Acronyms. Austria Belgium Bulgaria. Czech Republic Germany Denmark Estonia. Croatia Hungary Ireland. Lithuania Luxembourg.

Transcription

1

2 Acronyms AT BE BL CY CZ DE DK EE EL ES FIN FR HR HU IE IT LT LU LV ML NL PL PT RO SE SK SL UK Austria Belgium Bulgaria Cyprus Czech Republic Germany Denmark Estonia Greece Spain Finland France Croatia Hungary Ireland Italy Lithuania Luxembourg Latvia Malta Netherlands Poland Portugal Romania Sweden Slovakia Slovenia United Kingdom EU28 All current EU Member States EU15 EU Member States before 2004 enlargement EU13 EU Member States who joined the EU after 2004 EU8 Estonia, Latvia, Lithuania, Poland, Czech Republic, Hungary, Slovakia, Slovenia EU2 Romania and Bulgaria 2

3 Executive summary The accession of new Member States has profoundly changed the scale and patterns of intra-eu migration. The number of expatriate EU citizens living in EU Member States other than their home country has substantially increased and is continuing to grow. Yet, the minds of EU decision-makers have been primarily preoccupied with the inflows of legal (illegal) immigrants from the third countries, and relatively little attention has been paid so far on studying the new patterns of intra-eu migration, although there are strong indications that the recent intra-eu migration presents a qualitatively new phenomenon if compared to earlier episodes of intra-eu migration waves. What is more, the European agenda has completely omitted the emigration of EU nationals toward third countries like the US, Canada, and China, despite the fact that emigration is increasing due to economic hardship in many parts of Europe. In general, economic migration is being associated with more efficient allocation of human capital, thereby improving the prospects for economic growth. However, due to its multifaceted nature, the migration phenomenon is not easy to be studied. The existing research methods deal only with particular aspects of migration, and there is neither a single theory nor a method that would permit studying the economic aspects of people s mobility holistically. Another issue, hampering proper research, is unavailability of relevant data on aggregate and individual level. The latter is particularly important for migration studies, since a decision to move or to stay involves a myriad of personal choices. Indeed, each migrant s personal formation and employment history, as well as behaviour during migration, determine whether her/his migration will have a positive (negative) impact on the economies of both sending and destination countries. From the European perspective, it is relevant to assess whether intra-eu migration contributes to the building of an integrated EU labour market, helping to adjust to asymmetric shocks, or whether it leads to new forms of inequalities between different cohorts of European population and regions? If so, who wins and who loses? It is apparent that despite high political status attached to free movement of people, the notion of an integrated EU labour market is still a distant reality. The existing studies on intra-eu migration show that the overall impact on the EU from the post-enlargement migration has been positive, yet, serious negative side-effects exist, which impede full utilisation of migration s positive potential. Without properly devised policies aimed at reducing these negative externalities, despite the positive impact of migration, people will suffer and raise their voice against freedom of movement. The negative externalities include: (1) uneven distribution of economic gains between destination and sending countries; (2) overcrowding in certain municipalities in destination countries, (3) increased inequality between different cohorts of people; and (4) waste of human capital due to limited skills transferability across borders. Indeed, 3

4 untamed migration may, on the one hand, halt the economic development of source countries and, on the other hand, lead to obstruction of public infrastructure in overcrowded destinations. Understanding the causes and effects of migration in an enlarged EU is a precondition for designing effective migration policies in Europe, and thus, a precondition for reaching the targets of European jobs and growth strategy Europe 2020 as well as the key objectives of the European Employment Strategy. Indeed, the diverse post-enlargement migration flows of predominantly young people constitute an important policy issue that interacts with these targets in both receiving and sending countries, and, therefore, one has to agree with Marju Lauristin, the Estonian Member of the European Parliament, that a comprehensive policy concerning economic, social, and cultural implications of the free movement of EU citizens still needs to be developed (Lauristin, 2015). The analysis in this paper will begin with examination of migration statistics and challenges posed by incomplete data to migration research (Part I). Farther on, the focus will be laid on the examination of the dynamics of intra-eu people flows, e.g., circularity and volatility (Part II). A look at the theoretical and empirical aspects of migration on the European Union in toto (Part III) and on destination (host) (Part IV) and sending (origin) (Part V) countries economies will follow. At the end the case of Latvia will be studied in somewhat greater detail (art VI). This paper will conclude with a set of recommendations on how to improve the EU policy-making regarding migration. 4

5 Part I. People s mobility in the EU putting the figures right According to official statistics, in 2013, 13.6 million or 2.69% EU nationals lived in other EU Member State than their own. This number had slightly increased, e.g. since 2011, when only 12.4 million had residence in a foreign EU Member State (10% increase). The highest share 1 of foreign EU nationals in 2013 had been registered in Luxembourg (38%), followed by Cyprus (13%) and Ireland (8.3%). However, the number of EU nationals living in another EU Member State is dwarfed by the vast size of immigrant communities of third-country nationals. In 2013, the number of people born outside of the EU and living in the EU Member States was at least 20.4 million by almost twice exceeding the corresponding number of mobile EU citizens (see Table 1). In fact, only in Ireland, Hungary, Luxembourg, Slovakia, and Cyprus the number of immigrants born in other EU27 Member States was higher than the number of persons born outside of the EU (see Figure 1). The nationals of old Member States are considerably less mobile than the nationals of new Member States, while their countries remain the most popular destination for all immigrant groups. Thus, in 2013, only 5.89 million or 1.16% of the EU15 nationals lived in another EU Member State; besides, out of them overwhelming 98.8% lived in another EU15 Member State. The most popular destination countries for EU15 nationals are Germany, United Kingdom, and Spain. The picture is different for the EU13 countries. In their case one can see mostly a one-directional move from the East to the West. In 2013, out of 5.82 million EU13 nationals living in a foreign EU country, representing 5.61% of their population, 94.8% lived in the EU15. The most attractive countries are United Kingdom, Germany, and Spain. The attractiveness of the EU15 is also very high among the immigrant third-country nationals 77.8% lived in the EU15 in 2013 (see Table 2). Interestingly, the historic record of European people s mobility shows that the number of expatriate EU15 nationals residing in other EU Member States has not much changed since Thus, the number of EU15 citizens-foreigners had grown from 5.5 to 6.0 million between 1990 and What has changed though between the 1960 s and early 2000, is the composition of intra-eu15 migrants, namely, in the words of Recchi et al. (2003), the the European [EU15] migrant of the early twenty-first century is a highly qualified young professional replacing low-skilled migrants from earlier migration epochs. A look at recent cross-border flows of people in the EU reveals that the number of Europeans moving between the Member States is increasing. Thus, between 2008 and 2012, the annual scale of cross-border movements has jumped from 1.4 million to The share of foreign population in a Member State is calculated against the total number of population of the respective Member State. 5

