ranks last (0.80 %) with Sicilia (0.82 %), Bol- slight increase of 0.05 percentage points to 2017.

12

zano/Bozen (0.84 %), Abruzzo (0.91 %) and Um- Table 4: Intramural R&D expenditure as percentage of bria (1.03 %) only slightly better (see Table 4). For gross domestic product (GDP) 2017 Expenditures % of the MATILDE case study region Bolzano/Bozen the Region (€ mio.) GDP low level in R&D expenditures is particularly sur- Nord-Ovest 8,892.1 1.53% prising, as Bolzano/Bozen hast the highest GDP Piemonte 2,987.5 2.18% per capita in Italy. Valle d'Aosta 23.7 0.48% Liguria 672.7 1.35% In accordance to the research & development Lombardia 5,208.3 1.34% (R&D) funding structure almost one half (44.5 % or Nord-Est 6,706.4 1.64% 62,478 full-time equivalent (FTE)) of the research- Bolzano/Bozen 207.8 0.84% Trento 321.5 1.57% ers work in the business enterprise sector. 15.8 % Veneto 2,263.4 1.39% (or 22,118 FTE) work in the government sector Friuli-Venezia Giulia 634.7 1.67% and 3.1 % (or 4,421 FTE) in the private non-profit Emilia-Romagna 3,279.0 2.03% sector. The higher education sector (46.9 % or Centro 5,971.9 1.57% 62,542 FTE) is by far the greatest employer of re- Toscana 1,828.0 1.55% Umbria 231.1 1.03% searchers. This makes a total of 133,213 FTE. Fur- Marche 458.0 1.06% ther analysis on educational attainment level re- Lazio 3,454.7 1.75% veals that most researchers (67.0 %) have tertiary Sud 2,646.3 0.98% education but not doctoral or equivalent level; Abruzzo 307.1 0.91% Molise 81.6 1.26% 8.8 % belong to the latter whereas 24.3 % have Campania 1,404.0 1.30% less than primary, primary, secondary and post- Puglia 594.6 0.78% secondary non-tertiary education (see Figure 12). Basilicata 79.2 0.63% Calabria 179.8 0.54% As the group of migrants (EU-28 or non-EU-28) liv- Isole 1,015.5 0.82% ing in Italy has a lower share of tertiary education Sicilia 735.2 0.82% it is therefore likely that they are underrepre- Sardegna 280.4 0.80% sented in Italian R&D personnel. Italy 25,232.2 1.42% Source: Eurostat (2020u), own illustration

Figure 12: R&D researchers (full-time equivalent (FTE)) by sectors of performance and educational attainment level 2017 (ISCED2011)

45,000 41,853 40,000

35,000

30,000

25,000

20,000 15,154 14,530 15,000

10,000 7,213 5,471 5,000 2,567 1,794 375 60 0 Business enterprise sector Government sector Private non-profit sector

Less than primary, primary, secondary and post-secondary non-tertiary education (levels 0-4) Tertiary education excluding doctoral or equivalent level (levels 5-7) Doctoral or equivalent level

Source: Eurostat (2020v), own illustration

13

Summary Italy: the foreign population of third countries is disproportionately represented in primary educa- This Statistical Briefings examines the economic tion, whereas they are underrepresented in sec- and spatial impact of immigration with special re- ondary and tertiary education levels. gard to third-country nationals (TCN) in Italy. From a spatial point of view, the concentration of On January 1, 2020, a total of 5,039,218 persons better-educated migrants is highest in Italian cit- with non-Italian citizenship were officially living in ies. Provinces with a higher ratio of TCN are also Italy. This corresponds to a share of 8.4 % of Italy's characterized by higher economic growth, being total population. Moreover, estimations show more attractive for TNC on the one hand and of- 650.000 irregular migrants. Among non-Italian na- fering more labor supply on the other hand. tionals, almost one third (2.5 % of total popula- tion) came from EU-28 (incl. UK) and EFTA coun- The below average regional GVA growth rate, es- tries. A total of 3,525,766 persons (or 5.9 % of to- pecially in rural regions is indirectly linked to im- tal population) were third country nationals (TCN), migration from abroad but preferably to internal being individuals who are neither Italians, EU-28 migration from rural to cities and metropolitan ar- citizens nor EFTA citizens. By comparison, the eas. These emigration flows tend to reinforce the share of TCN in EU-27 (excl. UK) and EU- erosion of the economic basis, because they have 28 (incl. UK) is 5.7 % and 3.8 %, therefore Italy ly- a negative impact on the potential workforce and ing above the average. the economic attractiveness of the region. Eco- nomic performance is stagnating, accompanied by Within the last 20 years the population increase of declining population figures. Italy has been dominated by immigration. From 2006 to 2019 all Italian regions profited heavily Governments need to invest in education on the with regard to its working-age population. With- one hand and in research and development (R&D) out immigration, all Italian regions would have suf- on the other, to enable long-term prosperity and fered from a decrease in its population aged be- growth. Currently the national gross domestic ex- tween 20 and 65 years. penditures on R&D (as a percentage of GDP) are estimated at 1.42 % (2018) which is a slight in- Despite the relevance of migration for labor sup- crease of 0.05 pp to 2017. But there is again a ply, the employment rate of TCN is much lower striking regional variation in R&D expenditures. compared to Italians and EU-28 citizens. One rele- vant factor to be considered here is the lower ed- Summing up, immigrants in Italy play a pivotal role ucational attainment level of migrants, with the in reviving rural areas, and with the appropriate consecutive effects of lower income levels and a policies, they could play an even more significant higher risk of poverty. role in sustaining them.

The pattern of education of the EU-28 and non- EU-28 countries differs substantially from that of

14

Bibliography and source references

Becker, Gary Stanley (1964): Human Capital: A theoretical Eurostat (2020o): Mean and median income by broad group and empirical analysis, with special reference to education. of citizenship (population aged 18 and over), 3rd edition. New York. https://appsso.eurostat.ec.europa.eu/nui/show.do?da- taset=ilc_di15&lang=en Eurostat (2020a): Population on 1 January by age group, sex and citizenship, https://appsso.eurostat.ec.eu- Eurostat (2020p): Persons aged 18 and over by risk of pov- ropa.eu/nui/show.do?dataset=migr_pop1ctz&lang=en erty, material deprivation, work intensity of the household and most frequent activity status of the person - intersec- Eurostat (2020b): Population change - Demographic balance tions of Europe 2020 poverty target indicators, and crude rates at regional level (NUTS 3), https://appsso.eurostat.ec.europa.eu/nui/show.do?da- https://appsso.eurostat.ec.europa.eu/nui/show.do?da- taset=ilc_pees02&lang=en taset=demo_r_gind3&lang=en Eurostat (2020q): GDP and main components (output, ex- Eurostat (2020c): Population by sex, age, citizenship, labour penditure and income), https://appsso.eurostat.ec.eu- status and NUTS 2 regions, https://appsso.eurostat.ec.eu- ropa.eu/nui/show.do?dataset=nama_10_gdp&lang=en ropa.eu/nui/show.do?dataset=lfst_r_lfsd2pwn&lang=en Eurostat (2020r): Gross domestic product (GDP) at current Eurostat (2020d): Population on 1st January by age, sex, type market prices by NUTS 2 regions, https://appsso.euro- of projection and NUTS 3, https://appsso.eurostat.ec.eu- stat.ec.europa.eu/nui/show.do?da- ropa.eu/nui/show.do?dataset=proj_19rp3 taset=nama_10r_2gdp&lang=en Eurostat (2020e): City, Eurostat (2020s), Real growth rate of regional gross value https://ec.europa.eu/eurostat/statistics- added (GVA) at basic prices by NUTS 2 regions - percentage explained/index.php?title=Glossary:City change on previous year, https://appsso.eurostat.ec.eu- Eurostat (2020f): Town or suburb, https://ec.europa.eu/eu- ropa.eu/nui/show.do?dataset=nama_10r_2gvagr&lang=en rostat/statistics-explained/index.php?title=Glos- Eurostat (2020t): GERD by sector of performance and source sary:Town_or_suburb of funds, http://appsso.eurostat.ec.eu- Eurostat (2020g): Urban centre, ropa.eu/nui/show.do?dataset=rd_e_gerdfund&lang https://ec.europa.eu/eurostat/statistics- Eurostat (2020u): Intramural R&D expenditure (GERD) by explained/index.php?title=Glossary:Urban_centre NUTS 2 regions, https://ec.europa.eu/eurostat/data- Eurostat (2020h): Rural area, https://ec.europa.eu/euro- browser/view/tgs00042/default/table?lang=en stat/statistics-explained/index.php/Glossary:Rural_area Eurostat (2020v): Total R&D personnel and researchers by Eurostat (2020i): Rural grid cell, https://ec.europa.eu/euro- sectors of performance, educational attainment level stat/statistics-explained/index.php?title=Glossary:Ru- (ISCED2011) and sex, https://ec.europa.eu/eurostat/data- ral_grid_cell browser/view/rd_p_persqual11/default/table?lang=en

Eurostat (2020j): Population by educational attainment level, ISPI Istituto per gli Studi di Politica Internazionale (2020): Mi- sex, age, citizenship and degree of urbanization, grazioni in Italia: tutti i numeri, https://www.ispi- http://appsso.eurostat.ec.europa.eu/nui/show.do?da- online.it/it/pubblicazione/migrazioni-italia-tutti-i-numeri- taset=edat_lfs_9916&lang=en 24893 [09.05.2021]

Eurostat (2020k): Unemployment rates by sex, age, educa- Istat (2021): Employment (thousands): Nace 2007 - profes- tional attainment level and NUTS 2 regions (%), sional status, citizenship, http://dati.istat.it/Index.aspx?Que- https://appsso.eurostat.ec.europa.eu/nui/show.do?da- ryId=26887&lang=en# taset=lfst_r_lfu3rt&lang=en

Eurostat (2020l): Employment rates by sex, age, educational attainment level, citizenship and NUTS 2 regions, https://ec.europa.eu/eurostat/data- browser/view/tepsr_wc140/default/table?lang=en

Eurostat (2020m): Early leavers from education and training by sex and citizenship, http://appsso.eurostat.ec.eu- ropa.eu/nui/show.do?dataset=edat_lfse_01&lang=en

Eurostat (2020n): Mean and median income by educational attainment level - EU-SILC survey, http://appsso.euro- stat.ec.europa.eu/nui/show.do?lang=en&dataset=ilc_di08

15

ECONOMIC IMPACT OF MIGRATION

Statistical Briefings Norway

Call: H2020-SC6-MIGRATION-2019

Work Programmes:

H2020-EU.3.6.1.1. The mechanisms to promote smart, sustainable and inclusive growth H2020-EU.3.6.1.2. Trusted organisations, practices, services and policies that are necessary to build resilient, inclusive, participatory, open and creative societies in Europe, in particular taking into account migration, integration and demographic change

Deliverable 4.2 – Statistical briefing for Norway on economic impact of TCNs in MATILDE re- gions

Authors: Birgit Aigner-Walder, Albert Luger, Rahel M. Schomaker with contributions from Per Olav Lund and Veronica Isabel Blumenthal

Approved by Work Package Manager of WP4: Simone Baglioni, UNIPR Approved by Scientific Head: Andrea Membretti, UEF Approved by Project Coordinator: Jussi Laine, UEF

Version: 28.05.2021

DOI: 10.5281/zenodo.4817376

This document was produced under the terms and conditions of Grant Agreement No. 870831 for the European Commission. It does not necessary reflect the view of the European Union and in no way anticipates the Commission’s future policy in this area.

2

Introduction

This document presents the results of an assessment of the impact of migration and a measure- ment of the contribution provided by ‘third country nationals’ (TCNs) to the economic systems in the receiving contexts. In total, 10 statistical briefings for the MATILDE regions have been compiled, including the following countries: Austria, Bulgaria, Finland, Germany, Italy, Norway, Spain, Sweden, Turkey and the United Kingdom. Each document focuses on the following dimen- sions of economic development: economic growth, labour markets, innovation, and entrepre- neurship. Moreover, a short introduction on the relevance of TCNs in the respective country as well as the population development and structure are considered. The results of a literature review and a comparative analysis of the 10 statistical briefings are published separately.

Within the Matilde project there is a special focus on so-called ‘third country nationals’ (TCNs). MATILDE considers TCNs as Non EU citizens, who reside legally in the European Union (EU) and who are the target of EU integration policies. A TCN is “any person who is not a citizen of the European Union within the meaning of Art. 20(1) of TFEU and who is not a person enjoying the European Union right to free movement, as defined in Art. 2(5) of the Regulation (EU) 2016/399 (Schengen Borders Code).“ (European Commission 2020a). According to this definition, nationals of European Free Trade Association (EFTA) member countries Norway, Island, Liechtenstein and Switzerland are not considered to be TCNs. TCNs cover economic migrants, family migrants, stu- dents and researchers, highly skilled migrants, and forced migrants.

3

NORWAY Figure 1: Origin and share of total population of different nationalities in Norway

Source: Eurostat (2020a), own illustration Third Country Nationals (TCNs) in Norway total, more than 40 % (41.9 %) of TCN living in Nor- way belong to these nationalities. Figure 1 gives an As of January 1, 2020, a total of 940,580 persons overview of the origin and the share of the differ- with non-Norwegian citizenship were living in Nor- ent nationalities in Norway. way. This corresponds to a share of 11.3 % of Nor- way's total population. Among non-Norwegian na- Population Development & Population Structure tionals, 61.7 % (580,042 persons; 6.9 % of the to- The number of inhabitants in Norway increased in tal population) came from EU-28 (incl. UK) and the past and this development is projected to con- EFTA countries (excl. Norway; 15,615 persons; tinue according to current population forecasts. 0.2 %). A total of 426,535 persons (or 4.1 % of total As of January 1, 2020, there were 5,367,580 per- population) were third country nationals (TCN), sons registered in Norway. This corresponds to an being individuals who are neither Norwegian, EU- increase of 889,083 citizens (+ 19.9 %) within the 28 citizens nor EFTA citizens. By comparison, the last 20 years since 2000 (before eastward enlarge- share of TCN in EU-27 (excl. UK) and EU- ment of EU in 2004). 28 (incl. UK) is 5.7 % and 3.8 %, therefore the EFTA member state Norway lies below the EU average. A long-term analysis of population development since 2015 reveals clear regional differences Considering the nationalities of TCN, Syrians though all regions recorded a growth. Vestland (0.60 %) are above Eritreans (0.35 %), Philippians (+2.4 %), Nordland (+2.4 %) and Troms og Finn- (0.24 %), Somalians (0.22 %) and Thai (0.22 %). In mark (+3.3 %) grew at a comparably low level.

2

Compared to the under-average population labor supply, i.e. the population aged 15 to 64 growth recorded in the north, the situation is dif- years or 20 to 64 years in industrialized economies ferent in the south: Oslo (+ 30.9 %), Viken and nations. From an economic perspective, labor (+ 28.1 %) as well as Rogaland (+ 22.1 %) have very (done by human beings) is an essential element in high growth rates (see Figure 2). the production of goods and services. The term ‘la- bor force’ comprises all of those who work for While most discussions about changing demo- gain, whether as employees, employers, or as self- graphic structures focus on population ageing and employed, and it includes the unemployed who the sustainability of pension systems, urbanization are seeking for work. etc., another economically relevant aspect is the

Figure 2: Population development (Δ 2005 to 2019), NUTS 3

Source: Eurostat (2020b), own illustration

3

As Figure 3 (a)-(d) indicate, the increase in the facing a shrinking domestic working-age popula- working-age population (+397,800 inh.) is domi- tion in the past. From a regional perspective, Oslo nated by immigration (EU-28: +203,800 inh.; non- and Viken (+25.5 %), Agder and Rogaland EU-28: +85,800 inh.). Domestic population in- (+16.7 %), Trøndelag (+15.0 %) as well as Vestland creased (+106,100 inh.) as well but on a smaller (12.6 %) are fast-growing areas, with lower growth scale than immigrants. This makes Norway special, rates in the remaining regions. as most industrialized countries have already been

Figure 3: Working-age population (20 to 64 years; Δ 2006 to 2019) (a) (c)

(b) (d)

Source: Eurostat (2020c), own illustration

4

According to current forecasts, the increase of la- which are projected to shrink at a comparably low bor force potential does not stop in the future, rate in the future, i.e. Troms og Finnmark (- though the growth rates are lower compared to 3.69 %), Innlandet (-2.69 %), Vestfold og Telemark the past. Working-age population will increase by (-1.78 %), Trøndelag (-1.63 %) and Vestland (- +8.4 % until 2040. There are only five regions 1.08 %; see Figure 4).

Figure 4: Projections of Working-age Population (20 to 64 years; Δ 2020 to 2040) NUTS 3

Source: Eurostat (2020d), own illustration

The total-age dependency ratio is a measure of the (65 years old and upwards) – to the number of age structure of the population. It relates the num- those individuals who are, being working age from ber of individuals who are likely to be “dependent” 20 to 64 years old, capable of providing this sup- on the support of others for their daily living – the port. young (up to 19 years old) and the elderly

5

The total-age dependency ratio in 2020 in Nor- rural areas do. Primary educational attainment way is 69.1 %. The dependency rate is usually level has low relevance as approx. only 1 out of 5 lower in metropolitan areas and cities than in ru- residents in Norway belongs to this group. ral regions. The key factors here are a more pros- Concerning the educational attainment level of perous economy, i.e. more people in working- TCN, large differences can be recorded. In cities, age, and low child ratios. Most metropolitan re- where most of the Non-EU-28 citizens live, approx. gions in Norway (e.g., Oslo) are also characterized 23.7 % have only primary education. This share is by relatively low rates. High dependency ratios even higher in towns & suburbs (34.5 %) and rural are found in Vestland (80.08 %), Møre og Roms- areas (31.4 %). The shares of non-EU-28 citizens dal (76.50 %) and Innlandet (76.68 %). As a result (40.3 %), EU-28 citizens (45.0 %) and Norwegians of ageing societies, the total-age dependency ra- (49.6 %) with tertiary education living in cities are tios will rise in the future by 8.1 pp to 77.2 % in all on a very high level. On the contrary, only 2040 resulting in implications/challenges for pub- 25.7 % or 22.2 % of non-EU-28 citizens living in licly funded social security schemes (e.g., expend- towns and suburbs or rural areas have a university itures for pensions.) degree. Differences of educational attainment Education and (Un)Employment level between Norwegians and EU-28 residents range within a broad band from 2.1 pp to 12.7 pp. The tertiary education rate of Norway's domestic (see Table 1). In total, EU-28 (37.6 %) and non-EU- population is 38.2 %, which is slightly above that 28 (29.3 %) citizens have a lower tertiary educa- of the total population (37.7 %). Compared to the tion rate than domestic population (38.2 %). EU-28 (29.5 %) Norway has a way above-average tertiary education rate. If Norway was EU-28 Table 1: Working-age population by citizenship, degree of member it would approx. rank 6th ex aequo with urbanization and educational attainment level 2019 Lithuania (37.9 %), ahead of Estonia (36.5 %), Bel- Educational Attainment Level gium (36.0 %) and Spain (35.1 %). Luxembourg Citizenship (ISCED11) (41.0 %), Ireland (40.7 %) and United Kingdom

(40.6 %) rank at the top end, while the Czech Re- Urbanization Primary Secondary Tertiary public (21.6 %), Italy (17.4 %) and Romania Norwegians 19.8% 30.4% 49.8% (16.0 %) rank at the bottom. EU-28 11.9% 43.1% 45.0%

Cities Cities Non EU-28 23.7% 35.9% 40.3% Across the world, educational attainment is signif- Total 19.3% 31.8% 48.9% icantly higher in cities1 than in towns and suburbs2, Norwegians 24.1% 38.0% 37.9% EU-28 17.9% 46.3% 35.8% which in turn is higher than in rural areas. In rural Non EU-28 34.5% 39.9% 25.7% suburbs

3 & Towns areas in Norway, only 30.0 % of individuals have Total 24.2% 38.8% 37.0% university degrees, i.e. tertiary education Norwegians 26.9% 42.9% 30.2% (ISCED2011 levels 5 to 8), compared to 37.0 % in EU-28 18.7% 48.8% 32.5%

Rural Non EU-28 31.4% 46.4% 22.2% towns & suburbs and 48.9 % in cities. In contrast, Total 26.5% 43.5% 30.0% secondary degrees are more common in less Source: Eurostat (2020j), own illustration densely populated areas. While less than 31.8 % of city residents have secondary, i.e. upper second- ary and post-secondary non-tertiary education (ISCED2011 levels 3 to 4) attainment level, 38.8 % and 43.5 % of residents in towns & suburbs and

1 A city is a local administrative unit (LAU) where at least density of at least 1 500 inhabitants per km² and collectively 50 % of the population lives in one or more urban centres a minimum population of 50 000 inhabitants after gap-filling (Eurostat 2020e). (Eurostat 2020f; Eurostat 2020g). 2 Towns & suburbs are areas where less than 50 % of the 3 A rural area is an area where more than 50 % of its popula- population lives in rural grid cells and less than 50 % of the tion lives in ‘rural grid cells’, i.e. grid cells that are not identi- population lives in urban centres, i.e. a cluster of contiguous fied as urban centres or as urban clusters (Eurostat 2020h; grid cells of 1 km² (excluding diagonals) with a population Eurostat 2020i).