6 million representing 42% increase. But again, the third-country nationals still represent a major source of immigration to most of the EU Member States. Thus, in 2012, million foreign born nationals arrived to the EU (see Table 3). The EU15 still remains the most popular destination for all immigrants and, in 2012, out of total 3.38 million incoming migrants to the EU28, 83.5% stayed in EU15 countries. Yet, again, only 776 thousand or 1.82% of the EU15 citizens moved to another EU country: 1.77% stayed within the EU15, and only 0.05% moved to the EU13. In the meantime, 5.61% of EU13 nationals changed their residence country, and, again, 5.32% moved to the EU15 with the remaining 0.29% moving to another EU13 country (see Table 2). A noticeable recent tendency is the increasing level of return migration. In 2012, 880 thousand people, or 0.18% of the total EU population, returned to their home country. This represented 1/3 of intra-eu28 people s movement, and a large part of returnees, 358 thousands (41%), went to the EU13 countries. The outflows and inflows of people are quite unbalanced at the level of particular states. In 2013, the countries with the largest expatriate communities abroad compared to their population were Romania (11.3%), Lithuania (9.1%), and Ireland (8.0%). At the same time the countries that were most open to foreign EU nationals were Luxembourg (38.4%), Cyprus (13%), and Ireland (8.3%) (see Figure 2). Ireland, in fact, is the only EU country having simultaneously a high share of its own nationals abroad and a high share of foreign population, therefore, one can say that Ireland is a country most open to circularity of people in the EU. A look at the migration flows in a longer-term perspective shows that, between 2003 and 2012, the highest positive net migration rate was experienced by Luxembourg (increase by 15.7% of its population in 2003), Cyprus (13.7%), and Spain (8.2%). At the same time the greatest negative rates relate to Lithuania (-9.8%) and Latvia (-7.8%) (see Figure 3). Yet, a caveat should be introduced in relation to these migration figures. Namely, the precision and reliability of statistical data on people s migratory flows are to be qualified, as almost every study on migration reports problems related to data (e.g. Eurostat, 2014; CEPS, 2015; Recchi, 2003; Galgóczi, 2012). One of the major problems reported is difficulties in tracking mobile citizens. First, as pointed out by Eurostat, it is harder to count people leaving a country than those arriving, and the so-called mirror statistics return significant discrepancies between immigration and emigration numbers of EU residents moving around the EU. The explanation is that most EU Member States base their statistics on administrative data, while only some use sample surveys. However, since no incentives to deregister at the moment of departure exist, the overall picture as presented by authorities is skewed, as they slightly 2 The figures from 2012 are adopted to include also the numbers for Croatia which became the EU Member State as of 1 July

7 overestimate immigration (some people return home soon after arrival in destination country) and severely underestimate emigration. The example of Latvia is telling. Before 2011 census, according to official registers, only a moderate number of Latvian people had emigrated between 2006 and However, after the census, these figures had to be revised massively from 3 to 12 times (see Table 4). 3 Another major problem is the speed of processing of data and the content of information given by these data. In the preparation of this report in April 2015, the most recent data available from Eurostat in cross-country perspective were from the year 2012 on migration flows and from the year 2013 on stocks of foreign population in the EU Member States. 4 Yet, even these data are marked for low reliability and contain considerable data gaps at Member State level. No doubt, the quality of data on people s migration has improved substantially since Today, it is possible to obtain data on acquisition and loss of citizenship by sex, age group and former citizenship, and also on intermarriages. Besides, Eurobarometer conducts now regular public opinion polls involving themes related to migration. Notwithstanding these improvements, the data on migration contain considerable loopholes. Thus, the aggregated data on migration reveal only sex, age, and nationality of mobile people; however, there is no disaggregation in terms of status of the mobile people (e.g., salaried work, self-employed, family members) and educational level. Also, there is no data on short-term migration which includes seasonal and posted workers, not to mention longitudinal data on individuals (CEPS, 2015) which would allow assessing the circularity of people s movement. It is also to be mentioned that an increasing number of EU nationals have accepted double nationality, that is, they become nationals of two or even three EU Member States, and such incidences are not accounted either. 3 Eurostat points out that the post-census revision of migration tables had been carried out by 10 EU Member States and that another 5 were late with their new data submissions (Eurostat, 2014). 4 In fact, Eurostat publishes the population data every year in April. 5 According to Recchi et al. (2003), by 2003, the research and data on the (recent) history of intra-european migration and its causes had been patchy, incomplete and presented significant spots regarding some of the most elementary trends and facts, namely, on intermarriages and naturalizations and on structural accounts of migration stocks. Recchi et al. also note that there was a need for more studies of the migrants own experiences and outcomes, as migration decision always carries a subjective dimension back in

8 Table 1. Immigrant communities in the EU Member States, by nationality, 2013 Host Origin EU28 EU15 EU13 EU28 EU15 EU13 countries countries* countries** countries countries* countries** thousands % EU28 nationals EU15 nationals EU13 nationals Third-country nationals Host country nationals Total population Notes: (*) Represented only by 13 pre-2004 EU Member States, as data on Greece, Luxembourg, and France are not available. (**) Represented only by 11 post-2004 new EU Member States, as data on Cyprus and Malta are not available. Source: Eurostat, table [migr_pop1ctz]. Table 2. Intra-EU immigrants compared to population of origin countries, EU15 and EU12, 2013 Host Origin EU28 countries EU15 countries* EU13 countries** % EU15 countries* EU13 countries** Notes: (*) Represented only by 13 pre-2004 EU Member States, as data on Greece, Luxembourg, and France are not available. (**) Represented only by 11 post-2004 new EU Member States, as data on Cyprus and Malta are not available. Source: Eurostat, table [migr_pop3ctb]. Table 3. Flows of immigration to EU Member States, by nationality, 2012 Destination Sending EU28 countries EU15 countries* EU13 countries** EU28 countries EU15 countries* EU13 countries** thousands % of EU28 % of sending country s population population Total EU15 nationals EU13 nationals Return migration Third-country nationals Notes: (*) Represented only by 13 pre-2004 EU Member States, as data on Greece, France, and Portugal are not available. (**) Represented only by 11 post-2004 new EU Member States, as data on Cyprus and Malta are not available. Source: Eurostat, table [migr_imm1ctz]. 8

9 Table 4. Net migration to Latvia, , before and after 2011 census Reported before census Reported after 2011 census Scale of correction, times Source: Central Statistical Bureau of Latvia Figure 1. Share of EU28 and third-country nationals in population, %, 2013 Figure 2. Share of foreign EU15 and EU13 nationals in population, %, 2013 Source: Eurostat [migr_pop1ctz] Note: no data available for Greece, Cyprus, Luxembourg, Malta, and France. Source: Eurostat [migr_pop1ctz] 9

10 Eurostat does reports on the shares of foreign-born people, including from other EU Member States, within the EU Member States. Indeed, it is a proven fact that the firstgeneration immigrants retain close links with their country of birth; therefore, such statistics is relevant. As it can be seen from the Table 5, there were 50.8 million people in the EU born in a foreign country (compared to 34.1 million immigrants in the EU who still hold the nationality of their home country). Yet, Eurostat does not provide data on the second- and third-generation immigrant people, even though many of them may still have sympathies towards their ancestors home country. What is more, and not least relevant, is the magnitude of communities of EU nationals abroad in the third countries. Eurostat reports only on emigration flows, but not on the size of these European diaspora communities overseas. 6 Anyhow, a lack of timely, precise and reliable data is indeed an unfortunate situation and clearly a shortcoming caused by imperfect integration of national labour markets of the EU Member States. The available aggregate data provide a general picture, but fail to reveal the complexity of migration trajectories that would be required to measure such phenomena as brain gain, brain drain or brain waste. There is also no data collected on a regular basis on what constitutes the main barriers to mobility and on the economic impact of return migration. Shortage of reliable data leads to a skewed picture of social and economic reality (e.g., it returns false indications on the real rate of unemployment, flows of national income, and external competitiveness). This inhibits not only successful national policy-making, but also effective refutation of populist claims about harmfulness of people s migration. Table 5. Immigrants communities by nationality and origin of birth, thousands, 2012 Residing in EU28 Origin EU28 nationals EU15 nationals* EU13 nationals** Third-country nationals By current nationality countries By origin of birth n.a Notes: (*) Represented only by 13 pre-2004 EU Member States, as data on Greece, Luxembourg, and France are not available. (**) Represented only by 10 post-2004 new EU Member States, as data on Cyprus and Malta are not available. Source: Eurostat, table [migr_pop3ctb]. 6 According to a survey conducted by the Europeans Throughout the World among the Council of Europe Members States in 2014, the size of European diaspora communities overseas, be it in other EU Member State or in third-countries, could be estimated between 80 to 150 million people. 10