6

The unemployment rate in Norway is 3.3 %, being second on each level (89,1 %; 78.8 %; 61.1 %) and far below EU-28 average of 6.2 %. From a regional non-EU-28 citizens have the lowest employment perspective, Oslo and Viken has the highest unem- rates on each level (73.5 %; 54.0 %; 48.6 %; see ployment rate at 3.6 % (female: 3.4 %; Figure 7). male: 3.8 %), followed by Agder and Rogaland Figure 6: Unemployment rate by educational attainment (3.5 %; female: 3.4 %; male: 3.6 %). Troms of Finn- level 2019 (20 to 64 years; ISCED 2011) mark and Nordland (2.6 %) has the lowest unem- ployment rates (see Figure 5). 14 12.4 Figure 5: Unemployment rate 2019 (20 to 64 years) 12

10

8 6.3 6 5.6 4.0 4 3.4 2.2 2

0 Primary Secondary Tertiary

EU-28 Norway

Source: Eurostat (2020k), own illustration

Figure 7: Employment rate by educational attainment level and citizenship 2019 (20 to 64 years; ISCED 2011)

100 89.9 89.1 90 85.2 78.8 80 73.5 Source: Eurostat (2020k), own illustration 70 63.1 61.7 In general, there is a correlation between the level 60 54.0 48.6 of education and unemployment, according to 50 which the well-educated are significantly less 40 likely to be unemployed than people without a vo- 30 cational qualification. This correlation can also be 20 seen for Norway: Those with primary education 10 are most affected by unemployment (6.3 %). Con- 0 versely, people with tertiary education have the Primary Secondary Tertiary lowest unemployment rate (2.2 %), those with secondary a value of 3.4 %. Similar results are ob- EU-28 Non EU-28 Norway served at the European level (EU-28; see Figure 6). Source: Eurostat (2020l), own illustration

In contrast, the higher the educational attainment The abovementioned findings foster the theoreti- the higher is the employment rate, i.e. 88.6 % for cal conclusions by Becker (1964) that the educa- tertiary, 78.1 % for secondary and 60.8 % for pri- tional level has a significant influence on the pro- mary level. From an immigrational point of view, fessional career and the risk of unemployment. this observation holds for other citizenships as well. But EU-28 citizens have the highest employ- This theory and the empirical evidence can also be ment rate for all educational attainment levels applied to the “Early Leavers from Education and (89.9 %; 85.2 %; 63.1 %) while Norwegians rank Training”. This population group refers to persons

7

aged 18 to 24 who have completed a lower sec- citizens (mean: € 45,075 vs. € 38,604 and € 30,919; ondary education and are not involved in further median: € 42,069 vs. € 36,742 and € 28,110; see education or training. From a fiscal point of view, Figure 9). this specific group is of particular relevance as they A further analysis and breakdown of wage differ- are more likely to be (long-term) unemployed. In entials between domestic and foreign workers 2019, 10.3 % of the 18-24 year olds in the EU-28 done by Lehmer and Ludsteck (2013) reveals that were part of this group (male: 11.9 %; female: factors like seniority, promotions to better-paid 8.6 %). EFTA member state Norway ranges with a occupations as well as job stability also have to be value of 9.9 % below the EU average. In total, ap- taken into account. These factors cause an im- prox. one third (35.4 %) of early school leavers are provement in an individual's firm-specific human jobless but due to missing data, no definitive state- capital (training on the job) as the median age of ments can be made in regard to immigrational sta- foreign workers and therefore the length of time tus (Eurostat 2020m). within the company is lower than for domestic Income and Gross Domestic Product ones. However, these assumptions cannot be ver- ified in the context of this analysis due to a lack of As for the employment rate, the educational at- available data. tainment level has a positive impact on mean and median4 equivalized net income, i.e. the higher Figure 9: Mean and median equivalized net income by the educational attainment level the higher the in- broad group of citizenship 2018 (18 to 64 years) come. Working persons in Norway at primary level 50,000 earn less (mean: € 36,422; median: € 33,676) than 45,075 45,000 42,069 secondary (mean: € 42,843; median: € 41,226) 38,604 40,000 36,742 and tertiary levels (mean: 50,139; median: € 35,000 45,501; see Figure 8). 30,919 € 30,000 28,110 Figure 8: Mean and median equivalized net income by edu- 25,000 cational attainment level 2018 (18 to 64 years; ISCED 2011) 20,000 15,000 60,000 10,000 50,139 50,000 5,000 45,501 42,843 41,226 0 40,000 36,422 EU-28 Non EU-28 Norway 33,676

30,000 Mean equivalised net income Median equivalised net income

20,000 Source: Eurostat (2020o), own illustration

10,000 The indicator ‘people at risk of poverty or social exclusion’ corresponds to the sum of persons who 0 are: at risk of poverty after social transfers, se- Primary Secondary Tertiary verely materially deprived or living in households Mean equivalised net income with very low work intensity. As non-EU-28 citi- zens have a higher risk of unemployment and they Median equivalised net income Source: Eurostat (2020n), own illustration earn less than domestic and EU-28 citizens almost 3 out of 5 persons (58.2 %) belong to this group, Distinguished by citizenship, Norwegians have a though the value of EU-28 is quite high as well higher income compared to EU-28 and non-EU-28

4 The median is more robust, i.e. less sensitive to outliers than the mean. This explains why the mean equivalized income is always slightly higher for each educational level.

8

Table 2: Population aged 18 and over by risk of poverty, material deprivation, work intensity 2019

Risk of poverty At risk Not at risk At risk Not at risk At risk Not at risk Work Intensity Material Deprivation Norway EU-28 Non EU-28 Very low Severe 0.3 0.2 0.2 0.0 6.4 0.6 Not very low Severe 0.5 0.7 0.5 0.5 1.3 2.2 Very low Non severe 2.3 3.0 1.9 1.1 10.5 2.8 Not very low Non severe 8.1 12.7 18.5 People at Risk of Poverty or Social Exclusion 15.1 16.9 42.3 Not at risk Not at risk Not at risk Not very low Non severe 84.9 83.2 57.7 People at no Risk of Poverty or Social Exclusion 84.9 83.2 57.7 Total 100.0 100.0 100.0 Source: Eurostat (2020p), own illustration

(32.6 %). Norwegians face the lowest risk of social belong to the category ‘Others’ where nor further exclusion and poverty (13.8 %; see Table 2). information is available (see Table 3).

In 2019, compensation of employees (wages and A more detailed examination of the different sec- salaries plus employers’ social contributions) was tors of the economy reveals that there have also the largest income component of GDP in EU-28 been significant shifts within the sectors: the only and the Euro area, accounting for 47.8 % and declines in employment were recorded in the sec- 48.0 %. The EFTA country Norway (48.6 %) almost tions "Manufacture" (-17.7 %), "Transportation equals European average (both EU-28 and Euro and storage" (-8.6 %) and "Financial and insurance area). Taxes on production and imports (less sub- activities" (-8.1 %). "Food and beverage service ac- sidies) accounted for 10.5 % (EU-28: 11.9 %; Euro tivities" (+38.6 %), "Mining and quarrying" area: 11.5 %). Gross operating surplus and mixed (+35.7 %) as well as “Power and water supply, income accounted for 40.9 % of GDP in Norway, sewerage/remediation activities" (+24.6 %), on 40.3 % of GDP for the EU-28 and 40.5 % in the Euro the other hand, are among those economic sec- area. Norwegian GDP at current prices amounted tions with the largest growth rates. to approx. € 362.2 billion in 2019 and GDP per in- Due to non-available data no exact analyses are habitant equaled € 67,730 (Eurostat 2020q). possible in regard to TCN. The analysis by region of Economic structure and entrepreneurship birth shows that 19.5 % belong to the group of im- migrants (foreign-born with two foreign-born par- In general, a distinction is made between three dif- ents). Norwegian-born to immigrant parents are ferent sectors of the economy, the primary (agri- of minor relevance as they account for only 0.9 %. culture and forestry), secondary (manufacturing) and tertiary (services) sectors. Immigrants are overrepresented (above 19.5 %) in “Cleaning activities” (71.2 %), “Temporary em- In total, approx. 2.5 million people were employed ployment agency activities” (41.7 %; both subsec- in 2019 in Norway. This corresponds to an increase tions of “Administrative and support service activ- of +7.7 % compared to 2008. More than three ities” (37.1 %)), “Food and beverage service activ- quarters (77.6 %) are employed in the service sec- ities” (50.6 %), “Accommodation” (46.9 %), tor, where employment has increased by +9.8 % “Transportation and storage” (22.1 %) and “Con- since 2008. Employment in manufacturing evolved struction” (21.6 %). From a sectoral point of view, way below average (+3.1 %) and primary sector immigrants have almost the same relevance in the decreased by +25.4 % but it is of low importance, secondary (18.9 %) and tertiary sector (17.3 %; dif- i.e. only 1.8 % of total employment. Approx. 0.7 % ference of only 1.6 pp). But they are underrepre- sented in the primary sector (10.6 %; see Table 3).

9

Table 3: Employment by economic activity 2019 Employed Born to Economic Activity persons in Share Δ 08-19 Norwegian immigrant Immigrant 2019 parents Primary Sector 46,495 1.8% -25.4% 89.3% 0.1% 10.6% Agriculture, forestry and fishing 46,495 1.8% -25.4% 89.3% 0.1% 10.6% Secondary Sector 500,496 19.9% 3.1% 80.6% 0.5% 18.9% Mining and quarrying 57,551 2.3% 35.7% 87.9% 0.3% 11.7% Manufacture 199,588 7.9% -17.7% 80.1% 0.4% 19.4% Power and water supply, 30,900 1.2% 24.6% 89.2% 0.4% 10.3% sewerage/remediation activities Construction 212,457 8.4% 21.0% 77.8% 0.5% 21.6% Tertiary Sector 1,950,013 77.6% 9.8% 81.4% 1.4% 17.3% Domestic trade 306,302 12.2% -3.2% 82.8% 1.9% 15.3% Transportation and storage 126,876 5.0% -8.6% 76.4% 1.5% 22.1% Other passenger land transport 28,119 1.1% 3.1% 56.1% 2.3% 41.7% Accommodation 24,734 1.0% 7.0% 52.3% 0.9% 46.9% Food and beverage service activities 57,236 2.3% 38.6% 46.8% 2.6% 50.6% Information and communication 96,192 3.8% 13.8% 84.9% 1.4% 13.6% Financial and insurance activities 45,911 1.8% -8.1% 91.0% 1.7% 7.3% Real estate, professional, scientific 161,816 6.4% 15.7% 84.7% 1.1% 14.2% and technical activities Administrative and support service activities 127,550 5.1% 5.6% 61.0% 1.9% 37.1% Temporary employment agency activities 38,930 1.5% -7.5% 55.9% 2.4% 41.7% Cleaning activities 24,783 1.0% 28.3% 28.1% 0.6% 71.2% Public administration and defence: 159,222 6.3% 17.4% 92.7% 1.0% 6.4% compulsory social security Education 216,294 8.6% 15.0% 87.1% 0.9% 12.0% Human health and social work activities 532,807 21.2% 16.2% 83.5% 1.2% 15.4% Other service activities 95,073 3.8% 19.8% 81.6% 1.1% 17.3% Others 17,501 0.7% 72.6% 79.6% 0.9% 19.5% Total 2,514,505 100.0% 7.7% 81.4% 1.1% 17.5% Source: Statistics Norway (2021)

Research and Innovation funded by the private non-profit and the higher education sector (Eurostat 2020r). Following the neoclassical and endogenous growth theories, technological advance is believed Data on regional level (NUTS 2) is published with a to be one of the major drivers of economic one year time lag. Trøndelag is the province with growth. From this perspective, there is a growing the highest ratio in 2017 (4.79 %) succeeded by interest to investigate the link between research & Oslo og Akershus (3.18 %) and Vestlandet development (R&D), innovation, entrepreneur- (2.27 %). Except for Nord-Norge (2.00 %) all other ship and economic growth achieved by human provinces were well behind and below the all-Nor- capital from abroad (EU-28 and non-EU-28). wegian ratio of 2.10 %. Innlandet (0.92 %), Agder og Rogaland (1.50 %) and Sør-Østlandet (1.65 %) A well-known indicator provided to measure are the regions with the lowest ratios (see Table achievements of countries or regions in R&D is 4). GERD, i.e. regional/national gross domestic ex- penditure on R&D as a percentage of GDP. GERD In accordance to the research & development is estimated at 2.05 % for Norway (2018) which is (R&D) funding structure one half (48.9 %) of the a slight decrease of 0.05 pp to 2017. The total researchers work in the business enterprise sec- amount of research expenditures was € 7.58 bil- tor. Approx. one third (38.5 %) works in the higher lion, the largest share (47.8 %) was financed by the education sector and the remaining 12.6 % in the government sector and 42.0 % by the business en- governmental sector. This makes a total of terprise sector. 8.3 % were contributed by foreign 35,897 FTE (Eurostat 2020t). countries (rest of the world). 1.5 % and 0.5 % were

10

Available data for other countries has shown that and EU-28 (incl. UK) is 5.7 % and 3.8 %, therefore employees in R&D have an above average share of Norway records a comparably high rate of TCN. foreigners compared to the total number of em- Within the past two decades the increase in popu- ployed persons by economic activity. lation of working-age in Norway has strongly be Table 4: Intramural R&D expenditure as percentage of dominated by immigration, although also a natu- gross domestic product (GDP) 2017 ral population increase can be recorded for some Expenditures % of Province regions. (€ mio.) GDP Oslo og Akershus 3,098.3 3.18% Despite the relevance of migration for labor sup- Innlandet 155.3 0.92% Sør-Østlandet 718.4 1.65% ply, the employment rate of TCN is much lower Agder og Rogaland 622.9 1.50% compared to Norwegian nationals and EU-28 citi- Vestlandet 1,136.9 2.27% zens. One relevant factor to be considered here is Trøndelag 1,178.0 4.79% the lower educational attainment level of mi- Nord-Norge 507.1 2.00% grants, with the consecutive effects of lower in- Norway 7,416.8 2.10% Source: Eurostat (2020s), own illustration come levels and a higher risk of poverty.

In accordance to the research & development The pattern of education of TCN differs substan- (R&D) funding structure one half (48.9 %) of the tially from that of Norwegian nationals: The for- researchers work in the business enterprise sec- eign population of third country nationals are dis- tor. Approx. one third (38.5 %) works in the higher proportionately represented in primary and sec- education sector and the remaining 12.6 % in the ondary education, whereas they are underrepre- governmental sector. This makes a total of sented in tertiary education levels. 35,897 FTE (Eurostat 2020t). From a spatial point of view, the concentration of Available data for other countries has shown that better-educated migrants is highest in Norwegian employees in R&D have an above average share of cities. Regions with a higher ratio of TCN are also foreigners compared to the total number of em- characterized by higher economic growth, being ployed persons by economic activity. more attractive for TCNs but also offering more la- bor supply. Summary Governments need to invest in education on the This Statistical Briefings examines the economic one hand and in research and development (R&D) and spatial impact of immigration with special re- on the other, to enable long-term prosperity and gard to third-country nationals (TCN) in Norway. growth. Currently the national gross domestic ex- On January 1, 2020, a total of 940,580 persons penditures on R&D (as a percentage of GDP) are with non-Norwegian citizenship were living in Nor- estimated at 2.05 % 2018) which is a slight de- way. This corresponded to a share of 11.3 % of crease of 0.05 percentage points to 2017. But Norway's total population. Among non-Norwegian there is again a striking regional variation, from nationals, 61.7 % (6.9 % of total population) came densely populated areas with a high number of in- from EU-28 (incl. UK) and EFTA countries (excl. novative enterprises (e.g. Oslo) to less densely Norway; 0.2 %). A total of 426,535 persons (or populated provinces (e.g. Innlandet) 4.1 % of total population) were third country na- Summing up, immigrants in Norway play a pivotal tionals (TCN), being individuals who are neither role in reviving rural areas, and with the appropri- Norwegian, EU-28 citizens nor EFTA citizens. By ate policies, they could play an even more signifi- comparison, the share of TCN in EU-27 (excl. UK) cant role in sustaining them.

11

Bibliography and source references

Becker, Gary Stanley (1964): Human Capital: A theoretical Eurostat (2020o): Mean and median income by broad group and empirical analysis, with special reference to education. of citizenship (population aged 18 and over), 3rd edition. New York. https://appsso.eurostat.ec.europa.eu/nui/show.do?da- taset=ilc_di15&lang=en Eurostat (2020a): Population on 1 January by age group, sex and citizenship, https://appsso.eurostat.ec.eu- Eurostat (2020p): Persons aged 18 and over by risk of pov- ropa.eu/nui/show.do?dataset=migr_pop1ctz&lang=en erty, material deprivation, work intensity of the household and most frequent activity status of the person - intersec- Eurostat (2020b): Population change - Demographic balance tions of Europe 2020 poverty target indicators, and crude rates at regional level (NUTS 3), https://appsso.eurostat.ec.europa.eu/nui/show.do?da- https://appsso.eurostat.ec.europa.eu/nui/show.do?da- taset=ilc_pees02&lang=en taset=demo_r_gind3&lang=en Eurostat (2020q): GDP and main components (output, ex- Eurostat (2020c): Population by sex, age, citizenship, labour penditure and income), https://appsso.eurostat.ec.eu- status and NUTS 2 regions, https://appsso.eurostat.ec.eu- ropa.eu/nui/show.do?dataset=nama_10_gdp&lang=en ropa.eu/nui/show.do?dataset=lfst_r_lfsd2pwn&lang=en Eurostat (2020r): GERD by sector of performance, Eurostat (2020d): Population on 1st January by age, sex, type https://appsso.eurostat.ec.europa.eu/nui/show.do?da- of projection and NUTS 3, https://appsso.eurostat.ec.eu- taset=rd_e_gerdtot&lang=en ropa.eu/nui/show.do?dataset=proj_19rp3 Eurostat (2020s): Intramural R&D expenditure (GERD) by Eurostat (2020e): City, NUTS 2 regions, http://appsso.eurostat.ec.eu- https://ec.europa.eu/eurostat/statistics- ropa.eu/nui/show.do?dataset=rd_e_gerdreg explained/index.php?title=Glossary:City Eurostat (2020t): Total R&D personnel and researchers by Eurostat (2020f): Town or suburb, https://ec.europa.eu/eu- sectors of performance, educational attainment level rostat/statistics-explained/index.php?title=Glos- (ISCED2011) and sex, https://ec.europa.eu/eurostat/data- sary:Town_or_suburb browser/view/rd_p_persqual11/default/table?lang=en Eurostat (2020g): Urban centre, Lehmer, Florian; Ludsteck, Johannes (2013): Lohnanpassung https://ec.europa.eu/eurostat/statistics- von Ausländern am deutschen Arbeitsmarkt: Das explained/index.php?title=Glossary:Urban_centre Herkunftsland ist von hoher Bedeutung, IAB-Kurzbericht, Eurostat (2020h): Rural area, https://ec.europa.eu/euro- 01/2013), Nuremberg. stat/statistics-explained/index.php/Glossary:Rural_area Statistics Norway (2021): Employment among immigrants, Eurostat (2020i): Rural grid cell, https://ec.europa.eu/euro- register-based, https://www.ssb.no/en/statbank/ta- stat/statistics-explained/index.php?title=Glossary:Ru- ble/13215/ ral_grid_cell

Eurostat (2020j): Population by educational attainment level, sex, age, citizenship and degree of urbanization, http://appsso.eurostat.ec.europa.eu/nui/show.do?da- taset=edat_lfs_9916&lang=en

Eurostat (2020k): Unemployment rates by sex, age, educa- tional attainment level and NUTS 2 regions (%), https://appsso.eurostat.ec.europa.eu/nui/show.do?da- taset=lfst_r_lfu3rt&lang=en

Eurostat (2020l): Employment rates by sex, age, educational attainment level, citizenship and NUTS 2 regions, https://ec.europa.eu/eurostat/data- browser/view/tepsr_wc140/default/table?lang=en

Eurostat (2020m): Early leavers from education and training by sex and citizenship, http://appsso.eurostat.ec.eu- ropa.eu/nui/show.do?dataset=edat_lfse_01&lang=en

Eurostat (2020n): Mean and median income by educational attainment level - EU-SILC survey, http://appsso.euro- stat.ec.europa.eu/nui/show.do?lang=en&dataset=ilc_di08

12

ECONOMIC IMPACT OF MIGRATION

Statistical Briefings Spain

Call: H2020-SC6-MIGRATION-2019

Work Programmes:

H2020-EU.3.6.1.1. The mechanisms to promote smart, sustainable and inclusive growth H2020-EU.3.6.1.2. Trusted organisations, practices, services and policies that are necessary to build resilient, inclusive, participatory, open and creative societies in Europe, in particular taking into account migration, integration and demographic change

Deliverable 4.2 – Statistical briefing for Spain on economic impact of TCNs in MATILDE regions

Authors: Birgit Aigner-Walder, Albert Luger, Rahel M. Schomaker with contributions from Raúl Lardiés Bosque

Approved by Work Package Manager of WP4: Simone Baglioni, UNIPR Approved by Scientific Head: Andrea Membretti, UEF Approved by Project Coordinator: Jussi Laine, UEF

Version: 28.05.2021

DOI: 10.5281/zenodo.4817376

This document was produced under the terms and conditions of Grant Agreement No. 870831 for the European Commission. It does not necessary reflect the view of the European Union and in no way anticipates the Commission’s future policy in this area.

2

Introduction

This document presents the results of an assessment of the impact of migration and a measure- ment of the contribution provided by ‘third country nationals’ (TCNs) to the economic systems in the receiving contexts. In total, 10 statistical briefings for the MATILDE regions have been compiled, including the following countries: Austria, Bulgaria, Finland, Germany, Italy, Norway, Spain, Sweden, Turkey and the United Kingdom. Each document focuses on the following dimen- sions of economic development: economic growth, labour markets, innovation, and entrepre- neurship. Moreover, a short introduction on the relevance of TCNs in the respective country as well as the population development and structure are considered. The results of a literature review and a comparative analysis of the 10 statistical briefings are published separately.

Within the Matilde project there is a special focus on so-called ‘third country nationals’ (TCNs). MATILDE considers TCNs as Non EU citizens, who reside legally in the European Union (EU) and who are the target of EU integration policies. A TCN is “any person who is not a citizen of the European Union within the meaning of Art. 20(1) of TFEU and who is not a person enjoying the European Union right to free movement, as defined in Art. 2(5) of the Regulation (EU) 2016/399 (Schengen Borders Code).“ (European Commission 2020a). According to this definition, nationals of European Free Trade Association (EFTA) member countries Norway, Island, Liechtenstein and Switzerland are not considered to be TCNs. TCNs cover economic migrants, family migrants, stu- dents and researchers, highly skilled migrants, and forced migrants.

3

SPAIN Figure 1: Origin and share of total population of different nationalities in Spain

Source: Eurostat (2020a), own illustration Third Country Nationals (TCNs) in Spain Considering the nationalities of TCN, Moroccans (1.61 %) are above Colombians (0,55 %), Chinese On January 1, 2020, a total of 4,960,467 persons (incl. Hong Kong; 0.42 %), Venezuelans (0.40 %) with non-Spanish citizenship were living in Spain. and Ecuadorians (0.28 %). In total, more than half This corresponded to a share of 10.5 % of Spain's of TCN (51.8 %) living in Spain belong to these na- total population. tionalities. Figure 1 gives an overview of the origin Among non-Spanish nationals, 39.5 % and the share of the different nationalities in (1,958,662 persons; 4.1 % of total population) Spain. came from EU-28 (incl. UK) and EFTA countries Population Development & Population Structure (30,775 persons; 0.06 %). A total of 2,971,030 per- sons (or 6.3 % of total population) were third The number of inhabitants in Spain increased in country nationals (TCN), being individuals who are the past and this development is projected to con- neither Spanish, EU-28 citizens nor EFTA citizens. tinue in the future. As of January 1, 2020, there By comparison, the share of TCN in EU- were 47,332,614 persons registered in Spain. This 27 (excl. UK) and EU-28 (incl. UK) is 5.7 % and corresponds to an increase of 6,862,432 citizens 3.8 %, therefore Spain lying above the average. (+17.0 %) within the last 20 years since 2000 (be- fore eastward enlargement of EU in 2004).

4

A long-term analysis of population development of the country’s surface and concentrated barely since the turn of the new millennium reveals clear 3.1 % of the population. This reflects the "two regional differences. Fuerteventura (+108.15 %), Spains": the urbanized and populated (which in- Eivissa, Formentera (+81.33 %), Lanzarote cludes the Mediterranean coast and islands), ver- (+58.04 %), Guadalajara (+49.29 %) and Mallorca sus the empty interior. The rural exodus began at (+38.63 %) recorded the highest increase of popu- the end of the nineteenth century in zones close lation. Though the majority of regions (79.9 %) to urban and industrial centers of Catalonia and recorded population growth there are also some the Basque Country but, during the first half of the regions suffering from a decline. Zamora (- twentieth century, high fertility and declining mor- 13.45 %), Ourense (-9.25 %), Lugo (-8.15 %), Pa- tality rates assured continuity and even demo- lencia (-7.96 %) and León are most affected (see graphic growth in rural spaces in general (Re- Figure 2). caño 2017).