11 Part II. Future migration flows, circularity issue and volatility of flows In order to grasp the future trends of intra-eu people s migration flows a look at the forces behind the current migration patterns and possible adjustments to it is warranted. The question is about the trajectory of flows will they continue to be linear (onedirectional) or circular (multi-directional)? Another issue to be addressed is the volatility of migration flows. How predictable are people s intentions regarding migration and is it a challenge to macroeconomic forecasting? Due to its multifaceted nature, the migration phenomenon is not easy to be studied. The existing research methods deal only with particular aspects of migration, and there is neither a single theory nor a method that would permit studying the economic aspects of people s mobility holistically. Bodvarsson and Van den Berg (2009) attribute this problem to different scopes and time horizons between labour economists, who traditionally build single-sector models of a closed economy and focus on short-term effects, and international trade economists, who build multi-sector models of an open economy and focus on long-term effects. Thus, the two missing pieces from migration research appear to be a general equilibrium model that differentiates between short- and long-run adjustments and takes account of open and closed economies with single and multiple goods sectors. Currently there are no models examining secondary adjustments of economy to migration like capital responses, internal migration, changes to output mix, and technological and product market responses. Based on studies on migration, Bodvarsson and Van den Berg (2009) propose three categories or models, each corresponding to a particular motive for migration. Specifically, a migrant can be: (1) a supplier of her factor services or, effectively, a maximizing investor in her human capital, (2) a consumer of amenities and public goods, or (3) a producer of her own household goods and services. The first category comprises those who see migration as an investment in their human capital. These migrants goal is to maximize utility by choosing the location which offers the highest net income. For them, migration is essentially an investment decision, as moving involves certain costs followed by an uncertain payoff in the future, and consideration of regional income differences is the most essential motivator. The second category consists of those who migrate for consumption purposes. As desirable goods like attractive scenery, a pleasant climate, clean air, decent public services, cost of living, etc., are not universally available, people migrate to satisfy their utility function [that] includes goods and services that are not all available in each geographic market. And, at last, the third category comprises those who s main motive for individual and family migration is the cost of household production. The implications of the third group often match those of the human capital and consumption models of migration, and the assumption is that since there are significant spatial differences in goods and prices and amenities, and household s goal is to maximize utility by choosing the optimal combination of commodities to produce and 11

12 consume, subject to household s income to purchase goods and capital, the family will relocate to areas with higher income opportunities or greater levels of household production. Bodvarsson and Van den Berg (2009) also note that according to some researchers, psychological, social and information costs of migration are likely to fall when there is greater access to family, friends and other previous migrants in the destination country. The kinship (family) networks and migrant (shared ethnic or regional background) networks can greatly enhance the efficiency of migration. Besides, as Bodvarsson and Van den Berg (2009) underline, migration could also be modelled as an investment process undertaken at each stage of the life cycle rather than a one-investment decision. Young people may prefer to migrate to high income areas, whereas persons close to retirement may have a strong preference for locations with good climate and healthcare. The implication is that as there are multiple stages to the life cycle, it is very likely that there will be multiple migrations during person s life. The empirical data of intra-eu migration in general do confirm these theoretical assumptions. It seems that the main motivator behind migration of people in the EU has been income differences the larger the difference the more intense migration ( supplypush immigration). Indeed, the Baltic countries and also Bulgaria and Romania are most affected by outflow of people, and these are the countries with the lowest income level in the EU. Although the income gap is narrowing, it still remains considerable, therefore, from this perspective, the east west migration will continue. According to some authors (e.g., Kahanec et al., 2009), unemployment in source countries is also an important factor behind emigration; however, there is also evidence pointing to the fact that wage differences are a much more powerful driver of mobility than unemployment in the EU. Thus, Barslund and Scwarzwälder (2015), analyzing the impact of financial crisis on mobility, conclude that, except Ireland, high unemployment rates in the periphery (Spain, Italy, Greece, Portugal) have led only to limited mobility. Likewise, a study on patterns of emigration from Latvia finds out that most of emigrants had been employed shortly before departure (Hazans, 2011). Another relevant factor of migration is demand for labour in destination countries ( demand-pull immigration). Indeed, as is pointed out by Bodvarsson and Van den Berg (2009), the migration decision depends not just on the differences in wages across countries, but on where the immigrant would fit into the destination country labour market and how well the worker s abilities and other human capital can be applied. The post-enlargement migration from the new Member States to the old Member States provides a good example. It is argued that the transitional arrangements 7 applied by a 7 The so-called transitional arrangements, included in the Accession Treaties, set a seven-year period during which certain limitations in free movement of persons is possible. Initially, these arrangements were applied by all old EU Member States, except the United Kingdom, Ireland, and Sweden. Last restrictions towards

13 number of the old EU Member States restricting inflow of labour from the new Member States had a diversion effect on the east west migration flows (Barslund and Scwarzwälder, 2015). Thus, Germany, which hitherto was the most popular destination country for migrant labour, did loose its attractiveness because of applied restrictions, while the United Kingdom and Ireland, which along Sweden lifted restrictions on people s movement from the day one after the enlargement, saw a surge of immigrants from the Eastern Member States. Interestingly enough, the scale of immigration to Sweden, however, was considerably smaller. The suggested reasons for Sweden s apparent lack of attractiveness included fewer job vacancies, less flexible labour markets and language issues (Kahanec et al., 2009). Finally, Barslund and Scwarzwälder (2015) find out that the financial and subsequent crisis slowed down mobility flows from the new Member States, as unemployment rose in the EU15. Barslund and Scwarzwälder (2015) also point out that mobility flows have rebounded somewhat but that the rates have remained lower than before the crisis. Moreover, they observe that the crisis has triggered a redirection of mobility flows away from the periphery towards Germany, the United Kingdom and other northern European countries. What is more, the return migration has accelerated after the crisis, with Romania, Lithuania, and Poland being the main beneficiaries. In the case of the old EU Member States, cross-border exchange of people is very low, and its economic impact is negligible. Apparently, the tumbling costs of long-distance travel across Europe and different mobility programmes installed by the European institutions have not substantially increased the appetite of people to move. The motive of net return is working neither in the EU15, as people apparently prefer to stay unemployed for extended periods of time rather than to seek a job in the neighbouring country. This lack of enthusiasm for migration is explained by the language barrier, cultural differences, and a lack of cross-border ties (Barslund and Scwarzwälder, 2015). Take example of Germany and France. They both have largely equal level of income; yet they have developed a considerable gap between their respective unemployment rates. From 2008 to 2013, the difference in unemployment rate from roughly equal in both countries (7.4%) has widened to 5.2% in Germany and 10.3% in France. Despite that, the number of French nationals living in Germany has increased only by a slight 6% between 2008 and 2013 from to thousands (0.18% of France s total population). enlargement countries were lifted by the end of Regarding Croatia, limitations can be used for up to mid

14 Table 6. Unemployment and immigration of French citizens to Germany Unemployment rate, Germany Unemployment rate, France Number of French citizens living in Germany, 1000 Source: Eurostat Notwithstanding the general low level of mobility among the old Member States, several EU15 countries deserve some attention. First, as it has been already mentioned in this paper, Ireland stands out as a country most open to labour cross-border flows it has accepted large EU28 immigrant inflows (around 8% of its population) and has experienced equally large outflows of people to other EU countries. Second, Germany, the uncontested European economic powerhouse, has become the most popular destination country not only for the EU13 migrating nationals, but also to a large part of EU15 migrants (the immigrants from the EU15 represented 60% of the EU28 nationals living in Germany in 2013, except Germany s own nationals). Third, there are groups of countries that form chains of people s flows. Thus people move from Ireland to the United Kingdom and from the United Kingdom to Spain. Likewise, Romanians move towards Italy, while Italians towards Germany and Spain. It seems that, the south north migration flows are dominated by mostly young, economically active people who are in search for better paid jobs, while somewhat elderly people from the northern countries are relocating towards the South because of climate and scenery and, not less important, also because of lower cost-of-living reasons. Interestingly, in 2013, Spain alone had attracted 60% of the southward intra-eu migration, leaving other Southern Member States like Italy and Portugal far behind. 8 Returning to the east west migration, the issue of circularity and volatility of labour flows merits some attention. From the efficiency point of view, circular labour movement is much more preferable than one-time migration. Permanent settlement in a particular country may be beneficial from migrants family and personal comfort point of view, however, from the EU perspective, and in particular from the sending country s viewpoint, fluidity would foster fuller utilisation of human capital. There are already some good signs in this respect, indeed. As it has been noted, return migration has 8 According to Recchi et al. (2003), it is possible to discern three main intra-eu15 migration patterns: (1) the legacy of traditional south-north movements; (2) movements linked to geographical un cultural proximity (DE AT, BE FR, SE FIN, UK IE); (3) retirement migrations, which helps to explain the preponderance of Germans and Britons settled in southern EU Member States. 14

15 substantially increased in the aftermath of the crisis. At the same time, there are a number of factors which deter people from return or relocation to other countries. At first, there is the real estate and mortgage issue. Immigrant people, who have acquired a real estate and taken a mortgage loan, are less likely to resort to remigration. Second, generous and lengthy unemployment payouts may act as another drag on migration. Why to emigrate if prospective wage in the new destination country does not match the level of joblessness payouts? Volatility of labour flows is another issue. On the one hand, it may act as a valve from short-term asymmetric economic shocks. On the other hand, a lack of appropriate measurement techniques exacerbates the risk of a wrong policy response to uncontrolled and spontaneous flight of people from a crisis-struck countries thus leading to even greater imbalances. Already today a significant proportion of migration from the new member states is of a temporary or seasonal nature (Kahanec et al., 2009). One can expect volatility to increase, as every next cycle of migration is much less costly because of collected knowledge, gained experience and broader social contacts abroad. Kinship and migrant networks have a role, too, in enhancing the volatility of migrant flows. 9 Thus, in the case of Latvia, around 40% of those who had returned back to Latvia have emigrated again (Mierina, 2015). Summing up all arguments, put forward in literature, it is possible to distinguish three major groups of factors influencing the scale of future migration flows. These are: (1) accelerating factors, (2) decelerating factors and (3) factors having ambivalent impact. Accelerating factors: - Increase in wage difference between sending and destination countries; - Employment opportunities in destination countries; - Earlier migration experience, inter alia, foreign language proficiency and social contacts; - Consumption considerations. Decelerating factors: - Limited transferability of skills (as a result, low-skilled migration dominates over high-skill migration; it leads also to a perverse effect, as high-skilled migrants often accept low-skilled work in destination countries, eventually giving up their skills (brain-waste)); 9 A paper by the International Monetary Fund claims that there are also informational costs regarding the labour market in the destination country. This leads to an inverted U-shaped emigration pattern, whereby migration begins slowly, increases rapidly when information about employment conditions flows back to the sending country from previous migrants, and then tapers off once labour markets in the two countries converge (IMF, 2006). 15

16 - Large distance between the sending and destination country, high costs of travel; - Language and cultural barriers; - Generosity of welfare systems (lengthy payouts may discourage unemployed immigrants from searching new job either back at home or in other EU Member States). Ambivalent factors: - Family reasons (willingness to reunify may lead to both supplementary immigration or remigration; in countries with poor social protection, a willingness to hedge against the risk of loss of income may result in sending one spouse abroad); - Education (it is observed that increase in mean education increases emigration, while increase in a variance of educational performance decreases emigration (Bodvarsson and Van den Berg, 2009)). Kahanec et al. (2009), based on Brücker et al. model, report that the long-run migration stock for the first-wave new Member States (the EU8) stands at 6% of the population of the sending countries, and for the second-wave Member States (Romania and Bulgaria or the EU2) at about 10 12% of their population. Thus, in 2009, the migration from the new Member States amounted to about a half of their long-run potential. This finding roughly corresponds to the estimates of Holland et al. (2011) on total migration potential from the EU10 being between 6 10% of their population. Moreover, Kahanec et al. (2009) claim that labour migrants are typically very responsive to economic cycles, especially if there are no mobility restrictions to returning home and re-entering the host country. What is more, Kahanec et al. (2009) states that during the recent crisis high-skilled immigrants, who were young, single and with no family ties in the host country, were more likely to return at home than low-skilled immigrants who could not hope to find a job back home. The evidence from the Latvian emigrant communities, however, contradicts this observation, as in the case of Latvia the low-skilled expatriate males are most likely to return to Latvia (Mierina, 2015). 16

17 Part III. The impact of intra-eu migration on the European Union as a whole High geographical mobility of people can bring substantial benefits to an economically integrated area, because it balances labour supply and demand across borders, and helps to reduce labour and skill mismatches. Free movement of people not only leads to more efficient allocation of labour resources and human capital, but also fosters trade and technology transfer among the receiving and sending countries, thus generating growth and prosperity. The literature on migration puts forward ample ideas on advantages of migration. In sum, the positive impact of migration emerges through: - Increased labour markets flexibility in both sending and receiving countries; - Better balance between productivity and wage growth; - More international trade and capital flows, thus deepening integration of geographic areas; - Spillover of economic growth (positive demand shock) from core (destination) countries to periphery (sending); - More intense flow of skills and know-how which boosts productivity (positive supply shock); - Concentration of smartest people in regional technological centres thus enhancing innovation; - Easier adjustment to adverse and asymmetric economic shocks, as mobility reduces unemployment in depressed areas; - Revitalisation of depopulated areas (e.g., in East Germany) and decaying oldindustry sites (e.g., in England). Besides, in the EU, people s mobility carries a significant political significance. Namely, freedom of movement is expected to foster social and cultural interaction and better mutual understanding between EU citizens, thus serving the ideal of an ever closer union among the peoples of Europe. 10 However, people s mobility can have also a negative impact, as it may: - Lead to unemployment in certain cohorts of labour in the destination countries, on the one hand, and excessive wage increase in certain occupations in the sending countries, on the other hand, thus reinforcing inequality; - Intensify already existing demographic problems and, without compensatory return migration, may lead to permanent loss of purchasing power and shrinkage of economy of the sending country; - Add to existing overpopulation problems in the destination countries putting pressure on public infrastructure and public services; 10 Article 1 of the Treaty on the European Union. 17

18 - Result in brain waste as limitations on skill transfer lead to a loss of human capital; - In the worst case scenario, inhibit economic growth in periphery and result in permanent underdevelopment of sending countries. The impact of migration on the destination and sending countries will be discussed in greater detail in the following chapters. From the EU perspective, the effect of migration on general economic growth and its capacity to (1) elevate labour supply and demand mismatches and (2) smooth out asymmetric economic shocks will be addressed. The later is of particular relevance to Eurozone. In theory, labour migration can be expected to increase overall GDP in the receiving countries and decrease it in the sending countries, due to the increase or decrease, respectively, of labour as a production factor. The European Commission (2008) points to several studies which model the impact of intra-eu migration on GDP and other macroeconomic variables of the European economies after the 2004 enlargement. The overall conclusion is that most of these studies find relatively modest GDP effects in the short run and more substantial effects in the long run, although the exact results vary significantly depending on underlying assumptions concerning expected future migration flows, the skill mix of native versus migrant workers, the speed of adjustment of capital stocks and other factors. Thus, a study by Kahanec et al. (2009) finds that, on balance, the economic impact of post-enlargement migration has been very low, though positive. Thus, between 2004 and 2009, the post-enlargement labour flows have increased the GDP of the integrated area by about 0.2%, or by about 24 billions euro, in the long run. However, this has not been a win-win situation from the perspective of both receiving and sending countries. The EU15 countries have added extra 0.26% to their GDP, with main beneficiaries being Ireland (2.93%) and the United Kingdom (0.89%). The EU8, on the contrary, has lost 1.1% from its GDP, due to emigration. The major losers are Poland (-1.94%), Slovakia (- 1.51%) and Lithuania (-1.15%) Please, note that these figures do not take into account migration from Bulgaria and Romania. As both countries have seen large outflows of people, the disparity of gains between the old and the new Member States could be even bigger. 18