On 1 January 2016, 60 % of Spanish municipalities While most discussions about changing demo- had fewer than 1,001 inhabitants, occupied 40 % graphic structures focus on population ageing and

Figure 2: Population development (Δ 2002 to 2020), NUTS 3

Source: Eurostat (2020b), own illustration

5

the sustainability of pension systems, urbanization while the working-age population of Spanish peo- etc., another economically relevant aspect is the ple shrank since 2006 (-126,300 inh.). From a re- labor supply, i.e. the population aged 15 to 64 gional perspective, the areas with the highest pop- years or 20 to 64 years in industrialized economies ulation growths are Melilla (+25.8 %), Balearic Is- and nations. From an economic perspective, labor lands ( 19.1 %) and Canary Islands (+13.9 %). Prin- (done by human beings) is an essential element in cipality of Asturias (-10.9 %), Castile-Leon (-8.2 %) the production of goods and services. The term ‘la- and Basque Community (-7.4 %) recorded the bor force’ comprises all of those who work for highest declines. gain, whether as employees, employers, or as self- Whereas there was an increase of labor force po- employed, and it includes the unemployed who tential recorded in the past 20 years, this changes are seeking for work. in the future as working-age population decreases As Figure 3 (a)-(d) indicate, the increase in the by 6.2 % until 2040. There are only 5 regions working-age population (+416,500) within the re- (8.5 %) which are projected to grow in the future gions is mainly caused by the EU-28 (except Span- (see Figure 3 (a) and Figure 4). ish; +367,200 inh.) and non-EU 28 (+174,300 inh.)

Figure 3: Working-age Population (20 to 64 years; Δ 2006 to 2019) NUTS 2 (a) Total +416,500 inh. (c) Non EU-28 +174,300 inh.

(b) Spanish -126,300 inh. (d) EU-28 +367,200 inh.

Source: Eurostat (2020c), own illustration

6

Figure 4: Projections of Working-age Population (20 to 64 years; Δ 2020 to 2040) NUTS 3

Source: Eurostat (2020d), own illustration The total-age dependency ratio is a measure of the child ratios. Most metropolitan regions in Spain age structure of the population. It relates the num- are also characterized by relatively low rates. High ber of individuals who are likely to be “dependent” dependency ratios are found in Ourense on the support of others for their daily living – the (80.89 %), Zamora (78.35 %) and Lugo (74.76 %). young (up to 19 years old) and the elderly These provinces, together with others, constitute (65 plus years old) – to the number of those indi- the rural spaces at risk of irreversible depopula- viduals who are, being working age from 20 to 64 tion (Recaño, 2017). As a result of ageing societies, years old, capable of providing this support. the total-age dependency ratios will rise in the fu- ture by 18.5 pp to 83.0 % in 2040 resulting in im- The total-age dependency ratio in is plications/challenges for publicly funded social se- 64.6 %. The dependency rate is usually lower in curity schemes (e.g., expenditures for pensions). metropolitan areas and cities than in rural regions. The key factors here are a more prosperous econ- omy, i.e. more people in working-age, and low

7

Education and (Un)Employment The unemployment rate in Spain is 13.8 %, being much above the EU-28 average of 6.2 %. From a The tertiary education rate of Spain's domestic regional perspective, Melilla has the highest un- population is 36.5 %, which is only slightly above employment rate at 26.1 % (female: 31.9 %; that of the total population (35.1 %). Compared to male: 21.5 %), closely followed by Ceuta (25.4 %; the EU-28 (29.5 %) Spain ranks 10th. Luxembourg female: 28.9 %; male: 22.9 %) and Extremadura (41.0 %), Ireland (40.7 %) and United Kingdom (21.2 %; female: 27.4 %; male: 16.2 %). Navarre (40.6 %) rank at the top end, while the Czech Re- (8.0 %; 9.0 % female; 7.1 % male), Basque Com- public (21.6 %), Italy (17.4 %) and Romania munity (8.9 %; 9.2 % female; 8.7 % male) and La (16.0 %) rank at the bottom. Rioja (9.6 %; 9.7 % female; 9.6 % male) are the Across the world, educational attainment is signif- provinces with the lowest unemployment rates icantly higher in cities1 than in towns and suburbs2, (see Figure 5). 3 which in turn is higher than in rural areas . In rural Table 1: Working-age Population by citizenship, degree of areas in Spain, only 25.8 % of individuals have uni- urbanization and educational attainment level 2019 versity degrees, i.e. tertiary education (ISCED2011 Educational Attainment Level levels 5 to 8), compared to 29.8 % in towns & sub- Citizenship (ISCED11) Urba- urbs and 40.6 % in cities. Secondary degrees are nisation Primary Secondary Tertiary almost evenly spread over cities (25.7 %), towns Spanish 32.8% 24.9% 42.2% and suburbs (25.1 %) as well as rural areas EU-28 23.0% 34.9% 42.1%

Cities Cities Non EU-28 46.8% 28.5% 24.7% (23.9 %). But the regional differences in primary Total 33.8% 25.7% 40.6% educational levels are striking. While less than Spanish 44.6% 24.2% 31.2% 33.8 % of city residents have a primary, i.e. less EU-28 33.8% 38.9% 27.3% than primary, primary and lower secondary edu- Non EU-28 57.0% 25.6% 17.4% suburbs Towns & Towns cation (ISCED2011 levels 0 to 2) attainment level, Total 45.0% 25.1% 29.8% Spanish 50.2% 23.2% 26.6% 45.0 % and 50.3 % of residents in towns & suburbs EU-28 41.5% 36.8% 21.7% and rural areas do. Rural Non EU-28 65.1% 22.2% 12.7% Total 50.3% 23.9% 25.8% Concerning the educational attainment level of Source: Eurostat (2020j), own illustration TCN, big differences can be recorded. In cities, Figure 5: Unemployment rate 2019 (20 to 64 years) where most of the non-EU-28 citizens live, more than 45 % have only primary education. This share is even higher in towns & suburbs (57.0 %) and ru- ral areas (65.1 %). On contrary, only 24.7 % of non- EU-28 citizens have a degree from university, but this value is still much higher than in suburban (17.4 %) and rural (12.7 %) regions. Differences of educational attainment level between Spanish and EU-28 residents range within a broad band from 0.1 pp to 14.7 pp and are therefore scarcely comparable, too (see Table 1). In total, EU-28 (33.4 %) and non-EU-28 (21.6 %) citizens have a lower tertiary education rate than the domestic Source: Eurostat (2020k), own illustration population (36.5 %).

1 A city is a local administrative unit (LAU) where at least 50 density of at least 1 500 inhabitants per km² and collectively % of the population lives in one or more urban centres (Eu- a minimum population of 50 000 inhabitants after gap-filling rostat 2020e). (Eurostat 2020f; Eurostat 2020g). 2 Towns & suburbs are areas where less than 50 % of the 3 A rural area is an area where more than 50 % of its popula- population lives in rural grid cells and less than 50 % of the tion lives in ‘rural grid cells’, i.e. grid cells that are not identi- population lives in urban centres, i.e. a cluster of contiguous fied as urban centres or as urban clusters (Eurostat 2020h; grid cells of 1 km² (excluding diagonals) with a population Eurostat 2020i).

8

The approximation of unemployment rates by age, ranks first for primary (63.7 % vs. 57.8 %) and sec- nationality and educational level shows that they ondary educational attainment level (71.5 % vs. are influenced by age, education, and immigrant 64.8 %). non-EU-28 citizens have the lowest em- status. The youngest groups (but also the oldest ployment rate on each level (64.0 %; 62.0 %; ones) have higher unemployment rates than the 56.5 %; see Figure 7). central age groups. People with only Spanish na- Figure 7: Employment rate by educational attainment level tionality show lower unemployment rates than and citizenship (20 to 64 years; ISCED 2011) foreign nationals (CES, 2019). 90 In general, there is a correlation between the level 81.5 80 75.3 of education and unemployment, according to 71.5 70 which the well-educated are significantly less 63.7 62.0 64.8 64.0 57.8 likely to be unemployed than people without a vo- 60 56.5 cational qualification. This correlation can also be 50 seen for Spain. Those with primary education are 40 most affected by unemployment (19.8 %). Con- 30 versely, people with tertiary education have the 20 lowest unemployment rate (8.7 %), those with 10 secondary a value of 14.1 %. The same trend can 0 be observed at the European level (EU-28; Figure Primary Secondary Tertiary 6).

Figure 6: Unemployment rate by educational attainment EU-28 Non EU-28 Spain level 2019 (20 to 64 years; ISCED 2011) Source: Eurostat (2020l), own illustration 25 The abovementioned findings foster the theoreti- 19.8 cal conclusions by Becker (1964) that the educa- 20 tional level has a significant influence on the pro- fessional career and the risk of unemployment of 15 14.1 12.4 human beings.

10 8.7 This theory and the empirical evidence can also be

5.6 applied to the “Early Leavers from Education and 5 4.0 Training”. This population group refers to persons aged 18 to 24 who have completed a lower sec- 0 ondary education and are not involved in further Primary Secondary Tertiary education or training. From a fiscal point of view, this specific group is of particular relevance as they EU-28 Spain are more likely to be (long-term) unemployed. In 2019, 10.3 % of the 18-24 year olds in the EU-28 Source: Eurostat (2020k), own illustration were part of this group (male: 11.9 %; female: In contrast, the higher the educational attainment 8.6 %). Spain ranges with a value of 17.3 % above the higher the employment rate, i.e. 80.3 % for the European average. The share of EU-28 citizens tertiary, 65.0 % for secondary and 57.8 % for pri- residing in Spain is twice as high (27.9 %) and that mary level. From an immigrational point of view, of non-EU-28 is more than two times as high this observation holds for other citizenships as (27.9 %) than that of the domestic population well, as the risk of unemployment decreases by (14.7 %). In total, approx. 50 % (50.3 %) of all early the level of educational attainment. But Spanish leavers are jobless but no relying statements can have the highest employment rate for tertiary ed- be made in regard to immigrational status because ucational attainment level (81.5 %) while EU-28 of missing data (Eurostat 2020m).

9

Income and Gross Domestic Product capital (training on the job) as the median age of foreign workers and therefore the length of time As for the employment rate, the educational at- within the company is lower than for domestic tainment level has a positive impact on mean and ones. However, these assumptions cannot be ver- median4 equivalized net income, i.e. the higher ified in the context of this analysis due to a lack of the educational attainment level the higher the in- available data. come. Figure 9: Mean and median equivalized net income by Working persons in Spain at primary level earn less broad group of citizenship 2018 (18 to 64 years) (mean: € 13,252; median: € 12,143) than second- 20,000 ary (€ 16,518 or € 14,975) and tertiary levels 18,474 18,000 (€ 23,035 or € 20,999; see Figure 8). 16,303 16,000 Figure 8: Mean and median equivalized net income by edu- 14,000 cational attainment level 2018 (18 to 64 years; ISCED 2011) 12,182 12,000 9,897 10,328 25,000 10,000 8,954 23,035 20,999 8,000 20,000 6,000 16,518 4,000 14,975 2,000 15,000 13,252 12,143 0 EU-28 Non EU-28 Spain 10,000

Mean equivalised net income 5,000 Median equivalised net income

Source: Eurostat (2020o), own illustration 0 Primary Secondary Tertiary In Spain, most indicators show that vulnerability among foreign immigrants is higher than among Mean equivalised net income the local population, both in the workplace and in Median equivalised net income health, or other living conditions. Immigrants of- Source: Eurostat (2020n), own illustration ten work in low-income, precarious occupations, As working Spanish people have a higher educa- and tend to live in more overcrowded households tional attainment level than EU-28 and non-EU-28 (CES, 2019). this could be one reason (amongst others) that their mean (€ 18,474 vs. € 12,182 and € 10,328) The indicator ‘people at risk of poverty or social and median (€ 16,303 vs. € 9,897 and € 8,954) exclusion’ corresponds to the sum of persons who equivalized net income is higher (see Figure 9). are: at risk of poverty after social transfers, se- verely materially deprived or living in households A further analysis and decomposition of wage dif- with very low work intensity. As non-EU 28 citizens ferentials between domestic and foreign workers have a higher risk of unemployment and they earn done by Lehmer and Ludsteck (2013) reveals that less than domestic residents and EU-28 more than factors like seniority, promotions to better-paid one half (54.0 %) belongs to this group though the occupations as well as job stability also have to be value of EU-28 is quite high as well (46.7 %). Span- taken into account. These factors cause an im- ish face the lowest risk of social exclusion and pov- provement in an individual's firm-specific human erty (21.6 %; see Table 2).

4 The median is more robust, i.e. less sensitive to outliers than the mean. This explains why the mean equivalized income is always slightly higher for each educational level.

10

Table 2: Population aged 18 and over by risk of poverty, material deprivation, work intensity 2019 Risk of poverty At risk Not at risk At risk Not at risk At risk Not at risk Work Intensity Material Deprivation Spanish EU-28 Non EU-28 Very low Severe 1.0 0.3 1.4 0.1 4.9 0.2 Not very low Severe 1.0 1.3 2.9 3.5 8.5 2.3 Very low Non severe 3.2 3.5 4.0 0.6 4.5 1.4 Not very low Non severe 11.3 34.2 32.2 People at Risk of Poverty or Social Exclusion 21.6 46.7 54.0 Not at risk Not at risk Not at risk Not very low Non severe 78.5 53.3 46.0 People at no Risk of Poverty or Social Exclusion 78.5 53.3 46.0 Total 100.0 100.0 100.0 Source: Eurostat (2020p), own illustration

A way to calculate gross domestic product (GDP) south (- 12.6 %; Melilla: - 14.8 %). The highest re- is, to sum up, all the income earned by factors of gional GDP per capita at current prices occurred production from firms in the economy—the wages also in Madrid (approx. € 35,900) succeeded by earned by labor; the interest paid to those who northeastern regions (Basque Community (ap- lend their savings to firms and the government; prox. € 34,100); Navarre (approx. € 32,100)). the rent earned by those who lease their land or As in previous years, the southern provinces were structures to firms; and dividends, the profits paid far below the national level of € 26,400 (Melilla: to the shareholders, the owners of the firms’ phys- 19,200; Andalusi 21,600) ical capital. This is a valid measure because the € a: € 19,600; Murcia: € but also regions in the center (Extremadura: money firms earn by selling goods and services 19,500; Castile-L 21,000) as well as must go somewhere; whatever is not paid as € a Mancha: € 21,200) performed below average. Ex- wages, interest or rent is profit. Ultimately, profits Canarias (€ cept for Valencian Community 23,200) all east- are paid out to shareholders as dividends (€ ern regions rank above average (Balearic Islands: (Krugman and Wells 2017). € 28.200; Catalonia: € 31.100, see Table 3). In 2019, compensation of employees (wages and Provinces with a high share of TCN have a high salaries plus employers’ social contributions) was GDP per capita and vice versa. On the one hand, the largest income component of GDP in EU-28 prosperous regions are more attractive to mi- and the Euro area, accounting for 47.8 % and grants; on the other hand, these regions benefit 48.0 %. Spain (45.9 %) lies below European aver- from a better availability of labor supply, i.e. age (both EU-28 and Euro area). Taxes on produc- greater number of working age people (Amuedo- tion and imports (less subsidies) accounted for Dorantes & De la Rica 2007). In this context, one 10.2 % (EU-28: 11.9 %; Euro area: 11.5 %). Gross speaks of a so-called cross-fertilization, which re- operating surplus and mixed income accounted sults from the interaction of these two factors. for 43.9 % of GDP in Spain, 40.3 % of GDP for the EU-28 and 40.5 % in the Euro area. Spanish GDP at Expressing GDP in PPS (purchasing power stand- current prices amounted to approx. € 1,244.8 bil- ards) eliminates differences in price levels be- lion in 2019 and GDP per inhabitant equaled tween countries or regions. There are significant € 26,430 (Eurostat 2020q). differences in levels of prosperity among Spanish regions. The most prosperous region in Spain is Within 20 years Spain’s gross value added (GVA) also the region of the capital Madrid (124 %). It is increased by +29.8 % (real growth rate). In terms over 58/56/52 pp richer than Melilla/Andalu- of regional GVA almost all NUTS 2 regions also rec- sia/Castile-La Mancha and Ceuta, which are the orded positive real growth rates, ranging from poorest. Basque Community ranks second (118 %) +17.3 % in Principality of Asturias to +38.8 % in Co- and Navarre third (111 %; see Table 3). munidad de Madrid since 2000; but there are also regions with a decline in economic growth in the

11

Table 3: Nominal regional GDP (2019) Economic structure and entrepreneurship and real growth rate (Δ 2002 to 2019) Total Per Capita In general, a distinction is made between three dif- ferent sectors of the economy, the primary (agri- culture and forestry), secondary (manufacturing) and tertiary (services) sectors.

The majority (75.5 %) of the employed population works in the service sector in Spain in 2020, while

Region in 2008 that figure was 64.5%, according to data GDP (in €) GDP (in from the Active Population Survey of the National

GDP (in €) €) per capita GDP (in Institute of Statistics (INE; see Table 4). Among the at current marketat current prices marketat current prices 2015=100; Δ 2002 to 2019) 2002 to Δ 2015=100; TCN, 74.5 % also work in the service sector, so the of the EU27 (from 2020) average EU27 the 2020) of (from added (GVA) at basic prices (Index, prices at basic (GVA) added percentage is similar. 11.8 % of those employed in Purchasing power standard (PPS, EU27 (PPS, standard power Purchasing Real growth rate of regional gross value gross value growth rate regional Real of from 2020), per inhabitant in percentage per inhabitant from 2020), this sector are foreigners in Spain in 2020, while Noroeste 102,382 25.7 23,800 82 Galicia 64,430 29.7 23,900 82 TCN represent 7.74 %. Between 2008 and 2020, Principality of the number of foreigners and TCN employed in the 23,765 17.3 23,300 80 Asturias service sector has decreased in absolute numbers, Cantabria 14,187 21.4 24,400 84 due to the incidence of the Covid-19 pandemic. Noreste 142,380 26.8 31,800 110 Basque However, its proportion has increased in this eco- 74,496 26.5 34,100 118 Community nomic sector, from 57.3 % to 73.5% among for- Navarre 20,974 31.0 32,100 111 eigners, and from 57.8 % to 74.5 % among TCN. La Rioja 8,867 23.9 28,200 97 Aragon 38,044 25.9 28,700 99 Regarding other economic sectors, TCN employed Comunidad 240,130 38.8 35,900 124 in agriculture represent 7.9 % of the total em- de Madrid ployed in 2020, compared to only 4.0 % of all Centro 123,292 25.7 22,400 77 Castile-Leon 59,795 19.1 24,900 86 workers holding Spanish nationality (3.6 % among Castile-La employed persons with Spanish nationality) see 42,820 34.0 21,000 72 Mancha Table 4). Also, TCN employed in industry are less Extremadura 20,677 28.0 19,500 67 represented (9.6% among TCN), compared to Este 386,629 29.1 28,000 96 Catalonia 236,814 30.1 31,100 107 14.7% of the total employed workers with Spanish Valencian nationality. In short, the weight of TCN in the labor 116,015 26.6 23,200 80 Community market, compared to the native population, is sim- Balearic 33,799 29.6 28,200 97 Islands ilar in services, higher in agriculture and construc- Sur 201,608 28.5 19,900 69 tion, and lower in industry. This reflects the pre- Andalusia 165,865 28.0 19,600 68 dominance of these workers in manual and low- Region of 32,356 34.0 21,600 75 skilled activities, usually in jobs not demanded by Murcia the native population, which are difficult to fill Ceuta* 1,765 -12.6 20,900 72 Melilla* 1,622 -14.8 19,200 66 (Cuadrado et al. 2008). Canarias 47,164 24.5 21,200 73 Spain 1,244,772 29.8 26,400 91 Almost 65 % of all employed workers in Spain are Source: Eurostat (2020r), own calculations and illustration; concentrated in five Autonomous Communities - Eurostat (2020s), own calculations and illustration; NUTS 2 level - (Andalusia, Catalonia, Community of Madrid, Valencian Community and Galicia); most of these regions possess the highest growth and economic and demographic vitality in the country (see Figure 10). These five regions also concen- trate 69 % of the employed TCN.

12

Table 4: Employed by nationality and economic sectors 2008 and 2020 (*)(*) 2008 Economic Total Share Spanish Share Foreigners Share Non EU-28 Share sectors Agriculture 960,475 4.2% 762,700 3.9% 194,450 5.7% 141,250 5.8% Manufacturing 3,449,775 15.0% 3,051,025 15.7% 371,800 10.8% 247,625 10.2% Construction 2,882,000 12.5% 2,123,925 11.0% 729,625 21.2% 515,800 21.3% Services 14,903,625 64.6% 12,754,425 65.8% 1,975,300 57.5% 1,398,175 57.8% Total 23,065,550 100.0% 19,388,550 100.0% 3,437,050 100.0% 2,420,125 100.0% 2020 Economic Total Share Spanish Share Foreigners Share Non EU-28 Share sectors Agriculture 765,350 4.0% 589,075 3.6% 161,275 6.9% 118,750 7.9% Manufacturing 2698225 14.1% 2,381,650 14.7% 250,625 10.8% 145,350 9.6% Construction 1,244,075 6.5% 985,825 6.1% 203,325 8.8% 121,600 8.1% Services 14,494,750 75.5% 12,228,500 75.6% 1,705,925 73.5% 1,122,225 74.5% Total 19,202,425 100.0% 16,185,025 100.0% 2,321,100 100.0% 1,507,950 100.0% (*) On the table, only people with one nationality (only Spanish or other), but not with double nationality. Source: Economically Active Population Survey (EPA), 2008 and 2020; National Statistics Institute (INE).