19 Figure 3. Migration in EU Member States, , % of population in 2003 Figure 4. Monthly wages in industry, construction and services, EUR Notes: no data available data for Romania, Bulgaria, France, Greece, Malta, Belgium and Spain. Source: Eurostat [migr_imm1ctz], [migr_emi1ctz] Notes: no data available for Malta and Poland. Source: Eurostat [lc_ncost_r2] Figure 5. EU Member States ranked by immigration inflows, % share of population, 2012 Source: Eurostat [migr_imm3ctb] 19

20 Another study, commissioned by the European Commission (Holland, 2011), finds substantially larger impact of migration on the European economy. According to their estimates, the potential output of the EU has increased by 0.8% thanks to migration from the EU10 (Malta and Cyprus are not considered in this study). Most positively affected countries were Ireland (3%) and the United Kingdom (1.2%). On the other side, Romania, Bulgaria, and Lithuania had to expect a permanent reduction in their potential output by 5 10%, while Latvia and Estonia at least 3% reduction. In the EU, the development of labour markets still remains a national competence, with the EU institutions performing only a coordination role, despite high importance attached to the principle of free movement of people. This certainly hampers the development of a transnational EU labour market. Despite recent legislative progress, a number of problems continue to haunt mobile people from a lack of recognition of qualifications and access to certain occupations, to political disenfranchisement, red tape, and restrictions on cross-border transferability of social entitlements. Also family issues (unification, education of children, and divorce) frequently cause serious disturbances. 11 Besides, administrative hurdles, insufficient language skills and a lack of information on mobility opportunities abroad are enlisted among the most pertinent obstacles to low intra-eu mobility (European Commission, 2008). In 2008 the European Commission (2008) claimed that the intra-eu mobility had not led to serous disturbances on national labour markets: at least not until 2008, when the tide of post-accession migration had started to subside. The European Commission also found that a characteristic feature of post-enlargement mobility had been its short duration, that is, for a few months or years. This included posted and seasonal workers. Moreover, the European Commission also observed that the short-term mobility had become a sizeable phenomenon not only concerning the east west dimension, but for the EU as a whole. It was admitted, however, that posted workers from the EU12 countries had been suspected of circumventing the transitional labour restrictions imposed by the EU15, and social and other work-related rules of the receiving countries, which, in general, had been much stricter in the EU15 than in EU12 countries. Looking at the age and occupational structure of migrants, those who migrated from the EU15 were mostly young and employed as professionals (27% of all EU15 mobile citizens) or highly skilled technicians (17%). For comparison, from the EU12, the migrants were also young, but their occupation was limited to elementary works (>31%), craft and related trades (>16%) and machine operating (below 18%). From the EU12, above 70% were below 35 years, while in the case of the EU15, only 62% were under 35 (European Commission, 2008). What is more, the proportion of migrants with tertiary 11 Some of these problems are enlisted in the EU Citizenship Report 2013, which was published by the European Commission on 8 May

21 education had increased substantially after enlargements, leading to emergence of two distinct emigrant groups low-skilled and high-skilled ones (Kahanec et al., 2009). A theory on integrated economic areas sees open labour markets as a factor which can partially absorb asymmetric macroeconomic shocks, be they unexpected due to changing economic fortunes and a turn in business cycle, or expected like demographic shifts. This is particularly important in monetary and currency exchange rate unions, as people s migration may help to relieve downward (deflation) and upward (inflation) pressure on wages and prices. However, the theory also notes two occasions in which the relevance of labour mobility to successful functioning of a monetary union diminishes. The first occasion involves a fiscal transfer system among union s members enabling macro assistance to ailing members of the union. The second occasion refers to the situation in which the members of the union have achieved a high level of mutual convergence, and shock is equally distributed (symmetric) all over the union. Indeed, before the crisis of , the high degree of convergence in terms of low unemployment across the Eurozone s member states was used as a popular excuse in explaining low labour mobility among the euro countries. However, after the crisis this trend reversed. Alas, in Eurozone, there is almost no fiscal transfer system in place (albeit incipient safeguard mechanisms in the form of the European Stability Mechanism and European deposit insurance scheme have been introduced in response to recent financial turmoil). And, despite economic calamities in a number of Eurozone s Member States (e.g., Ireland, Portugal, Spain, and Greece), the immigration of people from these countries has not substantially increased. CEPS (2015) has a point here by insisting that intra-eu labour mobility is too low to support the single labour market as anything but a notion and too low to play anything other than a modest role in helping to rebalance the Eurozone after the crisis. Indeed, the intra-eu mobility is low compared with the US: less than 3% of EU citizens currently reside in another EU country, and the annual flows of people across the EU internal borders do not exceed 0.3% of the population. This makes only around one-tenth of the respective reference figure in the US 12 there the migration rather than unemployment is the major source of adjustment to state specific shocks (Blanchard, 2013, 29). In the EU, on the contrary, the unemployment is preferred over mobility. According to CEPS (2015), if anything but mostly wage differences are the main driver behind mobility in the EU, and, as observed by IMF (2006), most region-specific shocks in Europe are absorbed by changes in participation rates rather than by labour migration. It is relevant to note that in the view of the European Commission (see Bontout and 12 According to OECD, the rate of inter-state mobility in the US was 2.4% in However, even if internal migration within the EU countries is taken into account which increases the rate of migration in the EU up to 1% of population, the migration rate in the EU is still much below the US inter-state rate (Bontout and Lejeune, 2014). 21

22 Lejoune, 2013), in addition to traditional obstacles to free movement of people 13, also housing market regulations and the rise in home ownership, the lack of information about vacancies overseas, and prevalence of a dual-earner model in Europe increase inherent costs to migration. What is more, the restrictions on transferability of unemployment benefits discourage unemployed from seeking new job in other EU Member States These include low transferability of qualifications, lack of language skills, cultural differences, divergence among educational systems and administrative barriers. 14 Currently, EU law guarantees that unemployed citizens who are entitled to unemployment benefits and who go to look for a job in another EU country will continue to receive their unemployment benefits from their home country for a maximum period of three months. See EU Citizenship Report

23 Part IV. The impact on the destination countries Theoretical primer The theoretical literature on economic effects of immigration on destination countries has two strands. The first deals with macro effects and a general conclusion is that the immigration surplus 15 is usually positive although very small. The micro studies, on their part, focus on distributional effects, and the finding is that these have been substantial in magnitude. Notwithstanding the specificities of each theoretical model, the most pertinent research questions regarding destination countries are the following: 1. Does immigration raise or lower average destination country income? 2. Which native-born people in the destination country enjoy income gains, which suffer losses, and how much are the gains and losses? 3. How does immigration affect product markets in the destination country? 4. How do answers to the above differ from the short run vs. the long run? The standard labour supply and demand model predicts a positive impact from immigration on the destination country: the more people, the bigger the economy. In other words, immigration increases the returns to scale. However, the model shows also that despite general economic expansion, individuals may see their welfare diminishing, as limited wealth is distributed over an increased number of people. As a consequence, per capita income may fall, at least initially, until capital stock adjusts and labour-tocapital ratio returns to its natural level. Yet, this is not the whole story. Immigration not only helps to expand the economy, but also creates efficiency gains from improved allocation of labour. These gains are called immigration surplus (for more details see Bodvarsson and Van den Berg, 2009). The immigration surplus can be acquisitioned either by native workers, by owners of capital or by immigrants themselves. The actual distribution of the immigration surplus depends on labour market response to labour supply shock, capital market flexibility and also consumption market s reaction to immigration. It is quite feasible that under certain circumstances immigration may lead to increased inequality in the destination country. The economic immigrants are supposed to gain under all scenarios otherwise they would not immigrate. The fortune of capital owners depends on the country s openness to capital flows, on the one hand, and homogeneity of immigrants and local labour markets (a homogeneous labour market has no separation in low and high skilled market segments), on the other hand. If immigrants and native labour markets are homogenous, immigrants will substitute local workers and, if capital supply is fixed, locals will see 15 The immigration surplus is attributed to the general gain from the immigration to the destination economy. The crucial point is on how to increase this surplus and, not least important, how to distribute this surplus, so that every party immigrants, local workers, and capital owners gains. 23

24 their wages diminish. As a result, immigration surplus will be appropriated by capital owners and immigrants, while native workers will lose. However, if capital is flexible, more labour will attract more capital, thus complementing each other, and immigration, although initially pushing down wages, will lead to a surge in capital inflows after a while. In this case immigration surplus is short-termed and ephemeral. In fact, only the heterogeneity of immigrant and local labour market (e.g., containing low-skilled and high-skilled segments) gives a chance to local labour to improve its wellbeing from immigration. The rule is that, if skill levels differ between immigrants and locals, immigration moves wage levels in opposite directions, and immigration surplus will depend on the scale of disparity in skill distribution. More precisely immigrant skills must differ from natives for the destination country to benefit from the admission of immigrants. The message is that skilled immigration benefits unskilled and unskilled immigration benefits the skilled native labour. Thus, inflow of foreign labour may put pressure on wages and increase competition for jobs vis-a-vis the respective skill group of native workers, while at the same time creating opportunities for complementary skill groups of native workers (Kahanec et al., 2009). Apart from these standard models, Bodvarsson and Van den Berg (2009) show that immigration need not always have adverse effect on natives. First, immigration may encourage a larger share of population to invest in higher levels of education and training, if low-skilled immigration increases returns to high-skill occupations (the Johnson Model). Second, immigration boosts demand for local products, and, the more of local produce is consumed locally, the lesser the adverse wage effect of immigration on natives (the Altonji and Card Model). Third, immigrants and natives, ultimately, are not perfect substitutes, and natives mostly compete with other natives and immigrants with other immigrants. Even within the same occupation or skill level, the human capital endowment of a native will differ (be superior) from immigrant s because of source and destination countries differences in education, language, culture, and family environment. An important channel through which an economy adjusts to labour inflow is changes to demand for local products, because immigrants are not just workers, but also consumers. Hence, along labour markets, also the product markets reactions are to be studied. Interestingly, there are studies which claim that demand for local products (housing, food, services) and thus demand for local labour, may rise even before the supply of immigrant labour happens. A model by Hercowitz and Yashiv s suggests that product supply is a function of native population, immigrant labour and the relative price of domestic goods to imports. Which effect of immigration dominates increase in price of domestic goods because of surge in demand or decrease in price of domestic goods because of lower wages and production costs depends upon wage and price elasticities, labour and capital output shares, and the extent to which immigrants participate in the 24

25 goods market relative to the labour market. Another model, suggested by Bodvarssen and van den Berg (2006), estimates the impact of the rate of immigrants remittance on consumption of local produce, and make a distinction between two types of immigration, namely, between the demand-pull immigration caused by increase in destination country s export industries, which results in wage increase all over the economy, on the one hand, and the supply-push immigration, driven by conditions overseas and, therefore, exogenous to the local community, which depresses local wages and prices, on the other hand. To sum up, the labour and product market studies claim there are two separate effects on local wages: an input substitution effect and a consumer demand effect. The argument is that the ultimate effect of immigration on destination country s wages and unemployment will depend on: (1) immigrants substitutability for local workers; (2) the specific factor labour and capital price elasticities of demand; and (3) immigrants propensity to spend on local products. In addition, there are a series of studies analyzing how immigration affects the size and composition of demand. Immigrants often have different tastes, and these studies suggest that the size effect can be lessened by compositional changes in demand due to these differing tastes. The theoretical models discussed earlier assume that immigration has an exogenous effect, namely, that it delivers unexpected and instantaneous shock. However, in reality, immigration is an ongoing and anticipated process, and capital adjusts continually to actual and anticipated flow of new immigrants. Moreover, the long-term effects differ from the short-term effects because economies adjust to immigration shocks. Hence, labour market analyses should be compounded by reflections on secondary adjustments experienced by destination countries capital responses, internal migration, changes in output mix, and technological and product market responses. According to the neo-classical growth model (Solow s model), an economy cannot ensure permanent improvements in living standards solely through factor accumulation. Continuous improvements in per capita income can only occur if there is a technological progress. Correspondingly, Bodvarsson and Van den Berg (2009) argue that immigrants do not only work and consume, but they act also as innovators inducing technological progress. Accordingly, Bovarsson and Van den Berg (2009) suggest four channels through which immigration may influence technological progress. First, immigrants are carriers of ideas and knowledge, and therefore immigration increases the transfer of technologies. Second, immigrants often have talents and personalities that are especially appropriate for innovation. Indeed, immigration is a risky endeavour, for that reason immigrants are likely to be less risk averse than the average person in their countries of origin. As such, they can contribute to innovation as entrepreneurs and workers in innovative activities. Third, by changing the levels of population and gross domestic product of destination countries, immigrants change the size of economies. And, lastly, 25

26 by adding new, entrepreneurial people to population they increase innovative competition by reducing the ability of vested interests to take protectionist measures to slow the process of creative destruction. Empirical evidence According to a study by the European Commission (2008), from a destination country perspective, immigration is important because: (1) it helps to alleviate labour market and skills shortages; (2) it attenuates and spreads the effects of demographic change over a longer time; and (3) it fuels economic growth as labour pool grows. The following issues will be addressed hereinafter: 1. The impact on growth potential and structure of the receiving countries economy; 2. The impact of immigrants on job markets in the destination countries; 3. The impact of immigration on society s age structure in the destination countries; 4. The impact on macro-economic stabilization in the destination countries; 5. The impact of immigration on fiscal and welfare system in receiving countries. The impact on growth potential and structure of the receiving countries economy Baas et al., based on their research of migration from EU8 to EU10 countries between 2004 and 2007 (Kahanec and Zimmerman, 2009), conclude that the total factor income of the native population has increased slightly in the EU15 destination countries. They find that labour market effects are surprisingly small: in the short run, wages declined by about 0.1% in the EU15, and the unemployment rate increased by 0.1%. In the long run, after adjustment of capital stocks, migration is by and large neutral for wages and unemployment. Also, the European Commission finds that in the long run, the economic impact of immigration on the existing population is likely to be small (European Commission, 2008). Holland et al. (2009) are of similar opinion that the long-term impact on receiving EU15 countries will be slightly positive, and, like Baas et al. (2009), agree that among the receiving countries the largest impact on GDP growth from the immigration in the long run will accrue to Ireland and the United Kingdom. The impact of immigrants on job markets in the destination countries It is relevant to note that according to a study by OECD (2012), immigrants represented 15% of new entries in strongly growing occupations in Europe (health care occupations, science, technology, engineering, mathematics). Simultaneously, immigrants were also filling labour needs by taking up jobs regarded by domestic workers as unattractive or lacking career prospects. In Europe, such unattractive occupations include craft and related trades works, also machine operators and assemblers. Yet, as Begg (2015) points out, there is a hypocritical reluctance in the EU Member States to acknowledge the contribution of migrants to the effective functioning of many sectors of activity. Indeed, 26