Figure 10: Employment by nationalities and autonomous communities* 2020 (*) 4,000,000

3,500,000

3,000,000

2,500,000

2,000,000

1,500,000

1,000,000

500,000

0

Spanish EU-28 Non EU-28

(*) On the table, only people with one nationality (only Spanish or other), but not with double nationality. Source: Economically Active Population Survey (EPA); National Statistics Institute (INE)

13

Apart from the Community of Madrid (with the Region of Murcia and Balearic Islands) concentrate capital), the more touristic and coastal regions 82.3 % of the employed TCN in the country (see have the bigger presence of TCN workers, and only Figure 10). On the contrary, their presence is much seven of them (Catalonia, Community of Madrid, lower in inland and more depopulated regions. Andalusia, Valencian Community, Canary Islands,

Table 5: Workers' occupations by nationality and branches of economic activity 2020 (*) Workers' occupations Total Share Spanish Share EU-28 Share Non EU-28 Share Directors and managers 784,025 3.4% 693,275 3.7% 66,925 2.2% 25,299 1.2% Scientific and intellectual 3,852,250 16.9% 3,518,800 18.7% 247,550 8.0% 146,875 7.1% technicians and professionals Technicians; 2,296,600 10.1% 2,102,400 11.2% 147,875 4.8% 66,925 3.2% support professionals Accounting, administrative 2,185,050 9.6% 1,993,250 10.6% 128,375 4.2% 71,500 3.5% and other clerical employees Workers in catering, per- sonal, protective services 4,651,475 20.5% 3,638,025 19.4% 783,150 25.4% 583,375 28.3% and vendors Skilled workers in the agricul- tural, livestock, forestry and 450,900 2.0% 393,325 2.1% 49,300 1.6% 32,825 1.6% fishing sectors Craftsmen and skilled work- ers in manufacturing and construction industries (ex- 2,349,375 10.3% 1,905,900 10.2% 348,625 11.3% 201,500 9.8% cept plant and machinery op- erators) Plant and machinery opera- 1,613,850 7.1% 1,357,625 7.2% 200,400 6.5% 111,850 5.4% tors and assemblers Elementary occupations 2,911,550 12.8% 1,919,350 10.2% 791,125 25.7% 581,800 28.2% Military occupations 115,300 0.5% 111,250 0.6% 5,000 0.2% 1,750 0.1% They left their last job 1,219,200 5.4% 933,800 5.0% 233,225 7.6% 168,025 8.1% more than 1 year ago Unemployed seeking first job 303,800 1.3% 209,225 1.1% 81,775 2.7% 71,075 3.4% Total 22,733,325 100.0% 18,776,175 100.0% 3,078,825 100.0% 2,062,750 100.0% (*) On the table, only people with one nationality (only Spanish or other), but not with double nationality. Source: Economically Active Population Survey (EPA), 2020; National Statistics Institute (INE).

There is heavy demand in critical sectors of the job and more exposed to economic fluctuations, many market, with an immediate and evident stratifica- of them lost their jobs. This situation generated tion. This means that Spanish workers are in the great inequality between foreigners and nationals top segment of the job market, while TCN are in in terms of employment and unemployment rates, the second or bottom segment with insecure jobs, in terms of proportion of employees in elementary low wages, harsh working conditions, working on occupations, proportion of temporary contracts, a temporary or even seasonal basis (CES 2019). and income differences (Godenau et al. 2014). Also, TCN workers have the most insecure jobs - According to the mentioned stratification, workers normally from the countryside (from the agricul- with Spanish nationality occupy more qualified, tural and livestock sector), and this issue has be- technical and professional positions ("Scientific come particularly evident during the health pan- and intellectual technicians and professionals", demic. also "Technicians; support professionals", "Ac- For example, after the economic crisis between counting, administrative and other clerical em- 2007 and 2014, 20 % of jobs in Spain were de- ployees", and also "Workers in catering, personal, stroyed. As many immigrants worked concen- protective services and vendors "). On the con- trated in the less qualified segment, less protected

14

trary, the presence of TCN is greater in less quali- of the researchers work in the business enterprise fied jobs, in the service sector ("Workers in cater- sector. 15.6 % (or 20,844 FTE) work in the govern- ing, personal, protective services and vendors ") ment sector and 0.2 % (or 256 FTE) in the private and in "Elementary occupations "(see Table 5). non-profit sector. The higher education sector (46.9 % or 62,542 FTE) is by far the greatest em- Research and Innovation ployer of researchers. This makes a total of Following the neoclassical and endogenous 133,213 FTE. Further analysis on educational at- growth theories, technological advance is believed tainment level reveals that most researchers to be one of the major drivers of economic (55.5 %) have tertiary education but not doctoral growth. From this perspective, there is a growing or equivalent level; 43.8 % belong to the latter interest to investigate the link between research & whereas 1.2 % have less than primary, primary, development (R&D), innovation, entrepreneur- secondary and post-secondary non-tertiary edu- ship and economic growth achieved by human cation (see Figure 11). capital from abroad (EU-28 and non-EU-28). Table 6: Intramural R&D expenditure as percentage of A well-known indicator provided to measure gross domestic product (GDP) 2017 Expenditures % of achievements of countries or regions in R&D is Province (€ mio.) GDP GERD, i.e. regional/national gross domestic ex- Noroeste 897.0 0.90% penditure on R&D as a percentage of GDP. GERD Galicia 591.0 0.94% is estimated at 1.24 % for Spain (2018), which is a Principality of Asturias 188.0 0.81% slight increase of 0.03 percentage points to 2017. Cantabria 118.0 0.86% Noreste 2,205.0 1.60% The total amount of research expenditures was Basque Community 1,451.0 2.01% € 14.95 billion, the largest share (49.2 %) was fi- Navarre 344.0 1.70% nanced by the business enterprise sector. 37.9 % La Rioja 70.0 0.82% were financed by the government sector, 8.1 %by Aragon 340.0 0.92% Comunidad de Madrid 3,923.0 1.70% foreign countries (rest of the world) and 4.0 % by Centro 1,104.0 0.93% higher education and private non-profit sector Castile-Leon 763.0 1.32% (Eurostat 2020t). Castile-La Mancha 219.0 0.53% Extremadura 122.0 0.61% Regional data is published on NUTS 1 and NUTS 2 Este 4,816.0 1.29% level. Basque Community has the highest ratio in Catalonia 3,513.0 1.54% 2018 (2.01 %). The metropolitan area of Madrid Valencian Community 1,174.0 1.06% and Navarre are second (1.70 %) just ahead of Cat- Balearic Islands 129.0 0.40% Sur 1,787.0 0.91% alonia (1.54 %), Castile-Leon (1.32 %) and Valen- Andalusia 1,479.0 0.92% cian Community (1.06 %). All other regions have Region of Murcia 303.0 0.96% ratios lower than 1.00 %. Canarias ranks last Ceuta* : 0.00% (0.47 %) with Castile-La Mancha (0.53 %), Extre- Melilla* : 0.00% madura (0.61 %), Principality of Asturias (0.81 %) Canarias 215.0 0.47% Spain 14,946.0 1.24% and La Rioja (0.82 %) only slightly better, but there Source: Eurostat (2020u), own illustration is no reliable data available for Ceuta and Melilla (see Table 6). As the group of migrants (EU-28 or non-EU-28) liv- ing in Spain has a lower share of tertiary education In accordance to the research & development it is therefore likely that they are underrepre- (R&D) funding structure almost more than one sented in Spanish R&D personnel. third (37.2 % or 49.571 full-time equivalent (FTE))

15

Fig ure 11: R&D researchers (full-time equivalent (FTE)) by sectors of performance and educational attainment level 2017 (ISCED2011) 45,000 39,891 39,722 40,000

35,000

30,000

25,000 22,554

20,000

15,000 10,653 10,134 10,000 8,388

5,000 1,292 267 1 133 122 0 58 Business enterprise sector Government sector Higher education sector Private non-profit sector

Less than primary, primary, secondary and post-secondary non-tertiary education (levels 0-4) Tertiary education excluding doctoral or equivalent level (levels 5-7) Doctoral or equivalent level

Source: Eurostat (2020v), own illustration

Summary

This Statistical Briefings examines the economic Despite the relevance of migration for labor sup- and spatial impact of immigration with special re- ply, the employment rate of TCN is much lower gard to third-country nationals (TCN) in Spain. compared to Spanish nationals and EU-28 citizens. Also noteworthy is the strong segmentation of the On January 1, 2020, a total of 4,960,467 persons labor market, which pushes immigrants to occupy with non-Spanish citizenship were living in Spain. jobs not desired by the native population. Another This corresponded to a share of 10.5 % of Spain's relevant factor to be considered is the lower edu- total population. Among non-Spanish nationals, cational attainment level of migrants, with the 39.5 % (4.1 % of total population) came from EU- consecutive effects of lower income levels and a 28 (incl. UK) and EFTA countries (0.06 %). A total higher risk of poverty. of 2,971,030 persons (or 6.3 % of total population) were third country nationals (TCN), being individ- The pattern of education of TCN differs substan- uals who are neither Spanish, EU-28 citizens nor tially from that of Spanish nationals: The foreign EFTA citizens. By comparison, the share of TCN in population of third countries is disproportionately EU-27 (excl. UK) and EU-28 (incl. UK) is 5.7 % and represented in primary and secondary education, 3.8 %, therefore Spain lying above the average. whereas they are underrepresented in tertiary ed- ucation levels. Within the last 20 years the population increase of Spain has mainly be dominated by immigration. From a spatial point of view, the concentration of From 2006 to 2019 all Spanish regions profited better-educated migrants is highest in Spanish cit- heavily with regard to its working-age population. ies. Provinces with a higher ratio of TCN are also Without immigration, all regions would have suf- characterized by higher economic growth, being fered from a decrease in its population aged be- more attractive for TCN on the one hand and of- tween 20 and 64 years. fering more labor supply on the other hand.

16

The below average regional GVA growth rate, es- with a high number of innovative enterprises (e.g. pecially in rural provinces like Ceuta, Melilla is in- Madrid, Basque Community) to less densely popu- directly linked to immigration from abroad but pri- lated autonomous communities like Ceuta and marily to internal migration from rural areas to cit- Melilla. ies and metropolitan areas. These emigration Many rural areas in Spain suffer an acute problem flows tend to reinforce the erosion of the eco- of depopulation. In recent years, the arrival of for- nomic basis, because they have a negative impact eign immigrant workers has contributed to allevi- on the potential workforce and the economic at- ating the situation. Moreover, these immigrants tractiveness of the region. Economic performance find themselves in a very vulnerable situation with is stagnating, accompanied by declining popula- very limited resources; this is due to the lack of re- tion figures. sources and to the difficulties associated with the As a result, governments need to invest in educa- provision of social services in depopulated rural tion on the one hand and in research and develop- areas (Sampedro & Camarero 2018). ment (R&D) on the other, to enable long-term Summing up, immigrants in Spain play a pivotal prosperity and growth impulses. Currently, the na- role in reviving rural areas, and with the appropri- tional gross domestic expenditures on R&D (as a ate policies, they could play an even more signifi- percentage of GDP) are estimated at 1.24 % cant role in sustaining them. (2018), which is a slight increase of 0.03 percent- age points to 2017. But there is again a striking re- gional variation from densely populated areas

17

Bibliography and source references

Amuedo-Dorantes, C. & De la Rica, S. (2007): “Labour Market Eurostat (2020k): Unemployment rates by sex, age, educa- Assimilation of Recent Immigrants in Spain”. British Journal tional attainment level and NUTS 2 regions (%), of Industrial Relations, 45(2), 275-284. https://appsso.eurostat.ec.europa.eu/nui/show.do?da- taset=lfst_r_lfu3rt&lang=en Becker, Gary Stanley (1964): Human Capital: A theoretical and empirical analysis, with special reference to education. Eurostat (2020l): Employment rates by sex, age, educational 3rd edition. New York. attainment level, citizenship and NUTS 2 regions, https://ec.europa.eu/eurostat/data- CES (2019): La inmigración en España: efectos y oportunida- browser/view/tepsr_wc140/default/table?lang=en des. Informe 02|2019. Colección Informes. Madrid, Consejo Económico y Social-España (CES), 247 pp. Eurostat (2020m): Early leavers from education and training http://www.ces.es/documents/10180/5209150/Inf0219.pdf by sex and citizenship, http://appsso.eurostat.ec.eu- ropa.eu/nui/show.do?dataset=edat_lfse_01&lang=en Cuadrado, J.R.; Iglesias, C. & Llorente, R. (2008): Inmigración y mercado de trabajo. Análisis de algunas cuestiones funda- Eurostat (2020n): Mean and median income by educational mentales. Madrid, Fundación BBVA. attainment level - EU-SILC survey, http://appsso.euro- stat.ec.europa.eu/nui/show.do?lang=en&dataset=ilc_di08 EPA (2008; 2020): Economically Active Population Survey [Encuesta de Población Activa]. National Statistics Institute Eurostat (2020o): Mean and median income by broad group (INE), Madrid. https://www.ine.es/dyngs/INEbase/en/opera- of citizenship (population aged 18 and over), cion.htm?c=Estadistica_C&cid=1254736176918&menu=re- https://appsso.eurostat.ec.europa.eu/nui/show.do?da- sultados&idp=1254735976595 taset=ilc_di15&lang=en

Eurostat (2020a): Population on 1 January by age group, sex Eurostat (2020p): Persons aged 18 and over by risk of pov- and citizenship, https://appsso.eurostat.ec.eu- erty, material deprivation, work intensity of the household ropa.eu/nui/show.do?dataset=migr_pop1ctz&lang=en and most frequent activity status of the person - intersec- tions of Europe 2020 poverty target indicators, Eurostat (2020b): Population change - Demographic balance https://appsso.eurostat.ec.europa.eu/nui/show.do?da- and crude rates at regional level (NUTS 3), taset=ilc_pees02&lang=en https://appsso.eurostat.ec.europa.eu/nui/show.do?da- taset=demo_r_gind3&lang=en Eurostat (2020q): GDP and main components (output, ex- penditure and income), https://appsso.eurostat.ec.eu- Eurostat (2020c): Population by sex, age, citizenship, labour ropa.eu/nui/show.do?dataset=nama_10_gdp&lang=en status and NUTS 2 regions, https://appsso.eurostat.ec.eu- ropa.eu/nui/show.do?dataset=lfst_r_lfsd2pwn&lang=en Eurostat (2020r): Gross domestic product (GDP) at current market prices by NUTS 2 regions, https://appsso.euro- Eurostat (2020d): Population on 1st January by age, sex, type stat.ec.europa.eu/nui/show.do?da- of projection and NUTS 3, https://appsso.eurostat.ec.eu- taset=nama_10r_2gdp&lang=en ropa.eu/nui/show.do?dataset=proj_19rp3 Eurostat (2020s), Real growth rate of regional gross value Eurostat (2020e): City, added (GVA) at basic prices by NUTS 2 regions - percentage https://ec.europa.eu/eurostat/statistics- change on previous year, https://appsso.eurostat.ec.eu- explained/index.php?title=Glossary:City ropa.eu/nui/show.do?dataset=nama_10r_2gvagr&lang=en Eurostat (2020f): Town or suburb, https://ec.europa.eu/eu- Eurostat (2020t): GERD by sector of performance and source rostat/statistics-explained/index.php?title=Glos- of funds, http://appsso.eurostat.ec.eu- sary:Town_or_suburb ropa.eu/nui/show.do?dataset=rd_e_gerdfund&lang Eurostat (2020g): Urban centre, Eurostat (2020u): Intramural R&D expenditure (GERD) by https://ec.europa.eu/eurostat/statistics- NUTS 2 regions, https://ec.europa.eu/eurostat/data- explained/index.php?title=Glossary:Urban_centre browser/view/tgs00042/default/table?lang=en Eurostat (2020h): Rural area, https://ec.europa.eu/euro- Eurostat (2020v): Total R&D personnel and researchers by stat/statistics-explained/index.php/Glossary:Rural_area sectors of performance, educational attainment level Eurostat (2020i): Rural grid cell, https://ec.europa.eu/euro- (ISCED2011) and sex, https://ec.europa.eu/eurostat/data- stat/statistics-explained/index.php?title=Glossary:Ru- browser/view/rd_p_persqual11/default/table?lang=en ral_grid_cell Godenau, D.; Rinken, S., Martínez de Lizarrondo, A. & Eurostat (2020j): Population by educational attainment level, Moreno, G. (2014). La integración de los inmigrantes en Es- sex, age, citizenship and degree of urbanization, paña: una propuesta de medición a escala regional. Docu- http://appsso.eurostat.ec.europa.eu/nui/show.do?da- taset=edat_lfs_9916&lang=en

18

mentos del Observatorio, 30. Madrid, Observatorio Perma- nente de la Inmigración, Ministerio de Empleo y Seguridad Social.

Krugman, Paul, Wells, Robin (2017): Economics, Fifth Edition, New York.

Lehmer, Florian; Ludsteck, Johannes (2013): Lohnanpassung von Ausländern am deutschen Arbeitsmarkt: Das Herkunftsland ist von hoher Bedeutung, IAB-Kurzbericht, 01/2013), Nuremberg.

Recaño, Joaquín (2017): The Demographic Sustainability of Empty Spain. Perspectives Demogràfiques, 7, 1-4.

Sampedro, Rosario & Camarero, Luis (2018): Foreign Immi- grants in Depopulated Rural Areas: Local Social Services and the Construction of Welcoming Communities. Social Inclu- sion, 6 (3), 337-346.

19

ECONOMIC IMPACT OF MIGRATION

Statistical Briefings Sweden

Call: H2020-SC6-MIGRATION-2019

Work Programmes:

H2020-EU.3.6.1.1. The mechanisms to promote smart, sustainable and inclusive growth H2020-EU.3.6.1.2. Trusted organisations, practices, services and policies that are necessary to build resilient, inclusive, participatory, open and creative societies in Europe, in particular taking into account migration, integration and demographic change

Deliverable 4.2 – Statistical briefing for Sweden on economic impact of TCNs in MATILDE re- gions

Authors: Birgit Aigner-Walder, Albert Luger, Rahel M. Schomaker with contributions from Zuzana Macuchova and Ulf Hansson

Approved by Work Package Manager of WP4: Simone Baglioni, UNIPR Approved by Scientific Head: Andrea Membretti, UEF Approved by Project Coordinator: Jussi Laine, UEF

Version: 28.05.2021

DOI: 10.5281/zenodo.4817376

This document was produced under the terms and conditions of Grant Agreement No. 870831 for the European Commission. It does not necessary reflect the view of the European Union and in no way anticipates the Commission’s future policy in this area.

2

Introduction

This document presents the results of an assessment of the impact of migration and a measure- ment of the contribution provided by ‘third country nationals’ (TCNs) to the economic systems in the receiving contexts. In total, 10 statistical briefings for the MATILDE regions have been compiled, including the following countries: Austria, Bulgaria, Finland, Germany, Italy, Norway, Spain, Sweden, Turkey and the United Kingdom. Each document focuses on the following dimen- sions of economic development: economic growth, labour markets, innovation, and entrepre- neurship. Moreover, a short introduction on the relevance of TCNs in the respective country as well as the population development and structure are considered. The results of a literature review and a comparative analysis of the 10 statistical briefings are published separately.

Within the Matilde project there is a special focus on so-called ‘third country nationals’ (TCNs). MATILDE considers TCNs as Non EU citizens, who reside legally in the European Union (EU) and who are the target of EU integration policies. A TCN is “any person who is not a citizen of the European Union within the meaning of Art. 20(1) of TFEU and who is not a person enjoying the European Union right to free movement, as defined in Art. 2(5) of the Regulation (EU) 2016/399 (Schengen Borders Code).“ (European Commission 2020a). According to this definition, nationals of European Free Trade Association (EFTA) member countries Norway, Island, Liechtenstein and Switzerland are not considered to be TCNs. TCNs cover economic migrants, family migrants, stu- dents and researchers, highly skilled migrants, and forced migrants.

3

SWEDEN Figure 1: Origin and share of total population of different nationalities in Sweden

Source: Eurostat (2020a), own illustration Third Country Nationals (TCNs) in Sweden gives an overview of the origin and the share of the different nationalities in Sweden. On January 1, 2020, a total of 940,580 persons with non-Swedish citizenship were living in Swe- Population Development & Population Structure den. This corresponds to a share of 9.1 % of Swe- The number of inhabitants in Sweden increased in den's total population. Among non-Swedish na- the past and this development is projected to con- tionals, 34.3 % (322,324 persons; 3.1 % of total tinue according to current population forecasts. population) came from EU-28 (incl. UK) and EFTA As of January 1, 2020, there were 10,327,589 per- countries (40,855 persons; 0.4 %). A total of sons registered in Sweden. This corresponds to an 577,401 persons (or 5.60 % of total population) increase of 1,466,163 citizens (+16.5 %) within the were third country nationals (TCN), being individ- last 20 years since 2000 (before eastward enlarge- uals who are neither Swedish, EU-28 citizens nor ment of EU in 2004). EFTA citizens. By comparison, the share of TCN in EU-27 (excl. UK) and EU-28 (incl. UK) is 5.7 % and A long-term analysis of population development 3.8 %, therefore Sweden is at EU-28 average level. since the turn of the new millennium reveals clear regional differences. Norrbottens län (-1.8 %) is Considering the nationalities of TCN, Syrians the only Swedish region where the number of in- (1.13 %) are above Afghanis (0.48 %), Eritreans habitants declined, and stagnation took place in (0.42 %), Somalians (0,30 %), and Iranians Västernorrlands län (+0.1 %). With under-average (0.26 %). In total, almost half (46.2) % of TCN living in Sweden belong to these nationalities. Figure 1

2

population growth recorded in the north, the situ- According to current forecasts the increase of la- ation is different in the south: Uppsala län bor force potential does not stop in the future (+29.4%), Stockholms län (+29.3 %) as well as though the growth rates are not that high as they Skåne län (+ 21.2 %) have very high growth rates were in the past. Working-age population will (see Figure 2). evolve by 10.7 % until 2040. There are only two re- gions which are projected to shrink in the future, While most discussions about changing demo- i.e. Västernorrlands län (-2.11 %) and Norrbottens graphic structures focus on population ageing and län (-6.04 %; see Figure 4). the sustainability of pension systems, urbanization etc., another economically relevant aspect is labor The total-age dependency ratio is a measure of the supply, i.e. the population aged 15 to 64 years or age structure of the population. It relates the num- 20 to 64 years in industrialized economies and na- ber of individuals who are likely to be “dependent” tions. From an economic perspective, labor (done on the support of others for their daily living – the by human beings) is an essential element in the young (up to 19 years old) and the elderly production of goods and services. The term ‘labor (65 plus years old) – to the number of those indi- force’ comprises all of those who work for gain, viduals who are, being working age from 20 to 64 whether as employees, employers, or as self-em- years old, capable of providing this support. ployed, and it includes the unemployed who are The total-age dependency ratio in seeking for work. is 76.5 %. The dependency rate is usually lower in As Figure 3 (a)-(d) indicate, the increase in the metropolitan areas and cities than in rural regions. working-age population (+492,200 inh.) is domi- The key factors here are a more prosperous econ- nated by immigration (EU-28: +41,600 inh.; non- omy, i.e. more people in working-age, and low EU-28: +263,000 inh.). Domestic population in- child ratios. Most metropolitan regions (e.g., creased as well (+170,700 inh.) but on a smaller Stockholm) in Sweden are also characterized by scale than immigrants in total, though this makes relatively low rates. High dependency ratios are Sweden special, as most of the industrialized found in Dalarnas län (88.3 %), Kalmar län (88.1 %) countries had a tendency of shrinking domestic and Södermanlands län (87.1 %). working-age population. From a regional perspec- As a result of the ageing society, the total-age de- tive, Stockholm (+21.4 %), Sydsverige (+10.8 %) pendency ratios will rise in the future by 4.1 pp to and Östra Mellansverige (+7.5 %) are the growing 80.6 % in 2040, resulting in implications/chal- areas, stagnation takes place in Småland med lenges for publicly funded social security schemes öarna (1.3 %) and Norra Mellansverige (0.2 %), (e.g., expenditures for pensions). whereas Mellersta Norrland (-3.6 %) and Övre Norrland (-6.2 %) shrunk in the past.