27 without immigrants, many low-skilled jobs would be very hard to fill, from agriculture to care services, yet the public opinion in many EU15 countries has become negative towards immigration. True, as suggested by empirical studies, immigration creates winners and losers, but only in the short term. Winners are mainly the immigrants themselves and their employers, and losers are earlier migrants who are employed in low-paid jobs and, thus, are in direct competition with newly arrived migrants (European Commission, 2008; Kahanec et al., 2009). Kahanec and Zimmermann (2009) find that the skill structure of migrant workforce from the EU8 does not significantly differ from the native workforce in the receiving countries; therefore, only moderate distributional effect across different skill groups in the labour markets is felt. Besides, even those who had lost job because of immigration from the new Member States had soon found another occupation as inflow of people boosted the whole economy thus creating more jobs. The majority of migrants from the new Member States have secondary educational attainment (61% according to European Labour Force Survey from 2008), and they are proportionally less represented at the higher and lower ends of the skill spectrum compared to natives of the EU15 (22% versus 27% and 17% versus 27%, respectively). Relative to the population in the sending countries, the skill level of migrants from the new Member Sates is higher than that of natives who stayed behind. It is relevant to note, that there was a structural shift in the composition of migrant flows between 2004 and If after the enlargement the low-skilled labour dominated in the flows of labour from the eastern Member States, then around 2009 the share of high-skilled labour had started to expand (Kahanec et al., 2009). The worrisome news is that the migrants from the new Member States are frequently employed in jobs for which they are overqualified. This has twofold consequences. First, it effectively puts well-educated people in the category of low-skilled labour, and with time they lose their precious skills; second, their wage level, as a rule, is well below natives with similar educational and work experience. This situation clearly points to problems of cross-border transferability of human skills. Lower returns to foreign education and work experience, combined with limited mastery of host country s language, are to explain higher over-qualification rate for immigrants relative to nativeborn workers (Carrera et al., 2015). The impact of immigration on society s age structure in the destination countries The societies and policy-makers in Europe will face key labour market challenges in the years to come relating to economic recovery, ageing populations and shrinking workforce. According to Eurostat projections, without immigration, by 2030, the number of people aged will shrink by 30 million in the EU. In Germany alone the number 27

28 of economically active people will contract by 9 million. Certainly, immigration attenuates this demographic decline to some extent. The immigrants from the EU13 are mostly young people, and, as it has been shown, there is still some potential of further immigration from the EU13, in particular from Romania, Bulgaria, and Croatia. However, young people from the EU13 will not save the old Europe from aging, as many of EU13 countries will soon start aging even faster than EU15 societies. Eurostat (2014) also finds that international migration may be used as a tool to solve specific labour market shortages. However, migration alone will almost certainly not reverse the ongoing trend of population ageing experienced in many parts of the EU. The existing EU law regulating the flows of people cross-border European borders be it EU citizens or nationals of third-countries has been developed under the conditions of demographic boom and overpopulation and has become outdated. The stereotypes from the 1980 s and 1990 s still have strong influence on people s perception about immigrants. Yet, as noted by Bartha et al. (2015), Europe may have a surfeit of low- or unskilled workers at present, but once demographic transitions begin to bite, the position could change quite rapidly. The impact on fiscal and welfare systems of the destination countries OECD (2014) finds that in European countries immigrants in general are neither a burden to the public purse nor are they a panacea for addressing fiscal challenges. According to this study, the fiscal position of immigrant households is shaped: (1) by the design of tax and benefit systems and (2) by differences in the composition of the migrant population in terms of age and migrant-entry category (labour, consumer, family member). By working and consuming, immigrants contribute to recipient state s revenues and social insurance foundations. Immigration also helps to save public resources on nurturing and educating skilled people, as these costs are born by sending countries. Not least important is immigrants contribution to local markets, as a large part of their remuneration is spent on local goods and services, thus boosting local aggregate demand. In countries with recent labour immigration, immigrants have a much more favourable fiscal position than in countries where humanitarian migrants account for a significant part of immigrant population. Thus, language problems, psychological trauma, employer s discriminatory practices, or legislative and institutional factors that inhibit adjustment and economic outcomes of immigrants may make them more dependent on the local welfare system (Borjas, 1999; Brücker et al., 2002, as quoted in Kahanec et al., 2009). It is possible to argue that extensive inflows of immigrants can push up prices for local goods and properties, and test the limits of public services. In the case of intra-eu 28

29 migration, this is true only to a certain extent, and concerns only a small number of municipalities in popular destination countries suffering from overconcentration of immigrants. There is also counterfactual evidence showing that immigration has helped to revive a number of depressed areas within the EU15 countries. None of the studies on intra-eu migration has been able to deliver evidence on the socalled welfare tourism. Just on the contrary, research shows how social systems have gained from immigration in the receiving countries. This is happening because immigrants: (1) are younger than the native population in the receiving countries and will not need social services that soon; (2) they have higher participation rate in the labour force than local nationals have (in 2013, 67.7% versus 64.6%) thus being less dependant on the social security. Indeed, the scale of contributions of immigrants to social funds is lower than that of native workers; however, because immigrants are mostly occupied in low-skilled jobs they receive less in salaries, hence their lower capacity to contribute to social funds. Alas, the share of unemployment among the EU nationals working in another Member State tends to be larger than among the local workforce (in 2013, 12.5% against 9.8%), yet it is much lower than among third-country nationals (21.3%). Notwithstanding higher unemployment, the surveys of migrants indicate that people migrate out of search of better paid jobs, not because of social benefits. People want to improve their chances, and the higher rate of unemployment among immigrants is explained by problems of finding an appropriate job. After all, immigrants cannot claim welfare benefits upon their arrival in the destination country they have to spend some time working before they qualify for social benefits. 29

30 Part V. The impact on the sending countries Theoretical primer The simple labour supply model of migration suggests that the sending country should suffer a decline in total production because of people s outflows but see a rise in per capita income after immigrants depart. Thus, the net result should be increase in the real income of workers that remain behind and a decrease in earnings for the owners of the economy s other productive factors (land, capital, raw materials) whose marginal product declines with the departure of the immigrants. What is more, the movement of people shifts the demand from the sending to destination economy (Bodvarsson and van den Berg, 2009), with corresponding adjustments in capital stocks (on a condition that capital flows are flexible). The full picture, however, is much more complex, and, under certain conditions, the sending countries need not suffer from economic decline. The heterogeneity of labour markets, return migration and technology flows between sending and destination countries may have considerable implications, too. Hence, following Bodvarsson and van den Berg (2009), the issues to be addressed in respect to sending countries are the following: 1) The impact on sending country s wages and domestic demand; 2) The positive and negative externalities associated with the decline in population; 3) The effects of remittances; 4) The effects of the loss of innovative resources. The general rule is that the departure of people makes labour resources scarce in the source country, leading to wage increase. However, as pointed out by Kahanec et al. (2009), this wage increase will not be enjoyed by everyone and depends on who departs the country. Thus, the departure of high-skilled people increases the wages in the highskilled market, and reduces the wages (or increases unemployment) in the low-skilled sector because of complementarity effect and falling domestic demand. And vice versus, emigration of low-skilled workers causes the wages in the low-skilled market to increase (or decrease of unemployment) with negative effects on wages in the high-skilled sector. Many immigrants send a portion of their income earned abroad back to the sending country to sustain family members left behind, to repay debts or to store wealth for their own future return to their native country. In fact, remittances shift income, and thus labour demand, from the destination to the source countries. In fact, concludes Bodvarsson and van den Berg (2009), if immigrants remit a substantial portion of their foreign earnings (it is even feasible that foreign earnings compensate all losses related to departure of people), then the total demand for labour in the sending country could actually increase, even though people leave the country. Moreover, Bodvarsson and van 30