3

Figure 2: Population development (Δ 2002 to 2020), NUTS 3

Source: Eurostat (2020b), own illustration

4

Figure 3: Working-age population (20 to 64 years; Δ 2006 to 2019) (a) (c)

(b) (d)

Source: Eurostat (2020c), own illustration

5

Figure 4: Projections of working-age population (20 to 64 years; Δ 2020 to 2040) NUTS 3

Source: Eurostat (2020d), own illustration

6

Education and (Un)Employment 1.4 pp to 26.7 pp with the highest variations be- tween secondary and tertiary levels of education The tertiary education rate of Sweden's domestic (see Table 1). In total, EU-28 (56.4 %) have a higher population is 37.6 %, which is only slightly below tertiary education rate than domestic population that of the total population (37.8 %). Compared to (37.6 %) which ranks above non-EU-28 (32.6 %) the EU-28 (29.5 %) Sweden has a way above-aver- citizens. age tertiary education rate. Sweden ranks 7th within the EU-28 next to Lithuania (37.9 %), ahead Table 1: Working-age Population by citizenship, degree of of Estonia (36.5 %), Belgium (36.0 %) and Spain urbanization and educational attainment level 2019 (35.1 %). Luxembourg (41.0 %), Ireland (40.7 %) Educational Attainment Level and United Kingdom (40.6 %) rank at the top end, Citizenship (ISCED11) while the Czech Republic (21.6 %), Italy (17.4 %) and Romania (16.0 %) rank at the bottom. Urbanisation Primary Secondary Tertiary Swedish 15.9% 37.0% 47.1% Across the world, educational attainment is signif- EU-28 14.5% 21.8% 63.6% 1 2 icantly higher in cities than in towns and suburbs , Cities Non EU-28 39.8% 18.6% 41.6% which in turn is higher than in rural areas. In rural Total 17.7% 35.1% 47.3% areas3 in Sweden, only 28.5 % of individuals have Swedish 19.1% 46.5% 34.5% EU-28 22.6% 24.9% 52.5% university degrees, i.e. tertiary education Non EU-28 53.9% 17.7% 28.5% suburbs (ISCED2011 levels 5 to 8), compared to 34.6 % in & Towns Total 21.6% 43.8% 34.6% towns & suburbs and 47.3 % in cities. In contrast, Swedish 21.4% 50.4% 28.2% secondary degrees are more common in less EU-28 27.6% 23.7% 48.7%

Rural Non EU-28 60.4% 16.1% 23.5% densely populated areas. While less than 35.1 % of Total 24.0% 47.5% 28.5% city residents have secondary, i.e. upper second- Source: Eurostat (2020j), own illustration ary and post-secondary non-tertiary education (ISCED2011 levels 3 to 4) attainment level, 43.8 % The unemployment rate in Sweden is 6.0 %, being and 47.5 % of residents in towns & suburbs and close to EU-28 average of 6.2 %. From a regional rural areas do. Primary educational attainment perspective, Sydsverige has the highest unemploy- level has low relevance as approx. only 1 out of 5 ment rate at 8.1 % (female: 8.2 %; male: 7.9 %), residents in Sweden belongs to this group. followed by Norra Mellansverige (6.7 %; fe- male: 6.6 %; male: 6.8 %) and Östra Mellansverige Concerning the educational attainment level of (6.3 %; female: 6.3 %; male: 6.3 %). Småland med TCN, big differences can be recorded. In cities, öarna (5.1 %; 5.7 % female; 4.6 % male) and Stock- where most of the non-EU-28 citizens live, approx. holm (5.2 %) are the provinces with the lowest un- 40 % (39.8 %) have only primary education. This employment rates (see Figure 5). share is even higher in towns & suburbs (53.9 %) and rural areas (60.4 %). The shares of non-EU-28 In general, there is a correlation between the level citizens with primary or tertiary education of education and unemployment, according to (41.6 %) living in cities are almost the same. On which the well-educated are significantly less contrary, only 28.5 % or 23.5 % of non-EU-28 citi- likely to be unemployed than people without a vo- zens living in towns and suburbs or rural areas cational qualification. This correlation can also be have a degree from university. Differences of edu- seen for Sweden. Those with primary education cational attainment level between Swedish and are most affected by unemployment (17.5 %). EU-28 residents range within a broad band from Conversely, people with tertiary education have

1 A city is a local administrative unit (LAU) where at least 50 density of at least 1 500 inhabitants per km² and collectively % of the population lives in one or more urban centres (Eu- a minimum population of 50 000 inhabitants after gap-filling rostat 2020e). (Eurostat 2020f; Eurostat 2020g). 2 Towns & suburbs are areas where less than 50 % of the 3 A rural area is an area where more than 50 % of its popula- population lives in rural grid cells and less than 50 % of the tion lives in ‘rural grid cells’, i.e. grid cells that are not identi- population lives in urban centres, i.e. a cluster of contiguous fied as urban centres or as urban clusters (Eurostat 2020h; grid cells of 1 km² (excluding diagonals) with a population Eurostat 2020i).

7

the lowest unemployment rate (3.8 %), those with In contrast, the higher the educational attainment secondary a value of 4.9 %. Similar results are ob- the higher is the employment rate, i.e. 88.8 % for served at the European level (EU-28; see Figure 6). tertiary, 82.9 % for secondary and 61.2 % for pri- mary level. From an immigrational point of view, Figure 5: Unemployment rate 2019 (20 to 64 years) this observation holds for other citizenships as well, as the risk of unemployment decreases by the highest educational attainment. But Swedes have the highest employment rate for tertiary (90.5 %) and secondary (83.7 %) educational at- tainment level, while EU-28 ranks first for primary (69.2 % vs. 66.7 %). Non-EU-28 citizens have the lowest employment rate on each level (67.2 %; 62.6 %; 43.2 %; see Figure 7).

Figure 7: Employment rate by educational attainment level and citizenship 2019 (20 to 64 years; ISCED 2011)

100 90.5 90 83.7 85.5 80 77.8 69.2 66.7 67.2 70 62.6 60 50 43.2 40 30 20 10 0 Primary Secondary Tertiary

Source: Eurostat (2020k), own illustration EU-28 Non EU-28 Sweden Figure 6: Unemployment rate by educational attainment level 2019 (20 to 64 years; ISCED 2011) Source: Eurostat (2020l), own illustration

20 The abovementioned findings foster the theoreti- 18 17.5 cal conclusions by Becker (1964) that the educa- 16 tional level has a significant influence on the pro-

14 12.4 fessional career and the risk of unemployment of 12 human beings. 10 This theory and the empirical evidence can also be 8 applied to the “Early Leavers from Education and 6 5.6 4.9 4.0 3.8 4 Training”. This population group refers to persons aged 18 to 24 who have completed a lower sec- 2 ondary education and are not involved in further 0 Primary Secondary Tertiary education or training. From a fiscal point of view, this specific group is of particular relevance as they

EU-28 Sweden are more likely to be (long-term) unemployed. In 2019, 10.3 % of the 18-24 year olds in the EU-28 Source: Eurostat (2020k), own illustration were part of this group (male: 11.9 %; female: 8.6 %). Sweden ranges with a value of 6.5 % below the European average. The share of non-EU-28 is

8

more than four times as high (21.0 %) than of the with very low work intensity. As non-EU 28 citizens domestic population (4.8 %). In total, approx. 50 % have a higher risk of unemployment and they earn (46.2 %) of early school leavers are jobless but no less than domestic and EU-28 residents almost 3 relying statements can be made in regard to immi- out of 5 persons (58.2 %) belong to this group, grational status because of missing data (Eurostat though the value of EU-28 is quite high as well 2020m). (32.6 %). Swedish face the lowest risk of social ex- clusion and poverty (13.8 %; see Table 2). Income and Gross Domestic Product Figure 9: Mean and median equivalized net income by As for the employment rate, the educational at- broad group of citizenship 2018 (18 to 64 years) tainment level has a positive impact on mean and median4 equivalized net income, i.e. the higher 35,000 the educational attainment level the higher the in- 29,630 30,000 27,782 come. 25,000 23,253 Working persons in Sweden at primary level earn 21,590 less (mean: 20,805; median: 19,485) than sec- 20,000 € € 16,160 ondary (mean: € 28,288; median: € 27,107) and 15,000 14,108 tertiary levels (mean: € 32,825; median: € 29,901; 10,000 see Figure 8). Working Swedish earn more than EU-28 and non-EU-28 citizens (mean: € 29,630 vs. 5,000 16,160; median: 27,782 vs. € 23,253 and € € 0 € 21,590 and € 14,108; see Figure 9). EU-28 Non EU-28 Sweden Figure 8: Mean and median equivalized net income by edu- cational attainment level 2018 (18 to 64 years; ISCED 2011) Mean equivalised net income Median equivalised net income

35,000 32,825 Source: Eurostat (2020o), own illustration 29,901 30,000 28,288 27,107 In 2019, compensation of employees (wages and 25,000 salaries plus employers’ social contributions) was 20,805 the largest income component of GDP in EU-28 19,485 20,000 and the Euro area, accounting for 47.8 % and

15,000 48.0 % in the euro area. Sweden (47.8 %) almost equals European average (both EU-28 and Euro 10,000 area). Taxes on production and imports (less sub- 5,000 sidies) accounted for 20.3 % (EU-28: 11.9 %; Euro area: 11.5 %). Gross operating surplus and mixed 0 income accounted for 31.9 % of GDP in Sweden, Primary Secondary Tertiary 40.3 % of GDP for the EU-28 and 40.5 % in the Euro

Mean equivalised net income area. Swedish GDP at current prices amounted to Median equivalised net income approx. € 474.6 billion in 2019 and GDP per inhab-

Source: Eurostat (2020n), own illustration itant equaled € 46,170 (Eurostat 2020q).

The indicator ‘people at risk of poverty or social Within 20 years Sweden’s gross value added (GVA) exclusion’ corresponds to the sum of persons who increased by 35.9 % (real growth rate). In terms of are: at risk of poverty after social transfers, se- verely materially deprived or living in households

4 The median is more robust, i.e. less sensitive to outliers than the mean. This explains why the mean equivalized income is always slightly higher for each educational level.

9

Table 2: Population aged 18 and over by risk of poverty, material deprivation, work intensity 2019 Risk of poverty At risk Not at risk At risk Not at risk At risk Not at risk Work Intensity Material Deprivation Sweden EU-28 Non EU-28 Very low Severe 0.2 0.0 0.0 0.0 6.8 0.3 Not very low Severe 0.2 0.3 1.2 1.0 1.6 2.0 Very low Non severe 2.2 1.2 3.5 0.5 19.7 2.0 Not very low Non severe 9.7 26.4 25.8 People at Risk of Poverty or Social Exclusion 13.8 32.6 58.2 Not at risk Not at risk Not at risk Not very low Non severe 86.2 67.5 41.8 People at no Risk of Poverty or Social Exclusion 86.2 67.5 41.8 Total 100.0 100.0 100.0 Source: Eurostat (2020p), own illustration

Table 3: Nominal regional GDP (2019) previous years, the northern and southern prov- and real growth rate (Δ 2002 to 2019) inces were all below the national level of € 46,200 Total Per Capita see Table 3).

Provinces with a high share of TCN have a high GDP per capita and vice versa. On the one hand,

2019) prosperous regions are more attractive to mi- grants; on the other hand, these regions benefit from a better availability of labor supply, i.e. Region greater number of working age people. In this con- GDP (in €) GDP (in text, one speaks of a so-called cross-fertilization, GDP (in €) €) per capita GDP (in

at current marketat current prices marketat current prices which results from the interaction of these two 2015=100; Δ 2002 to 2002 to Δ 2015=100;

of the EU27 (from 2020) average EU27 the 2020) of (from factors. added (GVA) at basic prices (Index, prices at basic (GVA) added Purchasing power standard (PPS, EU27 (PPS, standard power Purchasing Real growth rate of regional gross value gross value growth rate regional Real of from 2020), per inhabitant in percentage per inhabitant from 2020), Expressing GDP in PPS (purchasing power stand- Östra Sverige 219,636 43.4 53,800 138 Stockholm 152,809 49.2 64,700 166 ards) eliminates differences in price levels be- Östra tween countries or regions. There are significant 66,827 30.1 38,900 100 Mellansverige differences in levels of prosperity among Swedish Södra Sverige 185,268 33.5 41,700 107 provinces. The capital is the most prosperous re- Småland med 33,683 21.4 38,900 100 öarna gion (166 %); it is over 72/66/65 pp richer than Sydsverige 60,234 32.9 39,400 101 Norra Mellansverige/Småland med öarna and Västsverige 91,351 38.4 44,600 115 Östra Mellansverige/Sydsverige. Västsverige and Norra Sverige 69,464 18.4 39,600 102 Norra Övre Norrland (both 115 %) rank second (see Ta- 31,358 17.0 36,600 94 Mellansverige ble 3). Mellersta 14,834 18.2 39,500 101 Norrland Economic structure and entrepreneurship Övre 23,272 20.6 44,700 115 Norrland In general, a distinction is made between three dif- Sweden 474,468 35.9 46,200 119 ferent sectors of the economy, the primary (agri- Source: Eurostat (2020r), own illustration; culture and forestry), secondary (manufacturing) Eurostat (2020s), own illustration] and tertiary (services) sectors. regional GVA all provinces recorded positive real In total, approx. 5.1 million people were employed growth, ranging from 17.0 % in Norra Mellans- in 2019 in Sweden. More than three quarters verige to 49.2 % in Stockholm since 2000. The (77.8 %) are employed in the service sector and highest regional GDP per capita at current prices 19.0 % in the production sector; only 1.9 % work occurred also in Stockholm (approx. € 64,700) suc- in the primary sector (see Table 4). ceeded by Övre Norrland (approx. € 44,700). As in A more detailed examination of the different sec- tors of the economy reveals that “Human health

10

and social work activities” (16.1 %), “Wholesale (14.4 %), “Accommodation and food service activ- and retail trade; repair of motor vehicles and mo- ities” (12.6 %), “Construction” (11.1 %), “Trans- torcycles” (11.7 %), “Education” (10.9 %) and portation and storage” (10.1 %), “Manufacturing” “Manufacturing” (10.6 %) are the most important (9.3 %) and “Human health and social work activi- economic sections in regard to employment. ties” (9.0 %) whereas all remaining sections are of lower importance. Citizens from the rest of the Due to non-available data no exact analyses are world tend to work in the same sections whereas possible in regard to TCN. The analysis by region of their importance must be highlighted in “Accom- birth shows that 8.9 % belong to the group of Eu- modation and food service activities” (30.5 %), ropeans except domestic born population which “Administrative and support service activities” accounts for 80.3 %. 10.9 % are from the rest of (17.7 %), “Transportation and storage” (15.7 %) the world. and “Human health and social work activities” Europeans are overrepresented (above 8.9 %) in (14.8 %; see Table 4). “Administrative and support service activities”

Table 4: Employment by economic activity 2019 Employed Rest-of- Economic Activity persons in Share Swedish Europeans the-World 2019 Primary Sector 97,686 1.9% 93.5% 4.8% 1.7% Agriculture, forestry and fishing 97,686 1.9% 93.5% 4.8% 1.7% Secondary Sector 962,500 19.0% 84.1% 9.7% 6.2% Mining and quarrying 8,855 0.2% 95.1% 3.6% 1.3% Manufacturing 538,988 10.6% 82.9% 9.3% 7.8% Electricity, gas, steam and air conditioning supply 28,420 0.6% 92.1% 4.5% 3.4% Water supply; sewerage, waste management and reme- 23,334 0.5% 87.8% 7.1% 5.1% diation activities Construction 362,903 7.2% 84.7% 11.1% 4.2% Tertiary Sector 3,938,031 77.8% 79.1% 8.8% 12.1% Wholesale and retail trade; repair of 594,358 11.7% 84.0% 7.2% 8.7% motor vehicles and motorcycles Transportation and storage 241,112 4.8% 74.2% 10.1% 15.7% Accommodation and food service activities 182,575 3.6% 56.9% 12.6% 30.5% Information and communication 217,689 4.3% 81.7% 8.2% 10.1% Financial and insurance activities 96,074 1.9% 87.9% 6.4% 5.7% Real estate activities 86,306 1.7% 87.4% 6.7% 5.8% Professional, scientific and technical activities 329,189 6.5% 84.4% 8.2% 7.4% Administrative and support service activities 286,590 5.7% 67.9% 14.4% 17.7% Public administration and defence; 305,484 6.0% 88.5% 5.7% 5.8% compulsory social security Education 552,818 10.9% 79.8% 8.8% 11.4% Human health and social work activities 814,709 16.1% 76.1% 9.0% 14.8% Arts, entertainment and recreation 105,785 2.1% 86.8% 6.8% 6.4% Other service activities, Activities of households as em- ployers; undifferentiated goods- and sevices-producing 125,342 2.5% 77.7% 8.8% 13.5% activities of households for own use; Activities of extra- territorial organisations and bodies Others 63,337 1.3% 75.4% 8.3% 16.4% Total 5,061,554 100.0% 80.3% 8.9% 10.9% Source Statistics Sweden (2021), own calculations and illustrations

11

Research and Innovation In accordance to the research & development (R&D) funding structure almost three quarters Following the neoclassical and endogenous (70.9 % or 51,843 FTE) of the researchers work in growth theories, technological advance is believed the business enterprise sector. Approx. one quar- to be one of the major drivers of economic ter (23.9 % or 17,501 FTE) works in the higher ed- growth. From this perspective, there is a growing ucation sector and the remaining in the govern- interest to investigate the link between research & ment sector (5.2 % or 3,788 FTE) in the higher ed- development (R&D), innovation, entrepreneur- ucation sector. This makes a total of 73,132 FTE ship and economic growth achieved by human (Eurostat 2020v). capital from abroad (EU-28 and non-EU-28). Summary A well-known indicator provided to measure achievements of countries or regions in R&D is This Statistical Briefings examines the economic GERD, i.e. regional/national gross domestic ex- and spatial impact of immigration with special re- penditure on R&D as a percentage of GDP. GERD gard to third-country nationals (TCN) in Sweden. is estimated at 3.36 % for Sweden (2017) which is On January 1, 2020, a total of 940,580 persons a slight increase of 0.11 percentage points to 2016. with non-Swedish citizenship were living in Swe- The total amount of research expenditures was den. This corresponds to a share of 9.1 % of Swe- € 16.14 billion, the largest share (60.7 %) was fi- den's total population. Among non-Swedish na- nanced by the business enterprise sector. 25.0 % tionals, 34.3 % (3.1 % of total population) came were funded by the government sector, 10.1 % by from EU-28 (incl. UK) and EFTA countries (0.4 %). foreign countries (rest of the world), 3.3 % by pri- A total of 577,401 persons (or 5.60 % of total pop- vate non-profit sector and only 0.6 % by the higher ulation) were third country nationals (TCN), being education sector (Eurostat 2020t). individuals who are neither Swedish, EU-28 citi- Regional data is published on NUTS 1 and NUTS 2 zens nor EFTA citizens. By comparison, the share level. Västsverige is the province with the highest of TCN in EU-27 (excl. UK) and EU-28 (incl. UK) is ratio in 2017 (4.83 %) succeeded by Stockholm 5.7 % and 3.8 %, therefore Sweden is at EU-28 av- (3.75 %) and Östra Mellansverige (3.51 %). Except erage. for Sydsverige (3.23 %) all other provinces were Within the last 20 years the population increase of well behind and below the all-Swedish ratio of Sweden has strongly be dominated by immigra- 3.36 %. Norran Mellansverige ranks last (1.20 %) tion. From 2006 to 2019 especially the southern with Övre Norrland (2.30 %) only slightly better regions profited heavily with regard to its working- (see Table 5). age population. Without immigration, many re- Table 5: Intramural R&D expenditure as percentage of gions (especially in the north) would have suffered gross domestic product (GDP) 2017 from a decrease in its population aged between 20 Expenditures % of Province and 64 years. (€ mio.) GDP Manner-Suomi 6,432.3 2.76% Despite the relevance of migration for labor sup- Östra Sverige 8,005.8 3.67% Stockholm 5,566.0 3.75% ply, the employment rate of TCN is much lower Östra Mellansverige 2,439.7 3.51% compared to domestic population and EU-28 citi- Södra Sverige 7,064.9 3.71% zens. One relevant factor to be considered here is Småland med öarna n.a. n.a the lower educational attainment level of mi- Sydsverige 1,962.2 3.23% grants, with the consecutive effects of lower in- Västsverige 4,535.8 4.83% Norra Sverige 1,052.8 1.48% come levels and a higher risk of poverty. Norra Mellansverige 389.4 1.20% The pattern of education of TCN differs substan- Mellersta Norrland n.a. n.a. Övre Norrland 540.3 2.30% tially from that of Swedish nationals: The foreign Sweden 16,142.2 3.36% population of third countries is disproportionately Source: Eurostat (2020u), own illustration represented in primary education, whereas they are underrepresented in tertiary education levels.

12

From a spatial point of view, the concentration of As a result, governments need to invest in educa- better-educated migrants is highest in Swedish cit- tion on the one hand and in research and develop- ies. Provinces with a higher ratio of TCN are also ment (R&D) on the other, to enable long-term characterized by higher economic growth, being prosperity and growth. Currently the national more attractive for TCN on the one hand and of- gross domestic expenditures on R&D (as a per- fering more labor supply on the other hand. centage of GDP) are estimated at 3.36 % (2017) which is a slight increase of 0.11 percentage points The below average regional GVA growth rates, es- to 2016. But there is again a striking regional vari- pecially in northern and southern provinces is in- ation from densely populated areas with a high directly linked to immigration from abroad but number of innovative enterprises (e.g. Stockholm) preferably to internal migration from rural areas to less densely populated provinces (e.g. Norran to cities and metropolitan areas. These emigration Mellansverige or Övre Norrland). flows tend to reinforce the erosion of the eco- nomic basis, because they have a negative impact Summing up, immigrants in Sweden play a pivotal on the potential workforce and the economic at- role in reviving rural areas, and with the appropri- tractiveness of the region. Economic performance ate policies, they could play an even more signifi- is stagnating, accompanied by declining popula- cant role in sustaining them. tion figures.

13

Bibliography and source references

Becker, Gary Stanley (1964): Human Capital: A theoretical Eurostat (2020o): Mean and median income by broad group and empirical analysis, with special reference to education. of citizenship (population aged 18 and over), 3rd edition. New York. https://appsso.eurostat.ec.europa.eu/nui/show.do?da- taset=ilc_di15&lang=en Eurostat (2020a): Population on 1 January by age group, sex and citizenship, https://appsso.eurostat.ec.eu- Eurostat (2020p): Persons aged 18 and over by risk of pov- ropa.eu/nui/show.do?dataset=migr_pop1ctz&lang=en erty, material deprivation, work intensity of the household and most frequent activity status of the person - intersec- Eurostat (2020b): Population change - Demographic balance tions of Europe 2020 poverty target indicators, and crude rates at regional level (NUTS 3), https://appsso.eurostat.ec.europa.eu/nui/show.do?da- https://appsso.eurostat.ec.europa.eu/nui/show.do?da- taset=ilc_pees02&lang=en taset=demo_r_gind3&lang=en Eurostat (2020q): GDP and main components (output, ex- Eurostat (2020c): Population by sex, age, citizenship, labour penditure and income), https://appsso.eurostat.ec.eu- status and NUTS 2 regions, https://appsso.eurostat.ec.eu- ropa.eu/nui/show.do?dataset=nama_10_gdp&lang=en ropa.eu/nui/show.do?dataset=lfst_r_lfsd2pwn&lang=en Eurostat (2020r): Gross domestic product (GDP) at current Eurostat (2020d): Population on 1st January by age, sex, type market prices by NUTS 2 regions, https://appsso.euro- of projection and NUTS 3, https://appsso.eurostat.ec.eu- stat.ec.europa.eu/nui/show.do?da- ropa.eu/nui/show.do?dataset=proj_19rp3 taset=nama_10r_2gdp&lang=en Eurostat (2020e): City, Eurostat (2020s), Real growth rate of regional gross value https://ec.europa.eu/eurostat/statistics- added (GVA) at basic prices by NUTS 2 regions - percentage explained/index.php?title=Glossary:City change on previous year, https://appsso.eurostat.ec.eu- Eurostat (2020f): Town or suburb, https://ec.europa.eu/eu- ropa.eu/nui/show.do?dataset=nama_10r_2gvagr&lang=en rostat/statistics-explained/index.php?title=Glos- Eurostat (2020t): GERD by sector of performance and source sary:Town_or_suburb of funds, http://appsso.eurostat.ec.eu- Eurostat (2020g): Urban centre, ropa.eu/nui/show.do?dataset=rd_e_gerdfund&lang https://ec.europa.eu/eurostat/statistics- Eurostat (2020u): Intramural R&D expenditure (GERD) by explained/index.php?title=Glossary:Urban_centre NUTS 2 regions, https://ec.europa.eu/eurostat/data- Eurostat (2020h): Rural area, https://ec.europa.eu/euro- browser/view/tgs00042/default/table?lang=en stat/statistics-explained/index.php/Glossary:Rural_area Eurostat (2020v): Total R&D personnel and researchers by Eurostat (2020i): Rural grid cell, https://ec.europa.eu/euro- sectors of performance, educational attainment level stat/statistics-explained/index.php?title=Glossary:Ru- (ISCED2011) and sex, https://ec.europa.eu/eurostat/data- ral_grid_cell browser/view/rd_p_persqual11/default/table?lang=en

Eurostat (2020j): Population by educational attainment level, Statistics Sweden (2021): Gainfully employed 16+ years by sex, age, citizenship and degree of urbanization, region of residence (RAMS) by sex, year, industrial classifica- http://appsso.eurostat.ec.europa.eu/nui/show.do?da- tion NACE Rev. 2 and region of birth, https://www.statis- taset=edat_lfs_9916&lang=en tikdata- basen.scb.se/pxweb/en/ssd/START__AM__AM0207__AM02 Eurostat (2020k): Unemployment rates by sex, age, educa- 07Z/NattSni07FoRN/table/tableViewLayout1/ tional attainment level and NUTS 2 regions (%), https://appsso.eurostat.ec.europa.eu/nui/show.do?da- taset=lfst_r_lfu3rt&lang=en

Eurostat (2020l): Employment rates by sex, age, educational attainment level, citizenship and NUTS 2 regions, https://ec.europa.eu/eurostat/data- browser/view/tepsr_wc140/default/table?lang=en

Eurostat (2020m): Early leavers from education and training by sex and citizenship, http://appsso.eurostat.ec.eu- ropa.eu/nui/show.do?dataset=edat_lfse_01&lang=en

Eurostat (2020n): Mean and median income by educational attainment level - EU-SILC survey, http://appsso.euro- stat.ec.europa.eu/nui/show.do?lang=en&dataset=ilc_di08

14

ECONOMIC IMPACT OF MIGRATION

Statistical Briefings Turkey

Call: H2020-SC6-MIGRATION-2019

Work Programmes:

H2020-EU.3.6.1.1. The mechanisms to promote smart, sustainable and inclusive growth H2020-EU.3.6.1.2. Trusted organisations, practices, services and policies that are necessary to build resilient, inclusive, participatory, open and creative societies in Europe, in particular taking into account migration, integration and demographic change

Deliverable 4.2 – Statistical briefing for Turkey on economic impact of TCNs in MATILDE re- gions

Authors: Birgit Aigner-Walder, Albert Luger, Rahel M. Schomaker with contributions from Koray Akay and Kubra Dogan Yenisey

Approved by Work Package Manager of WP4: Simone Baglioni, UNIPR Approved by Scientific Head: Andrea Membretti, UEF Approved by Project Coordinator: Jussi Laine, UEF

Version: 28.05.2021

DOI: 10.5281/zenodo.4817376

This document was produced under the terms and conditions of Grant Agreement No. 870831 for the European Commission. It does not necessary reflect the view of the European Union and in no way anticipates the Commission’s future policy in this area.

2

Introduction

This document presents the results of an assessment of the impact of migration and a measure- ment of the contribution provided by ‘third country nationals’ (TCNs) to the economic systems in the receiving contexts. In total, 10 statistical briefings for the MATILDE regions have been compiled, including the following countries: Austria, Bulgaria, Finland, Germany, Italy, Norway, Spain, Sweden, Turkey and the United Kingdom. Each document focuses on the following dimen- sions of economic development: economic growth, labour markets, innovation, and entrepre- neurship. Moreover, a short introduction on the relevance of TCNs in the respective country as well as the population development and structure are considered. The results of a literature review and a comparative analysis of the 10 statistical briefings are published separately.

Within the Matilde project there is a special focus on so-called ‘third country nationals’ (TCNs). MATILDE considers TCNs as Non EU citizens, who reside legally in the European Union (EU) and who are the target of EU integration policies. A TCN is “any person who is not a citizen of the European Union within the meaning of Art. 20(1) of TFEU and who is not a person enjoying the European Union right to free movement, as defined in Art. 2(5) of the Regulation (EU) 2016/399 (Schengen Borders Code).“ (European Commission 2020a). According to this definition, nationals of European Free Trade Association (EFTA) member countries Norway, Island, Liechtenstein and Switzerland are not considered to be TCNs. TCNs cover economic migrants, family migrants, stu- dents and researchers, highly skilled migrants, and forced migrants.

3

TURKEY Figure 1: Origin and share of total population of different nationalities in Turkey

Source: Eurostat (2020a), own illustration Third Country Nationals (TCN) in Turkey 3.8 %, therefore Turkey lying way below the aver- age. On January 1, 2020, a total of 1,528,271 persons with non-Turkish citizenship1 were living in Turkey. Considering the nationalities of TCN, Iraqis This corresponded to a share of 1.8 % of Turkey's (0.38 %), Afghanis (0.18 %), Turkmens (0.16 %), total population. Among non-Turkish nationals, Syrians (0.14 %) and Iranians (0.11 %) are the big- only 11.3 % (172,989 persons; 0.2 % of total popu- gest groups by citizenship. In total, more than half lation) came from EU-28 (incl. UK) and EFTA coun- (59.7 %) of TCN living in Turkey belong to these na- tries (3,459 persons; 0.004 %). A total of tionalities. Figure 1 gives an overview of the origin 1,351,823 persons (or 1.6 % of total population) and the share of the different nationalities in Tur- were third country nationals (TCN), being individ- key. uals who are neither Turks, EU-28 citizens nor EFTA citizens. By comparison, the share of TCN in EU-27 (excl. UK) and EU-28 (incl. UK) is 5.7 % and

1 Foreign population covers individuals who are holding a his/her Turkish Republic citizenship and who have a valid ad- valid residence/work permit at the reference day and indi- dress declaration at the reference day. Foreigners holding vi- viduals who have a valid address declaration at the refer- sas or residence permits shorter than 3 months with the pur- ence day while holding an identity document equivalent to pose of training, tourism, scientific research, etc. are not residence permit such as international protection identity covered. Also Syrians (3,671,811 as of March 21, 2021) un- document and the individuals who have already renounced der temporary protection are not included in the figures (wee Turkish Statistical Institute 2020a).

4

Population Development & Population Structure analysis of population development since the turn of the new millennium reveals differences, though The number of inhabitants in Turkey has increased the number of inhabitants increased in 86.4 % of in the past and this development is projected to all NUTS 3 regions. Yalova (+85.28 %), Tekirdag continue in the future. As of January 1, 2020, there (+81.39 %), Antalya (+75.55 %), Sanliurfa were 83,154,997 persons registered in Turkey. (+ 67.12 %) and Gaziantep (+62.03 %) experi- This corresponds to an increase of 16,265,572 cit- enced the highest population growth in the past. izens (+24.3 %) within the last 20 years since 2000. Eleven regions shrunk since 2001. Relative reduc- The main cause of the population increase are tions were highest for Yozgat (-22.0 %), Ardahan (- high fertility rates before the 1990s. A long-term 19.1 %) and Kars (-11.3 %; see Figure 2).

Figure 2: Population development (Δ 2001 to 2020), NUTS 3

Source: Eurostat (2020b), own illustration

While most discussions about changing demo- labor force potential are recorded in Istanbul graphic structures focus on population ageing and (+5.9 %), Ankara (+2.3 %), Kocaeli, Sakarya, Düzce, the sustainability of pension systems, urbanization Bolu, Yalova (+1.7 %), Bursa, Eskisehir, Bilecik etc., one relevant aspect is labor supply, i.e. the (+1.5 %), Antalya, Isparta, Burdur (+1.5 %), San- population aged 15 to 64 years or 20 to 64 years in liurfa, Diyarbakir (+1.5 %), Izmir (+1.3 %) and Gazi- industrialized economies and nations. From an antep, Adiyaman, Kilis (+1.1 %). In all other regions economic perspective, labor (done by human be- the labor force potential stagnated, i.e. range be- ings) is an essential element in the production of tween +0.1 and 0.9 %. goods and services. The term ‘labor force’ com- As Figure 4 (a)-(b) indicate, the increase in the prises all of those who work for gain, whether as population from 2016 to 2019 within the regions employees, employers, or as self-employed, and it is mainly caused by an increase in the Turkish pop- includes the unemployed who are seeking for ulation (78.6 %) whereas foreigners account for work. only 21.4 % of population growth. Summing up, al- As indicates, there are regional disparities among most all areas are experiencing an overall growth the NUTS 2 regions in Turkey in regard to the in- in population. crease of working-age population though it in- creased in all of them. The highest growth rates in

5

Figure 3: Working-age population 20 to 64 years; Δ 2006 to 2019)

Source: Eurostat (2020c), own illustration

Figure 4: Population development by citizenship (Δ 2016 to 2019) (a)

(b)

Source: Turkish Statistical Institute (2020b), own illustration

6

Education and (Un)Employment Hakkari (25.9 %; female: 27.1 %; male: 25.4 %) and Sanliurfa, Diyarbakir (23.6 %; female: 18.6 %; The unemployment rate in Turkey is 13.5 %, being male: 25.3 %). Kastamonu, Çankiri, Sinop (7.7 %; below the EU-28 average of 6.2 %. From a regional 10.5 % female; 6.0 % male) and Konya, Karaman perspective, the southeastern region Mardin, Bat- (7.9 %; 9.3 % female; 7.3 % male) are the prov- man, Sirnak, Siirt has by far the highest unemploy- inces with the lowest unemployment rates (see ment rate at 29.9 % (female: 42.1 %; Figure 5). male: 25.7 %), followed by Van, Mus, Bitlis,

Figure 5: Unemployment rate 2019 (20 to 64 years)

Source: Eurostat (2020d), own illustration

The tertiary education rate of the total population observed at the European level, as the risk of un- living in Turkey is 18.4 %. Compared to the EU-28 employment decreases the higher the educational (29.5 %) Turkey has a far below-average tertiary level is (EU-28; Figure 6). education rate. If Turkey was EU-28 member it Figure 6: Unemployment rate by educational attainment th would approx. ranks 26 , ahead of Italy (17.4 %) level 2019 (20 to 64 years; ISCED 2011) and Romania (16.0 %). Luxembourg (41.0 %), Ire- 16 land (40.7 %) and United Kingdom (40.6 %) rank at 14.6 the top end, while the Czech Republic (21.6 %), It- 14 13.1 13.5 aly (17.4 %) and Romania (16.0 %) rank at the bot- 12.4 12 tom. 10 In general, there is a correlation between the level 8 of education and unemployment, according to which the well-educated are significantly less 6 5.6 4.0 likely to be unemployed than people without a vo- 4 cational qualification. This correlation does not 2 hold for Turkey but for EU-28: In Turkey, those 0 with secondary education are most affected by Primary Secondary Tertiary unemployment (14.6 %). Conversely, people with primary education have the lowest unemploy- EU-28 Turkey ment rate (13.1 %), and those with tertiary

(13.5 %) rank in the midfield. Contrary results are Source: Eurostat (2020d), own illustration

7

The abovementioned findings foster, at least for a lack of data no statements about the causality EU-28, the theoretical conclusions by Becker can be made as this is quite surprising in regard to (1964) that the educational level has a significant the comparable high income of foreigners in Tur- influence on the professional career and the risk key. of unemployment of human beings. Figure 7: Mean and median equivalized net income by edu- This theory and the empirical evidence can also be cational attainment level 2019 (18 to 64 years; ISCED 2011) applied to the “Early Leavers from Education and 8,000 7,566 Training”. This population group refers to persons 7,000 aged 18 to 24 who have completed a lower sec- ondary education and are not involved in further 6,000 5,671 education or training. From a fiscal point of view, 5,000 4,476 this specific group is of particular relevance as they 4,000 3,352 3,586 are more likely to be (long-term) unemployed. In 3,000 2,780 2019, 10.3 % of the 18-24 year olds in the EU-28 were part of this group (male: 11.9 %; female: 2,000 8.6 %). Turkey ranges with a value of 28.7 % way 1,000 above the European average. In total, more than a 0 half (55.1 %) of all early leavers are jobless but no Primary Secondary Tertiary relying statements can be made in regard to immi- grational status because of missing data (Eurostat Mean equivalised net income Median equivalised net income (2020e). Source: Eurostat (2020f), own illustration

Income and Gross Domestic Product Figure 8: Mean and median equivalized net income by The educational attainment level has a positive broad group of citizenship 2019 (18 to 64 years) 2 impact on mean and median equivalized net in- 7,000 6,356 6,374 come, i.e. the higher the educational attainment level the higher the income. Working persons in 6,000 Turkey at primary level earn less (mean: € 3,352; 5,000 4,414 median: € 2,780) than secondary (€ 4,476 or 4,000 3,504 3,451 € 3,586) and tertiary levels (€ 7,566 or € 5,761; see 3,285 Figure 7). Working Turkish earn less than foreign- 3,000 ers and non-EU-28 citizens which are a subgroup 2,000 of foreigners in Turkey (mean: € 4,414 vs. € 6,356 1,000 and € 6,374; median: € 3,285 vs. € 3,504 and € 3,451; see Figure 8). 0 Foreign country Non EU-28 Turkey The indicator ‘people at risk of poverty or social exclusion’ corresponds to the sum of persons who Mean equivalised net income are: at risk of poverty after social transfers, se- Median equivalised net income verely materially deprived or living in households Source: Eurostat (2020g), own illustration with very low work intensity. Almost one half of A way to calculate gross domestic product (GDP) foreigners (43.1 %) as well as the subgroup of non- is, to sum up, all the income earned by factors of EU-28 citizens (43.0 %) belong to this group production from firms in the economy—the wages whereas the values for the domestic population earned by labor; the interest paid to those who (36.4 %) are comparable low (see Table 1). Due to lend their savings to firms and the government;

2 The median is more robust, i.e. less sensitive to outliers than the mean. This explains why the mean equivalized income is always slightly higher for each educational level.

8

the rent earned by those who lease their land or must go somewhere; whatever is not paid as structures to firms; and dividends, the profits paid wages, interest or rent is profit. Ultimately, profits to the shareholders, the owners of the firms’ phys- are paid out to shareholders as dividends ical capital. This is a valid measure because the (Krugman and Wells 2017). money firms earn by selling goods and services

Table 1: Population aged 18 and over by risk of poverty, material deprivation, work intensity 2019 Risk of poverty At risk Not at risk At risk Not at risk At risk Not at risk Work Intensity Material Deprivation Turkey Foreign country Non EU-28 Very low Severe 1.9 1.4 7.0 2.1 7.4 2.2 Not very low Severe 8.2 12.4 12.1 9.7 12.9 9.1 Very low Non severe 1.1 4.8 0.1 8.4 0.1 8.9 Not very low Non severe 6.6 3.7 2.4 People at Risk of Poverty or Social Exclusion 36.4 43.1 43, 0 Not at risk Not at risk Not at risk Not very low Non severe 63.5 56.9 56.9 People at no Risk of Poverty or Social Exclusion 63.5 56.9 56 .9 Total 100.0 100.0 100.0 Source: Eurostat (2020h), own illustration

In 2019, compensation of employees (wages and Table 2: Nominal regional GDP (2019) salaries plus employers’ social contributions) was and real growth rate (Δ 2004 to 2019) Per Total the largest income component of EU-28 GDP ac- Capita counting for 47.8 % and 48.0 % in the euro area. EU candidate country Turkey (31.3 %) lies below European average (both EU-28 and Euro area).

Taxes on production and imports (less subsidies) accounted for 9.0 % (EU-28: 11.9 %; Euro 4 to 2019)4 to

area: 11.5 %). Gross operating surplus and mixed Region income accounted for 59.6 % of GDP in Turkey, GDP (in €) GDP (in 40.3 % of GDP for the EU-28 and 40.5 % in the Euro GDP (in €) €) per capita GDP (in

area. Turkish GDP at current prices amounted to marketat current prices marketat current prices 2015=100; Δ 200 Δ 2015=100; approx. € 679.5 billion in 2019 and GDP per inhab- added (GVA) at basic prices (Index, prices at basic (GVA) added

itant equaled € 8,230 (Eurostat 2020i). gross value growth rate regional Real of Istanbul 208,791 63.7 13,700 Within 15 years Turkey’s gross value added (GVA) Bati Marmara 30,838 59.5 8,600 increased by 59.9 % (real growth rate). In terms of Tekirdag, Edirne, 17,713 63.3 9,700 regional GVA all provinces also recorded positive Kirklareli real growth, ranging from 34.1 % in Zonguldak, Balikesir, Çanakkale 13,125 54.5 7,400 Karabük, Bartin to 77.3 % in Mardin, Batman, Sir- Ege 85,817 57.2 8,100 nak, Siirt since 2004. The highest regional GDP per Izmir 41,372 57.0 9,500 Aydin, Denizli, capita at current prices occurred also in Istanbul 22,799 56.9 7,300 Mugla 13,700) succeeded by Ankara (approx. (approx. € Manisa, € 11,200) and Kocaeli, Sakarya, Düzce, Bolu, Afyonkarahisar, 21,645 57.9 7,000 Yalova (approx. € 10,300). As in previous years, all Kütahya, Usak regions east of Bati Anadolu (approx. € 9,700) Dogu Marmara 78,037 60.1 9,700 Bursa, Eskisehir, were below the national level of € 8,200 (see Table 37,733 58.1 9,100 2). Bilecik Kocaeli, Sakarya, 40,304 62.0 10,300 Düzce, Bolu, Yalova

9

Bati Anadolu 78,383 62.4 9,700 Economic structure and entrepreneurship Ankara 62,243 62.1 11,200 Konya, Karaman 16,140 63.6 6,500 In general, a distinction is made between three dif- Akdeniz 70,418 58.4 6,700 ferent sectors of the economy, the primary (agri- Antalya, Isparta, culture and forestry), secondary (manufacturing) 28,218 64.8 8,900 Burdur and tertiary (services) sectors. Adana, Mersin 25,379 57.7 6,300 Hatay, 145,232 work permits were granted to foreigners Kahramanmaras, 16,820 49.9 5,100 in Turkey in 2019. Almost all (94,68 %) permissions Osmaniye were given to TCNs, especially to workers from Orta Anadolu 24,425 52.0 6,000 Syria (43.92 % of total permissions granted), Kyr- Kirikkale, Aksaray, Nigde, Nevsehir, 9,252 55.5 5,800 gyzstan (7.58 %), Ukraine (4.27 %), Turkmenistan Kirsehir (4.22 %), Georgia (3.59 %) and Uzbekistan (3.1 %). Kayseri, Sivas, In total, these countries account for exactly two 15,173 49.9 6,200 Yozgat thirds (66.66 %) of work permission in 2019. Bati Karadeniz 25,086 43.2 5,400 3.55 % of all permits are given to EU-28 citizens Zonguldak, 5,968 34.1 5,700 and there is no further information available for Karabük, Bartin Kastamonu, 1.77 %. 4,630 48.1 5,700 Çankiri, Sinop In total, approx. 26.8 million people were em- Samsun, Tokat, 14,488 46.0 5,100 Çorum, Amasya ployed in . This corresponds to an Dogu Karadeniz 14,936 52.3 5,500 increase of +3.4 % compared to 2014. More than Kuzeydogu Anadolu 9,931 53.3 4,500 one half (56.2 %) are employed in the service sec- Erzurum, Erzincan, tor (15.1 million persons), where employment has 5,835 53.1 5,400 Bayburt increased by +13.8 % since 2014 while it de- Agri, Kars, Igdir, 4,096 53.7 3,700 creased in primary (-13.8 %) and secondary sector Ardahan (-2.6 %). Ortadogu Anadolu 16,073 62.6 4,100 Malatya, Elazig, 8,685 57.8 4,900 A more detailed examination of the different sec- Bingöl, Tunceli tors of the economy reveals that there have also Van, Mus, Bitlis, 7,388 68.5 3,400 Hakkari been significant shifts within the sectors: the only Güneydogu Anadolu 36,775 64.7 4,100 declines in employment were recorded in the sec- Gaziantep, tors Administrative and support service activi- 14,990 67.9 5,300 “ Adiyaman, Kilis ties" (-24.4 %), “Construction” (-19.6 %), “Agricul- Sanliurfa, 12,132 52.6 3,200 ture, forestry and fishing” (-13.8 %) and “Arts, en- Diyarbakir -5.3 %). "Human Mardin, Batman, tertainment and recreation” ( 9,654 77.3 4,200 Sirnak, Siirt health and social work activities" (+50.9 %), “Pub- Turkey 679,510 59.9 8,200 lic administration and defence” (+42.0 %), “Real Source: Eurostat (2020j), own illustration; estate activities” (42.0 %), “Education" (+36.2 %), Eurostat (2020k), own illustration “Professional, scientific and technical activi- Provinces with a high share of foreigners have a ties” (+28.4 %) as well as "Electricity, gas, steam, high GDP per capita and vice versa. On the one water supply, sewerage etc." (+20.0 %), on the hand, prosperous regions are more attractive to other hand, are among those economic sections migrants; on the other hand, these regions benefit with the largest growth rates. from a better availability of labor supply, i.e. Further analysis focusing on work permissions greater number of working age people. In this con- granted to foreigners in 2019 shows that more text, one speaks of a so-called cross-fertilization, than one half (56.3 %) work in the tertiary and ap- which results from the interaction of these two prox. one quarter (27.3 %) in the secondary sector factors. whereas foreigners are of no relevance in the pri-

10

mary sector (0.3 %). Those with permission prefer- “Wholesale and retail trade” (13.1 %)) which ac- ably work in “Manufacturing” (23.4 %), “Accomo- count for more than 50.0 % (56.0 %; see Table 3). dation and food service activities” (19.5 %)” and

Table 3: Employment by economic activity 2019 Employed Share Work Economic Activity persons Share Δ 14-20 permissions (in 1,000) 2019 Primary Sector 4,716 17.6% -13.8% 0.3% Agriculture, forestry and fishing 4,716 17.6% -13.8% 0.3% Secondary Sector 7,036 26.2% -2.6% 27.3% Mining and quarrying 134 0.5% 0.0% 0.6% Manufacturing 5,070 18.9% 2.7% 23.4% Electricity, gas, steam, water supply, sewerage etc. 294 1.1% 20.0% 0.2% Construction 1,538 5.7% -19.6% 3.1% Tertiary Sector 15,061 56.2% 13.8% 56.3% Wholesale and retail trade 3,723 13.9% 3.8% 13.1% Transportation and storage 1,209 4.5% 8.0% 1.8% Accommodation and food service activities 1,377 5.1% 1.9% 19.5% Information and communication 241 0.9% 6.2% 1.2% Financial and insurance activities 315 1.2% 4.7% 0.3% Real estate activities 291 1.1% 42.0% 1.1% Professional, scientific and technical activities 878 3.3% 28.4% 2.1% Administrative and support service activities 873 3.3% -24.4% 6.6% Public administration and defence 1,965 7.3% 42.0% 0.0% Education 1,798 6.7% 36.2% 3.0% Human health and social work activities 1,467 5.5% 50.9% 3.5% Arts, entertainment and recreation 124 0.5% -5.3% 1.0% Other social, community and personal service activities 800 3.0% 0.0% 3.1% Total 26,812 100.0% 3.4% 100.0% Source: Turkish Statistical Institute (2020c), own calculations and illustration; Turkish Statistical Institute (2020d), own calculations and illustration

Research and Innovation amount of research expenditures was € 7,23 bil- lion, the largest share (56,6 % or € 4.07 billion) was Following the neoclassical and endogenous financed by the business enterprise sector. 29.2 % growth theories, technological advance is believed (or € 2.12 billion) were financed by the govern- to be one of the major drivers of economic ment sector. 13.2 % (or € 923,5 millions) were fi- growth. From this perspective, there is a growing nanced by higher education sector and 1.9 % (or € interest to investigate the link between research & 109.0 millions) were funded by foreign countries development (R&D), innovation, entrepreneur- (rest of the world; Eurostat 2020l) ship and economic growth achieved by human capital from abroad (EU-28 and non-EU-28). From a regional perspective (NUTS 2) Ankara (31.6 %) and Istanbul (26.4 %) account for almost A well-known indicator provided to measure 60 % (58.0 %) of total research spending. The ratio achievements of countries or regions in R&D is in the remaining provinces is 9.5 % (Kocaeli, Sa- GERD, i.e. regional/national gross domestic ex- karya, Düzce, Bolu, Yalova), 6.4 % (Bursa, penditure on R&D as a percentage of GDP. A well- Eskişehir, Bilecik) and 5.2 % (İzmir; see Table 4). known indicator provided to measure achieve- ments of countries or regions in R&D is GERD, i.e. regional/national gross domestic expenditure on R&D as a percentage of GDP. GERD is estimated at 1.06 % for Turkey (2019) which is a slight increase of 0.11 percentage points to 2017. The total

11

Table 4: Intramural R&D expenditure as percentage of In accordance to the research & development gross domestic product (GDP) 2017 (R&D) funding structure more than half (55.7 % or Expenditures % of 62,305 full-time equivalent (FTE)) of the research- Province (Share GDP of Total) ers work in the business enterprise sector. Approx. Ankara 31.6 % 0.34% one third (38.4 % or 42.916 FTE) works in the Istanbul 26.4 % 0.28% higher education sector and the rest in the govern- Kocaeli, Sakarya, Düzce, 9.5 % 0.11% ment sector (6.0 % or 6,672 FTE). This makes a to- Bolu, Yalova Bursa, Eskişehir, Bilecik 6.4 % 0.07% tal of 111,893 FTE. Further analysis on educational İzmir 5.2 % 0.06% attainment level reveals that most researchers Diğer (Other) 20.8 % 0.22% (75.1 %) have tertiary education but not doctoral Turkey 100.0% 1.06% or equivalent level; 23.7 % belong to the latter Source: Eurostat (2020m), own calculations and illustration; Turkish Statistical Institute (2020e); own calculations and il- whereas 1.2 % have less than primary, primary, lustration secondary and post-secondary non-tertiary edu- cation (see Figure 9).

Figure 9: R&D researchers (full-time equivalent (FTE)) by sectors of performance and educational attainment level 2017 (ISCED2011) 70,000

60,000 58,107

50,000

40,000

30,000 22,286 20,631 20,000

10,000 2,940 5,282 1,258 1,334 0 56 Business enterprise sector Government sector Higher education sector

Less than primary, primary, secondary and post-secondary non-tertiary education (levels 0-4) Tertiary education excluding doctoral or equivalent level (levels 5-7) Doctoral or equivalent level

Source: Eurostat (2020n), own illustration

12

Conclusion

This Statistical Briefings examines the economic of TCN in EU-27 (excl. UK) and EU-28 (incl. UK) is and spatial importance and impact of immigration 5.7 % and 3.8 %, therefore Turkey lying way below with special regard to third-country nationals the average. (TCN) in Turkey. Within the last 20 years the population growth On January 1, 2020, a total of 1,528,271 persons was dominated by high fertility rates of Turkish cit- with non-Turkish citizenship were living in Turkey. izens, which makes Turkey a young nation (espe- This corresponds to a share of 1.8 % of Turkey's to- cially with regard to the developments in industri- tal population. Among non-Turkish nationals, only alized nations). From 2006 to 2019 all NUTS 2 re- 11.3 % came from EU-28 (incl. UK) and EFTA coun- gions in Turkey recorded a growth in working-age tries. A total of 1,351,823 persons (or 1.6 % of to- population though the development stagnates in tal population) were third country nationals (TCN), most of the regions. As the share of immigrants is being individuals who are neither Turks, EU-28 cit- quite low, one can assume that their impact was izens nor EFTA citizens. By comparison, the share of low importance in the past.

13

Bibliography and source references

Becker, Gary Stanley (1964): Human Capital: A theoretical Eurostat (2020n): Total R&D personnel and researchers by and empirical analysis, with special reference to education. sectors of performance, educational attainment level 3rd edition. New York. (ISCED2011) and sex, https://ec.europa.eu/eurostat/data- browser/view/rd_p_persqual11/default/table?lang=en Eurostat (2020a): Population on 1 January by age group, sex and citizenship, https://appsso.eurostat.ec.eu- Krugman, Paul, Wells, Robin (2017): Economics, Fifth Edition, ropa.eu/nui/show.do?dataset=migr_pop1ctz&lang=en New York.

Eurostat (2020b): Population change - Demographic balance Turkish Statistical Institute (2020a), Foreign population by and crude rates at regional level (NUTS 3), sex and the first year of residence in Turkey 2019, https://appsso.eurostat.ec.europa.eu/nui/show.do?da- https://data.tuik.gov.tr/Kategori/GetKate- taset=demo_r_gind3&lang=en gori?p=nufus-ve-demografi-109&dil=2

Eurostat (2020c): Population by sex, age, citizenship, labour Turkish Statistical Institute (2020b): Immigrants and emi- status and NUTS 2 regions, https://appsso.eurostat.ec.eu- grants by citizenship and provinces, 2016-2019, ropa.eu/nui/show.do?dataset=lfst_r_lfsd2pwn&lang=en https://data.tuik.gov.tr/Bulten/Index?p=Uluslararasi-Goc- Istatistikleri-2019-33709 Eurostat (2020d): Unemployment rates by sex, age, educa- tional attainment level and NUTS 2 regions (%), Turkish Statistical Institute (2020c): Economic Activity of Em- https://appsso.eurostat.ec.europa.eu/nui/show.do?da- ployment, Nace Rev.2, https://data.tuik.gov.tr/a256ab2e- taset=lfst_r_lfu3rt&lang=en cb6d- 482c-ad7c-b90f6ceb8807

Eurostat (2020e): Early leavers from education and training Turkish Statistical Institute (2020d): Work permits of foreign- by sex and citizenship, http://appsso.eurostat.ec.eu- ers, https://www.csgb.gov.tr/media/63117/yabancii- ropa.eu/nui/show.do?dataset=edat_lfse_01&lang=en zin2019.pdf

Eurostat (2020f): Mean and median income by educational Turkish Statistical Institute (2020e): Distribution of R&D ex- attainment level - EU-SILC survey, http://appsso.euro- penditures according to NUTS 2 level 2019, stat.ec.europa.eu/nui/show.do?lang=en&dataset=ilc_di08 https://tuikweb.tuik.gov.tr/PreHaberBultenleri.do?id=33676

Eurostat (2020g): Mean and median income by broad group of citizenship (population aged 18 and over), https://appsso.eurostat.ec.europa.eu/nui/show.do?da- taset=ilc_di15&lang=en

Eurostat (2020h): Persons aged 18 and over by risk of pov- erty, material deprivation, work intensity of the household and most frequent activity status of the person - intersec- tions of Europe 2020 poverty target indicators, https://appsso.eurostat.ec.europa.eu/nui/show.do?da- taset=ilc_pees02&lang=en

Eurostat (2020i): GDP and main components (output, ex- penditure and income), https://appsso.eurostat.ec.eu- ropa.eu/nui/show.do?dataset=nama_10_gdp&lang=en

Eurostat (2020j): Gross domestic product (GDP) at current market prices by NUTS 2 regions, https://appsso.euro- stat.ec.europa.eu/nui/show.do?da- taset=nama_10r_2gdp&lang=en

Eurostat (2020k): Real growth rate of regional gross value added (GVA) at basic prices by NUTS 2 regions - percentage change on previous year, https://appsso.eurostat.ec.eu- ropa.eu/nui/show.do?dataset=nama_10r_2gvagr&lang=en

Eurostat (2020l): GERD by sector of performance and source of funds, http://appsso.eurostat.ec.eu- ropa.eu/nui/show.do?dataset=rd_e_gerdfund&lang

Eurostat (2020m): Intramural R&D expenditure (GERD) by NUTS 2 regions, https://ec.europa.eu/eurostat/data- browser/view/tgs00042/default/table?lang=en

14

ECONOMIC IMPACT OF MIGRATION

Statistical Briefings United Kingdom

Call: H2020-SC6-MIGRATION-2019

Work Programmes:

H2020-EU.3.6.1.1. The mechanisms to promote smart, sustainable and inclusive growth H2020-EU.3.6.1.2. Trusted organisations, practices, services and policies that are necessary to build resilient, inclusive, participatory, open and creative societies in Europe, in particular taking into account migration, integration and demographic change

Deliverable 4.2 – Statistical briefing for United Kingdom on economic impact of TCNs in MA- TILDE regions

Authors: Birgit Aigner-Walder, Albert Luger, Rahel M. Schomaker with contributions from Maria Luisa Caputo, Michele Bianchi and Simone Baglioni

Approved by Work Package Manager of WP4: Simone Baglioni, UNIPR Approved by Scientific Head: Andrea Membretti, UEF Approved by Project Coordinator: Jussi Laine, UEF

Version: 28.05.2021

DOI: 10.5281/zenodo.4817376

This document was produced under the terms and conditions of Grant Agreement No. 870831 for the European Commission. It does not necessary reflect the view of the European Union and in no way anticipates the Commission’s future policy in this area.

2

Introduction

This document presents the results of an assessment of the impact of migration and a measure- ment of the contribution provided by ‘third country nationals’ (TCNs) to the economic systems in the receiving contexts. In total, 10 statistical briefings for the MATILDE regions have been compiled, including the following countries: Austria, Bulgaria, Finland, Germany, Italy, Norway, Spain, Sweden, Turkey and the United Kingdom. Each document focuses on the following dimen- sions of economic development: economic growth, labour markets, innovation, and entrepre- neurship. Moreover, a short introduction on the relevance of TCNs in the respective country as well as the population development and structure are considered. The results of a literature review and a comparative analysis of the 10 statistical briefings are published separately.

Within the Matilde project there is a special focus on so-called ‘third country nationals’ (TCNs). MATILDE considers TCNs as Non EU citizens, who reside legally in the European Union (EU) and who are the target of EU integration policies. A TCN is “any person who is not a citizen of the European Union within the meaning of Art. 20(1) of TFEU and who is not a person enjoying the European Union right to free movement, as defined in Art. 2(5) of the Regulation (EU) 2016/399 (Schengen Borders Code).“ (European Commission 2020a). According to this definition, nationals of European Free Trade Association (EFTA) member countries Norway, Island, Liechtenstein and Switzerland are not considered to be TCNs. TCNs cover economic migrants, family migrants, stu- dents and researchers, highly skilled migrants, and forced migrants.

3

UNITED KINGDOM (UK) Figure 1: Origin and share of total population of different nationalities in United Kingdom

Source: Eurostat (2020a), own illustration Third Country Nationals (TCN) in the UK

On January 1, 2019, a total of 6,200,181 persons Data about third country nationalities are not with non-British citizenship were living in the made easily available. According to our findings In- United Kingdom (UK), former EU-28 member state dians (0.54 %), Pakistanis (0.29 %), Chinese (incl. (until 2020). This corresponds to a share of 9.3 % Hong Kong; 0.21 %), US Americans (0.23 %), Nige- of the UK's total population. Among non-British rians (0.14 %) and Bangladeshi (0.13 %) are the nationals, 59.4 % (3,681,859 persons; 5.5 % of to- most relevant nationalities. In total, more than tal population) came from the EU- 40 % (41.5 %) of TCN living in the UK belong to 28 (2013 – 2020; excl. reporting country UK) and these nationalities. Figure 1 gives an overview of EFTA countries (35,947 persons; 0.05 %). A total of the origin and the share of the different nationali- 2,482,375 persons (or 3.7 % of the total popula- ties in the country. tion) were third country nationals (TCN), being in- Population Development & Population Structure dividuals who are neither British, EU-28 citizens nor EFTA citizens. By comparison, the share of TCN The number of inhabitants in the UK increased in in EU-27 (excl. UK) and EU-28 (incl. UK) is 5.7 % the past and this development is projected to con- and 3.8 %, therefore the UK is lying below the av- tinue in the future. As of January 1, 2019, there erage. were 66,647,112 persons registered. This corre- sponds to an increase of 6,862,432 citizens

2

(+13.4 %) within the last 20 years (e.g. before the There was an increase of labor force potential rec- eastward enlargement of the EU in 2004). orded in the past 15 years and this development continues in the future as working-age population A long-term analysis of population development is projected to increase at by +1.1 % until 2040. from 2002 to 2019 reveals clear regional differ- Half of the regions tend to grow in the future with ences. Manchester (+30.00 %), Hackney & New- highest percentual change in West Midlands ham (+38.4 %) and Tower Hamlets (56.7 %) record (+5.8 %), East Midlands (+4.5 %), London (+3.5 %) the highest population growth rates. Though the and South West (+3.1 %). In Scotland (-4.0 %), majority of regions (95.0 %) recorded a population North East (-3.5 %), Wales (-2.9 %) and Northern growth, there are also some regions suffering Ireland (-2.4 %) labor force is projected to shrink from a decline. Sunderland (-2.1 %), Blackpool (- (see Figure 4). 2.2 %) and Sefton (-2.2 %) are among those most affected by depopulation (see Figure 2). The total-age dependency ratio is a measure of the age structure of the population. It relates the num- From an economic perspective, labor (done by hu- ber of individuals who are likely to be “dependent” man beings) is an essential element in the produc- on the support of others for their daily living – the tion of goods an d services. The term ‘labor force’ young (up to 19 years old) and the elderly comprises all of those who work for gain, whether (65 plus years old) – to the number of those indi- as employees, employers, or as self-employed, viduals who are, being working age from 20 to 64 and it includes the unemployed who are seeking years old, capable of providing this support. work. The total-age dependency ratio in United Kingdom As Figure 3 (a)-(d) indicate, the increase in the is 72.5 %. The dependency rate is usually lower in working-age population (+2,403,200 inh.) from metropolitan areas and cities than in rural regions. 2006 to 2019, i.e. longest available period, within The key factors here are a more prosperous econ- the regions is mainly caused by immigration from omy, i.e. more people in working-age, and low the EU (+1,600,900 inh.) and from outside the EU child ratios. Densely populated regions in United (+254,600 inh.), though the British citizens in Kingdom are also characterized by relatively low working-age increased substantially as well since rates. High dependency ratios are found in South 2006 (+ 539,900 inh.). West (79.5 %), East (77.3 %) and Wales (76.6 %) From a regional perspective, the areas with the whereas lowest values are reported from London highest working-age population growths are Lon- (58.3 %) and Scotland (67.3 %). As a result of age- don (+18.1 %), East of England (+7.2 %), South East ing societies, the total-age dependency ratios will (+6.5 %), Northern Ireland (+6.2 %) and West Mid- rise in the future by 10.2 pp to 82.6 % in 2040 re- lands. The North East (+0.8 %) and Wales (+2.4 %) sulting in implications/challenges for publicly recorded the lowest growth rates. Scotland and funded social security schemes (e.g., expenditures Northern Ireland grew by +4.7 % and +6.2 %. for pensions).

3

Figure 2: Population development (Δ 2002 to 2019), NUTS 3

Source: Eurostat (2020b), own illustration

4

Figure 3: Working-age population (20 to 64 years; Δ 2006 to 2019) (a) Total +2,403,200 inh. (c) Non-EU-28 +254,600 inh.

(b) British +539,900 inh. (d) EU-28 +1,600,900 inh.

Source: Eurostat (2020c), own illustration

5

Figure 4: Projections of working-age population (20 to 64 years; Δ 2020 to 2040)

Source: Office for National Statistics (2020a), own illustration; Office for National Statistics (2020b), own illustration; Office for Na- tional Statistics (2020c), own illustration; Office for National Statistics (2020d), own illustration

6

Education and (Un)Employment (5.2 %; female: 5.0 %; male: 5.3 %) and Northum- berland and Tyne and Wear (5.1 %). Dorset and The tertiary education rate of United Kingdom's Somerset (1.8 %; 2.2 % female; 1.4 % male), Cum- nationals is 39.6 %, which is slightly below that of bria (1.9 %) and Highlands and Islands (2.0 %) are the total population (40.6 %). Compared to the the provinces with the lowest unemployment EU-28 (29.5 %) the United Kingdom has an above- rates (see Figure 5). average tertiary education rate. In 2020 the coun- try ranked 3rd among the EU-28 member states Table 1: Working-age Population by citizenship, degree of for share of tertiary educated residents, following urbanization and educational attainment level 2019 Ireland (40.7 %) and Luxembourg (41.0 %) and ahead of Cyprus (40,0 %), Finland (38.5 %) and Educational Attainment Level Lithuania (37.9 %). While the Czech Republic Citizenship (ISCED11) (21.6 %), Italy (17.4 %) and Romania (16.0 %) have Urbanisation the lowest tertiary education rate. Primary Secondary Tertiary British 19.9% 39.9% 40.2% Across the world, educational attainment is signif- EU-28 15.7% 38.3% 46.0%

Cities Cities Non-EU-28 18.5% 29.7% 51.7% icantly higher in cities1 than in towns and suburbs2, Total 19.5% 39.1% 41.5% 3 which in turn is higher than in rural areas . This is British 19.3% 42.5% 38.2% not the case for the United Kingdom where pri- EU-28 13.0% 44 .9% 42.1% mary, secondary and tertiary levels of education Non-EU-28 14.4% 31.4% 54.2% suburbs Towns & Towns are almost evenly spread around cities, towns and Total 18.9% 42.3% 38.8% British 18.3% 41.3% 40.5% suburbs as well as rural areas. Approx. 40 % of the EU-28 13.4% 41.1% 45.5% population have a university degree, i.e. tertiary Rural Non-EU-28 15.7% 33.2% 51.1% education (ISCED2011 levels 5 to 8), and another Total 18.1% 41.2% 40.8% 40 % hold a secondary, i.e. upper secondary and Source: Eurostat (2020i), own illustration post-secondary non-tertiary educational In general, there is a correlation between the level (ISCED2011 levels 3 to 4) attainment level. Pri- of education and unemployment, according to mary educational attainment level has low rele- which the well-educated are significantly less vance as approx. only 1 out of 5 residents in UK likely to be unemployed than people without a vo- belongs to this group. Differences of educational cational qualification. This correlation can also be attainment level between British, non-EU-28 resi- seen for the UK: those with primary education are dents and EU-28 residents range within a narrow most affected by unemployment (5.6 %). Con- band from 0.2 pp to 6.3 pp (see Table 1; cf. Caputo, versely, people with tertiary education have the Bianchi & Baglioni 2021). In total, EU-28 (45.2 %) lowest unemployment rate (2.5 %), those with and domestic population (39.6 %) have a lower secondary a value of 3.5 %. Similar results are ob- tertiary education rate than non-EU-28 citizens served at the European level (EU-28; Figure 6). (52.1 %). In contrast, the higher the educational attainment The unemployment rate in UK is 3.4 % (2019), be- the higher is the employment rate, i.e. 86.0 % for low the EU-28 average of 6.2 %. From a regional tertiary, 78.8 % for secondary and 64.6 % for pri- perspective, West Midlands has the highest unem- mary level. From an immigration point of view, ployment rate at 5.3 % (female: 5.1 %; this observation holds for other citizenships as male: 5.5 %), followed by Tees Valley and Durham well, as the risk of unemployment decreases by

1 A city is a local administrative unit (LAU) where at least 50 density of at least 1 500 inhabitants per km² and collectively % of the population lives in one or more urban centres (Eu- a minimum population of 50 000 inhabitants after gap-filling rostat, 2020d). (Eurostat, 2020e; Eurostat, 2020f). 2 Towns & suburbs are areas where less than 50 % of the 3 A rural area is an area where more than 50 % of its popula- population lives in rural grid cells and less than 50 % of the tion lives in ‘rural grid cells’, i.e. grid cells that are not identi- population lives in urban centres, i.e. a cluster of contiguous fied as urban centres or as urban clusters (Eurostat, 2020g; grid cells of 1 km² (excluding diagonals) with a population Eurostat, 2020h).

7

Figure 5: Unemployment rate 2019 (20 to 64 years) EU-28 citizens have the lowest employment rate on each level (74.5 %; 64.7 %; 51.9 %; see

Figure 7).

Figure 7: Employment rate by educational attainment level and citizenship (20 to 64 years; ISCED 2011)

100 89.5 90 84.8 86.5 78.8 79.0 80 74.5 70 64.4 64.7 60 51.9 50 40 30 20 10 0 Primary Secondary Tertiary

EU-28 Non EU-28 United Kingdom

Source: Eurostat (2020k), own illustration

The abovementioned findings foster the theoreti- Source: Eurostat (2020j), own illustration cal conclusions by Becker (1964) that the educa- Figure 6: Unemployment rate by educational attainment tional level has a significant influence on the pro- level (20 to 64 years; ISCED 2011) fessional career and the risk of unemployment.

14 This theory and the empirical evidence can also be 12.4 12 applied to the “Early Leavers from Education and Training”. This population group refers to persons 10 aged 18 to 24 who have completed a lower sec- 8 ondary education and are not involved in further education or training. From a fiscal point of view, 5.6 6 5.6 this specific group is of particular relevance as they 4.0 4 3.5 are more likely to be (long-term) unemployed, and 2.5 thus causing costs for the economy. 2 In 2019, 10.3 % of the 18-24 year-old in the EU-28 0 Primary Secondary Tertiary were part of this group (male: 11.9 %; female: 8.6 %). The UK ranges with a value of 10.9 % slightly above the European average. The share of EU-28 United Kingdom EU-28 citizens residing in UK is above (12.5 %) and Source: Eurostat (2020j), own illustration that of non-EU-28 is below (9.3 %) that of the do- mestic population (10.9 %). In total, approx. 50 % the highest educational attainment also for immi- (45.9 %) of all early leavers are jobless, but no re- grants. But EU-28 citizens have the highest em- lying statements can be made in regard to nation- ployment rate for all educational attainment lev- ality because of missing or unreliable data (Euro- els (89.9 %; 84.8 %; 78.8 %) while British rank sec- stat 2020l). ond on each level (86.5 %; 79.0 %; 64.4 %). Non-

8

Income and Gross Domestic Product (20.3 %) and British (22.0 %) face the lowest risk of social exclusion and poverty (see Table 2; cf. Ca- As for the employment rate, the educational at- puto, Bianchi & Baglioni 2021). tainment level has a positive impact on mean and median4 equivalized net income, i.e. the higher Figure 9: Mean and median equivalized net income by the educational attainment level the higher the in- broad group of citizenship 2018 (18 to 64 years) come. 30,000 27,186 26,892 Working persons in the UK at primary level earn 26,439 25,000 less (mean: € 20,007; median: € 17,764) than sec- 22,774 22,924 19,947 ondary (€ 23,837 or € 21,120) and tertiary levels 20,000 (€ 32,406 or € 28,046; see Figure 8). 15,000 Figure 8: Mean and median equivalized net income by edu- cational attainment level 2018 (18 to 64 years; ISCED 2011) 10,000

35,000 32,406 5,000 30,000 28,046 0 25,000 23,837 EU-28 Non EU-28 United Kingdom 21,120 20,007 20,000 17,764 Mean equivalised net income Median equivalised net income 15,000 Source: Eurostat (2020n), own illustration 10,000 In 2019, compensation of employees (wages and 5,000 salaries plus employers’ social contributions) was 0 the largest income component of GDP in EU-28 Primary Secondary Tertiary and the Euro area, accounting respectively for 47.8 % and 48.0 %. The UK (49.5 %) is above Euro- Mean equivalised net income pean average (both EU-28 and Euro area). Taxes Median equivalised net income on production and imports (less subsidies) ac- Source: Eurostat (2020m), own illustration counted for 11.9 % (EU-28: 11.9 %; Euro Working British citizens have a comparable educa- area: 11.5 %). Gross operating surplus and mixed tional attainment level as EU-28 and non-EU-28 income accounted for 38.5 % of GDP in UK, 40.3 % residents, nevertheless the mean (€ 26,892 vs. of GDP for the EU-28 and 40.5 % in the Euro area. € 27,186 and € 26,439) and median (€ 22,924 vs. British GDP at current prices amounted to approx. € 22,774 and € 19,947) equivalized net income of € 2,526.6 billion in 2019 and GDP per inhabitant TCN are lower than the British ones (see Figure 9). equaled € 37,830 (Eurostat 2020p).

The indicator ‘people at risk of poverty or social Within almost 20 years gross domestic product exclusion’ corresponds to the sum of persons who (GDP) increased by 36.0 % (real growth rate). In are: at risk of poverty after social transfers, se- terms of regional GDP all provinces recorded pos- verely materially deprived or living in households itive real growth since 2000, ranging from 27.6 % with very low work intensity. As non-EU 28 citizens in Northern Ireland to 50.0 % in London. Scotland have a higher risk of unemployment and they earn lies slightly below British average, i.e. 33.2 %. The less than domestic residents and EU-28 approx. further away we move from London, the less the one third (31.8 %) belong to this group. EU-28 economic growth. Solely North West (35.0 %) and

4 The median is more robust, i.e. less sensitive to outliers than the mean. This explains why the mean equivalized income is always slightly higher for each educational level.

9

the East of England (35.0 %) almost reach the na- has the lowest regional GDP per capita in United tional level of 36.0 %. Kingdom (see Table 3).

Analyzing GDP per capita reveals that all regions Provinces with a high share of foreigners have a except for London (£ 54,686) and South East high GDP per capita and vice versa. On the one ((£ 34,083) lie below the national level of £ 31,976. hand, prosperous regions are more attractive to GDP per capita in North East (£ 23,569), Wales migrants; on the other hand, these regions benefit (£ 23,866), Yorkshire and The Humber (£ 25,859), from a better availability of labor supply, i.e. East Midlands (£ 25,946), Northern Ireland greater number of working age people. In this con- (£ 25,981) West Midlands (£ 27,087) is less than text, one speaks of a so-called cross-fertilization, half of London. Scotland ranks in the midfield which results from the interaction of these two (£ 29,660) with highest regional value in the North factors. Eastern (£ 40,914). Southern Scotland (£ 19,937)

Table 2: Population aged 18 and over by risk of poverty, material deprivation, work intensity 2019 Risk of poverty At risk Not at risk At risk Not at risk At risk Not at risk Work Intensity Material Deprivation United Kingdom EU-28 Non-EU-28 Very low Severe 1.1 0.4 0.1 0.1 2.2 0. 4 Not very low Severe 1.1 1.5 1.3 1.9 2.4 3.3 Very low Non severe 1.9 2.4 2.0 2.0 3.2 1.7 Not very low Non severe 13.6 12.9 18.6 People at Risk of Poverty or Social Exclusion 22.0 20.3 31.8 Not at risk Not at risk Not at risk Not very low Non severe 78. 1 79.7 68.2 People at no Risk of Poverty or Social Exclusion 78.1 79.7 46.0 Total 100.0 100.0 100.0 Source: Eurostat (2020o), own illustration

Table 3: Nominal regional GDP (2018) East Midlands 124,647 32.4 25,946 2000 to 2018) Derbyshire and and real growth rate (Δ 55,999 28.3 25,367 Nottinghamshire Leicestershire, Rutland and 51,021 35.3 27,717 Northamptonshire Lincolnshire 17,627 38.4 23,322 Region West Midlands 159,832 31.0 27,087

GDP (in £) GDP (in Herefordshire, Worcestershire and 40,716 37.7 30,045 annual growth rates annual GDP (in £) £) per capita GDP (in Gross Domestic Product Product Gross Domestic at current marketat current prices marketat current prices Warwickshire chained measuresvolume chained Shropshire and 41,664 29.1 25,574 North East 62,644 26.4 23,569 Staffordshire Tees Valley and Durham 26,286 20.5 21,882 West Midlands 77,452 28.5 26,557 Northumberland and 36,357 31.6 24,960 East of England 186,462 35.0 30,069 Tyne and Wear East Anglia 71,915 35.2 28,597 North West 207,452 35.0 28,449 Bedfordshire and 65,065 41.1 35,100 Cumbria 14,028 39.1 28,118 Hertfordshire Greater Manchester 78,918 35.8 28,059 Essex 49,482 28.2 26,999 Lancashire 39,124 33.9 26,112 London 487,145 50.3 54,686 Cheshire 37,427 38.4 40,207 Inner London - West 210,224 63.0 176,015 Merseyside 37,956 29.7 24,464 Inner London - East 115,826 56.9 48,143 Yorkshire and Outer London - 141,698 29.0 25,859 43,390 24.0 22,626 The Humber East and North East East Yorkshire and 24,727 20.0 26,529 Outer London - South 36,356 23.5 27,910 Northern Lincolnshire Outer London - 81,348 41.7 38,968 North Yorkshire 23,368 32.6 28,345 West and North West South Yorkshire 31,010 33.2 22,104 South East 311,300 30.8 34,083 West Yorkshire 62,593 29.6 26,977

10

Berkshire, Economic structure and entrepreneurship Buckinghamshire and 103,321 35.7 42,915 Oxfordshire In general, a distinction is made between three dif- Surrey, East and West 93,700 27.8 32,380 ferent sectors of the economy, the primary (agri- Sussex Hampshire and Isle of culture and forestry), secondary (manufacturing) 65,712 26.9 33,091 Wight and tertiary (services) sectors. Kent 48,567 33.4 26,302 South West 158,084 32.2 28,231 EU-28 citizens are overrepresented (above 7.1 %) Gloucestershire, in “Manufacturing” (11.0 %), “Wholesale and re- Wiltshire and 83,158 34.0 33,188 %), Bath/Bristol area tail trade hotels and restaurants” (9.0 Dorset and Somerset 33,881 24.7 25,442 “Transport and communication” (8.8 %), “Con- Cornwall and 12,555 43.2 22,096 struction” (8.7 %), “Financial and business ser- Isles of Scilly vices” (7.4 %) and “Agriculture, forestry and fish- Devon 28,490 30.7 23,858 Wales 74,906 28.4 23,866 ing” whereas they are of low relevance in “Energy West Wales and and water” (4.5 %) and “Public administration, ed- 41,924 26.5 21,274 The Valleys ucation and health” (4.0 %). East Wales 32,982 30.4 28,238 Scotland 161,295 33.2 29,660 Non-EU-28 citizens preferably work in “Transport North Eastern 20,008 47.7 40,914 and communication” (5.6 %), “Wholesale and re- Scotland tail trade hotels and resta Highlands and Islands 13,786 35.7 29,372 urants” (4.5 %)”, “Finan- Eastern Scotland 65,917 32.7 33,094 cial and business services” (4.5 %) and “Other Ser- West Central Scotland 42,707 29.2 27,713 vices” (4.1 %) whereas they are underrepresented Southern Scotland 18,877 27.4 19,937 (below 4.0 %) in “Public administration, education Northern Ireland 48,887 27.6 25,981 and health” (3.7 %), “Energy and water” (3.0 %) United Kingdom 2,124,351 36.0 31,976 %; see Table 4). Source: Office for National Statistics (2019), own calculations and “Construction” (2.2 and illustrations

Table 4: Employment by economic activity 2016 Non Economic Activity British EU-28 n.a. EU-28 Primary Sector Agriculture, forestry and fishing 88. 3 % 7. 4 % n.a. 4. 4 % Secondary Sector Energy and water 91.9% 4.5% 3.0% 0.6% Manufacturing 86.0% 11.0% 2.9% 0.1% Construction 89.1% 8.7% 2.2% 0.1% Tertiary Sector Wholesale and retail trade hotels and restaurants 86.4% 9.0% 4.5% 0.1% Transport and communication 85.5% 8.8% 5.6% 0.1% Financial and business services 88.0% 7.4% 4.5% 0.1% Public administration, education and health 92.2% 4.0% 3.7% 0.0% Other services 90.3% 5.5% 4.1% 0.1% Total 88.8% 7.1% 4.0% 0.1% Source: Office for National Statistics (2017), own illustration

Research and Innovation development (R&D), innovation, entrepreneur- ship and economic growth achieved by human Following the neoclassical and endogenous capital from abroad (EU-28 and non-EU-28). growth theories, technological advance is believed to be one of the major drivers of economic A well-known indicator provided to measure growth. From this perspective, there is a growing achievements of countries or regions in R&D is interest to investigate the link between research & GERD, i.e. regional/national gross domestic ex- penditure on R&D as a percentage of GDP. GERD is estimated at 2.05 % for United Kingdom in 2018

11

which is a slight increase of 0.05 pp compared to North Yorkshire 406.6 1.55% 2017. The total amount of research expenditures South Yorkshire 461.5 1.33% 41.90 billion, the largest share (54.9 %) was West Yorkshire 760.6 1.08% was € East Midlands 2,498.6 1.81% financed by the business enterprise sector. 26.0 % Derbyshire and 1,751.1 2.85% were financed by the government sector. 13.9 % Nottinghamshire Leicestershire, Rutland were contributed by foreign countries (rest of the 664.0 1.15% and Northamptonshire world) and 5.2 % and 0.6 % were funded by the Lincolnshire 83.5 0.44% private non-profit and the higher education sector West Midlands 3,712.8 2.07% (Eurostat 2020q). Herefordshire, Worcestershire and 1,869.1 3.98% Regional data is published on NUTS 1 and NUTS 2 Warwickshire Shropshire and level. The East of England has the highest ratio in 358.2 0.81% Staffordshire 2018 (3.57 %) followed by the South East (2.26 %), West Midlands 1,485.5 1.68% West (2.07 %) and East Midlands (1.81 %). Except East of England 7,458.2 3.57% for Scotland (1.69 %) all other provinces on East Anglia 4,462.4 5.47% Bedfordshire and NUTS 1 level are well behind and below the all- 2,158.4 2.95% Hertfordshire British ratio of 1.73 %. Wales ranks last (1.08 %) Essex 837.5 1.55% with London (1.16 %), Yorkshire and The Humber London 6,651.1 1.16% (1.18 %), North (1.27 %) and West (1.43 %) doing Inner London - West 4,154.1 1.63% only slightly better, see Table 5). Inner London - East 915.6 0.67% Outer London - East and 145.8 0.31% In accordance to the research & development North East (R&D) funding structure almost more than one Outer London - South 190.1 0.46% Outer London West – 1,245.4 1.36% half (54.4 %) perform research activities in the and North West higher education sector followed by the business South East 7,945.6 2.26% enterprise sector (41.9 ). The remaining are lo- Berkshire, cated in the government (2.2 %) and private non- Buckinghamshire and 4,388.5 3.65% Oxfordshire profit sector (1.5 %). This makes a total of 317,472 Surrey, East and West 1,641.3 1.55% full-time equivalent (FTE; Eurostat 2020s). Sussex Hampshire and Isle of 1,191.6 1.65% Available data for other countries has shown that Wight employees in R&D have an above average share of Kent 724.3 1.34% South West 2,807.6 1.59% foreigners compared to the total number of em- Gloucestershire, ployed persons by economic activity. Wiltshire and 2,033.7 2.19% Bristol/Bath area Table 5: Intramural R&D expenditure as percentage of Dorset and Somerset 318.0 0.83% gross domestic product (GDP) 2017 Cornwall and Expenditures % of 49.1 0.35% Province Isles of Scilly (€ mio.) GDP Devon 406.8 1.27% North East 881.1 1.27% Wales 891.6 1.08% Tees Valley and Durham 324.9 1.13% West Wales and 447.9 0.99% Northumberland and The Valleys 556.3 1.37% Tyne and Wear East Wales 443.7 1.19% North West 3,333.1 1.43% Scotland 3,058.8 1.69% Cumbria 208.2 1.37% North Eastern Scotland 342.0 1.48% Greater Manchester 1,104.1 1.22% Highlands and Islands 69.7 0.45% Lancashire 357.1 0.82% Eastern Scotland 1,710.9 2.33% Cheshire 979.1 2.37% West Central Scotland 863.3 1.80% Merseyside 684.6 1.60% Southern Scotland 72.8 0.36% Yorkshire and 1,855.9 1.18% Northern Ireland 808.7 1.51% The Humber United Kingdom 41,903.4 1.73% East Yorkshire and 227.3 0.88% Source: Eurostat (2020r), own illustration Northern Lincolnshire

12

Conclusion

This Statistical Briefings examines the economic From a spatial point of view, the concentration of and spatial impact of immigration with special re- better-educated migrants is highest in cities. Prov- gard to third-country nationals (TCN) in the UK. inces with a higher ratio of TCN are also character- ized by higher economic growth, being more at- On January 1, 2019, a total of 6,200,181 persons tractive for TCN on the one hand and offering with Non-British citizenship were living in the more labor supply on the other hand. country, a former EU-28 member state (until 2020). This corresponds to a share of 9.3 % of the The below average regional GDP growth rate, es- UK's total population. Among non-British nation- pecially in diffusely populated regions like Wales als, 59.4 % (5.5 % of total population) came from and Northern Ireland is indirectly linked to immi- EU and EFTA countries (0.05 %). 3.7 % of total pop- gration from abroad but primarily to internal mi- ulation were third country nationals (TCN), being gration from rural areas to cities and metropolitan individuals who are neither British, EU nor EFTA areas. These emigration flows tend to reinforce citizens. By comparison, the share of TCN in EU- the erosion of the economic basis, because they 27 (excl. UK) and EU-28 (incl. UK) is 5.7 % and have a negative impact on the potential workforce 3.8 %, therefore UK lying below the average. and the economic attractiveness of the region. Economic performance is stagnating, accompa- Within the last 20 years the development of the nied by declining population figures. population in the country has mainly been due to immigration. From 2006 to 2019 all regions prof- Governments need to invest in education on the ited heavily from immigration with regard to its one hand and in research and development (R&D) working-age population. Without immigration, on the other, to enable long-term prosperity and many regions would have suffered from a de- growth. In 2018 the national gross domestic ex- crease in its population aged between 20 and 64 penditures on R&D (as a percentage of GDP) were years. estimated at 2.05 %, which is a slight increase of 0.05 percentage points to 2017. But there is again Despite the relevance of migration for labor sup- a striking regional variation from densely popu- ply, the employment rate of TCN is lower com- lated areas with a high number of innovative en- pared to UK nationals and EU-28 citizens. One terprises (e.g. The East and South East) to less would consider the educational attainment level densely populated regions (e.g. Wales). as relevant aspect, but in the country EU citizens as well as TCN are better educated than British Summing up, immigrants in the UK play a pivotal people. Hence, other reasons have to be consid- role in reviving rural areas, and with the appropri- ered to explain the gap. ate policies, they could play an even more signifi- cant role in sustaining them.

13

Bibliography and source references

Becker, Gary Stanley (1964): Human Capital: A theoretical Eurostat (2020n): Mean and median income by broad group and empirical analysis, with special reference to education. of citizenship (population aged 18 and over), 3rd edition. New York. https://appsso.eurostat.ec.europa.eu/nui/show.do?da- taset=ilc_di15&lang=en Caputo, Maria Luisa; Bianchi, Michele; Baglioni, Simone (2021): Statistical briefing for Scotland on social impact of Eurostat (2020o): Persons aged 18 and over by risk of pov- TCNs in MATILDE regions (= MATILDE Deliverable 3.2). erty, material deprivation, work intensity of the household and most frequent activity status of the person - intersec- Eurostat (2020a): Population on 1 January by age group, sex tions of Europe 2020 poverty target indicators, and citizenship, https://appsso.eurostat.ec.eu- https://appsso.eurostat.ec.europa.eu/nui/show.do?da- ropa.eu/nui/show.do?dataset=migr_pop1ctz&lang=en taset=ilc_pees02&lang=en Eurostat (2020b): Population change - Demographic balance Eurostat (2020p): GDP and main components (output, ex- and crude rates at regional level (NUTS 3), penditure and income), https://appsso.eurostat.ec.eu- https://appsso.eurostat.ec.europa.eu/nui/show.do?da- ropa.eu/nui/show.do?dataset=nama_10_gdp&lang=en taset=demo_r_gind3&lang=en Eurostat (2020q): GERD by sector of performance and source Eurostat (2020c): Population by sex, age, citizenship, labour of funds, http://appsso.eurostat.ec.eu- status and NUTS 2 regions, https://appsso.eurostat.ec.eu- ropa.eu/nui/show.do?dataset=rd_e_gerdfund&lang ropa.eu/nui/show.do?dataset=lfst_r_lfsd2pwn&lang=en Eurostat (2020r): Intramural R&D expenditure (GERD) by Eurostat (2020d): City, NUTS 2 regions, https://ec.europa.eu/eurostat/data- https://ec.europa.eu/eurostat/statistics- browser/view/tgs00042/default/table?lang=en explained/index.php?title=Glossary:City Eurostat (2020s): Total R&D personnel and researchers by Eurostat (2020e): Town or suburb, https://ec.europa.eu/eu- sectors of performance, educational attainment level rostat/statistics-explained/index.php?title=Glos- (ISCED2011) and sex, https://ec.europa.eu/eurostat/data- sary:Town_or_suburb browser/view/rd_p_persqual11/default/table?lang=en Eurostat (2020f): Urban centre, Office for National Statistics (2017): International immigra- https://ec.europa.eu/eurostat/statistics- tion and the labour market, UK: 2016, explained/index.php?title=Glossary:Urban_centre https://www.ons.gov.uk/peoplepopulationandcommu- Eurostat (2020g): Rural area, https://ec.europa.eu/euro- nity/populationandmigration/internationalmigration/arti- stat/statistics-explained/index.php/Glossary:Rural_area cles/migrationandthelabourmarketuk/2016

Eurostat (2020h): Rural grid cell, https://ec.europa.eu/euro- Office for National Statistics (2019): Regional gross domestic stat/statistics-explained/index.php?title=Glossary:Ru- product all NUTS level regions, ral_grid_cell https://www.ons.gov.uk/economy/grossdomes- ticproductgdp/datasets/regionalgrossdomesticproductall- Eurostat (2020i): Population by educational attainment level, nutslevelregions sex, age, citizenship and degree of urbanization, ht tp://appsso.eurostat.ec.europa.eu/nui/show.do?da- Office for National Statistics (2020a): Subnational population taset=edat_lfs_9916&lang=en projections for England: 2018-based, https://www.ons.gov.uk/releases/subnationalpopula- Eurostat (2020j): Unemployment rates by sex, age, educa- tionprojectionsforengland2018based tional attainment level and NUTS 2 regions (%), https://appsso.eurostat.ec.europa.eu/nui/show.do?da- Office for National Statistics (2020b): Principal projection - taset=lfst_r_lfu3rt&lang=en Scotland summary, https://www.ons.gov.uk/peoplepopula- tionandcommunity/populationandmigration/populationpro- Eurostat (2020k): Employment rates by sex, age, educational jections/datasets/tablea16principalprojectionscotlandsum- attainment level, citizenship and NUTS 2 regions, mary https://ec.europa.eu/eurostat/data- browser/view/tepsr_wc140/default/table?lang=en Office for National Statistics (2020c): Principal projection - Wales summary, https://www.ons.gov.uk/peoplepopula- Eurostat (2020l): Early leavers from education and training tionandcommunity/populationandmigration/populationpro- by sex and citizenship, http://appsso.eurostat.ec.eu- jections/datasets/tablea15principalprojectionwalessummary ropa.eu/nui/show.do?dataset=edat_lfse_01&lang=en Office for National Statistics (2020d): Principal projection – Eurostat (2020m): Mean and median income by educational Northern Ireland summary, https://www.ons.gov.uk/people- attainment level - EU-SILC survey, http://appsso.euro- populationandcommunity/populationandmigration/popula- stat.ec.europa.eu/nui/show.do?lang=en&dataset=ilc_di08 tionprojections/datasets/tablea17principalprojectionnorth- ernirelandsummary

14