31 den Berg (2009) observe that immigrant remittances are not large enough to substantially reduce the demand for labour in high-income destination countries (high-income countries are costly for living, therefore, immigrants are forced to spend a large portion of their income on daily needs). Rather, these remittances make possible the case where immigration generates income gains in both the destination and source countries. However, the welfare effect from remittances experienced by a sending country is dependent on how remittances are spent. The point is that remittances can raise the longterm welfare by much more if they are not spent on consumption, but are used for acquisition of consumer durables, housing improvement or, most importantly, for higher education attainment. In the worst case, remittances may create significant moral hazard both for individuals and also for governments. Thus, while boosting purchasing power of individuals, they reduce work incentives. Likewise, if remittances are huge, governments may succumb to moral hazard and delay steps to be taken to restructure the economy needed, e.g., to limit people s outflow. The real impact of remittances, though, is very difficult to establish because of inaccuracy of data on these remittances. The departure of people can have both positive and negative externalities in the source country. First, immigration may serve as a safety valve when employment opportunities are lacking in the sending economy. In more cynical terms through emigration an economy is shedding its spare labour. Second, the departure of unemployed or under-employed reduces the need for housing and makes other infrastructure more accessible, including social security bonuses. Third, on the negative side, economies of scale may be lost if population shrinks, raising costs and suppressing per capita real income. Fourth, emigration may lead to lower wage level. Finally, due to emigration there may no longer be a market for certain products and infrastructure (Bodvarsson and van den Berg, 2009). A specific negative externality is related to the fact that industrial and innovation activities tend to be subject to economies of scale. Departure of people reduces the economies of scale in the sending country and depletes the stock of talented people. The economies of scale, therefore, exaggerate income differences that may already exist for other reasons, thus further stimulating immigration. Innovation is a costly and human capital intense activity, therefore, agglomeration of talented people in high-tech centres is critical for success. Immigration may facilitate and also hinder the spread of technology at the same time. First, immigration will reduce capital and human resources in the sending country and this may delay the adoption of new technologies. Second, innovative activities in the centre countries generate new technologies that potentially become available to everyone. Finally, to the extent that immigrants build networks through which new ideas and technologies pass, immigration driven by innovative agglomeration can both increase innovation and speed the spread of 31

32 innovative new ideas and technologies. In fact, the transfer of knowledge from technologically more advanced countries to less advanced countries requires entrepreneurial and skilled people on the both ends of transfer. Certainly, immigration helps build networks and channels that facilitate the flow of technology. However, these networks may be of little use, if immigration leads to the departure of a considerable fraction of the source country s most skilful workers, who are critical to applying these newly transferred technologies. Bodvarsson and van den Berg (2009) find that the relationship between immigration and investment flows is complex. Traditional trade theory suggests that capital should move to labour abundant countries, which are the sending countries. However, this is not always the case, and the lack of investment mobility is attributed to institutional failures (information asymmetry, high disposition to risk aversion, etc.). Hence, the immigrants through ethnic and cultural networks can mitigate these institutional failures which prevent capital from crossing the borders. What is more, immigration stimulates international investment by reducing the transaction costs associated with language, business practices, and culture. And, last but not least important, there is evidence that immigration may foster international trade between destination and sending countries, because: (1) immigrants create trade networks; (2) they stimulate presence of multinational companies from the destination in source countries; and (3) they induce governments in both the source and destination countries to reduce trade barriers. A particular problem for source countries is brain drain. When people emigrate, they take with them not only their own labour, but also their accumulated human capital. The problem is that the skilled emigrants take with them the education and training that was, at least in part, paid for by people remaining in the source country. In fact, the departure of highly educated immigrants carries serious costs for source countries, and, bizarrely, by paying for educating skilled migrant people, capital scarce developing countries may end up subsidising developed countries. However, it has to be recognised that immigrants leave not only because they see better opportunities elsewhere, but also because they see no opportunities at home. At last, the overseas job opportunities may provide strong incentive for people in the sending countries to acquire better education, and the highly educated are likely to remit more income back to the sending economies. In reality, though, the remitting capacity depends on migrant s ability to apply her skills abroad, and, as observed by a number of researchers, highly skilled immigrants do no always use their skills abroad. Moreover, there may even be huge differences in the earnings of skilled immigrants coming from different sending countries, and, as attested by empirical evidence, immigrants are more likely to fully exploit their talents and skills, when the same language is spoken in the sending and destination countries, reflecting the importance of social capital for a person s economic success (Bodvarsson and van den Berg, 2009). 32

33 Empirical evidence In line with theoretical literature, migration can affect the path of development of sending countries in a number of possible ways. Some possible paths (IMF, 2006): 1. Convergence occurs in terms of income of production factors, as capital flows toward the East and people move to the West until capital-to-labour ratios are equalized across countries (naive neo-classical model); 2. If capital is perfectly mobile and labour is not, capital from the West relocates instantly to the East, and convergence takes place without any labour movement between the East and West; 3. More realistic is that capital is not perfectly mobile and labour is able to relocate. The more that capital is immobile or the more that labour is mobile, the greater the labour shedding toward the West. Nevertheless, if technologies are identical in both regions, convergence will eventually take place, with capital-to-labour and other parameters becoming equal (EU pre-enlargement situation); 4. Human capital has increasing returns to scale and innovation provides new and more productive investment opportunities. This leads to accumulation of highlyskilled or highly-educated workers in some specific regions (at the expense of other regions). The result is permanent divergence and technological underdevelopment of the East (endogenous growth theory); 5. Finally, the changes in capital and labour locations are determined by interplay of two opposing forces agglomeration and congestion. Agglomeration tends to pull capital and labour (together) to central locations. At the same time, the agglomeration tendencies are moderated by poor business climates, congestion and other decreases in the quality of life in these same centres. Which force is stronger for a given region is a complex function of technology and tastes, historical accident, government policies, and institutions. As a result, agglomeration forces could continue to keep capital in the West once trade barriers are removed. Or, if the West is overcrowded, the East might look more attractive for capital and labour inflows (low taxes, ease of doing business, and good public infrastructure could help in this respect) (new economic geography theory). Which of these scenarios will materialise in the case of the new EU Member States? Will convergence in terms of capital-to-labour continue despite emigration? If yes, how long will it take to achieve this convergence? Or, how high is the risk of technological underdevelopment and, hence, permanent divergence? Historically the EU has had several high migration episodes or waves. With some oversimplification, looking back to the last decade, it is possible to discern the following three experiences: 33

34 1. Pre-enlargement experience; 2. Post-2004 experience; 3. Post-crisis experience. Before 2004 enlargement, the main geographical direction in intra-eu migration was from the South to North. The industrialisation of Western countries in the 1960 s led to shortage of poorly paid work, and Southern European (Italian, Spanish, Portuguese, and Greek) migrants moved to the North and West to find job in construction, manufacturing, and service industries. Luckily, successful economic development in the then sending countries led to equalization of income levels between the South and North and substantial return migration in the 1970 s. In the 1990 s, the median intra-eu migrant was already a highly qualified professional, and third-country migrants had substituted Europeans in low-paid jobs. After 2004, when the EU began to embrace the countries from Central and Eastern Europe, the number of EU citizens living in another EU Member State has considerably grown and it continues to grow though at a slower speed than during the first years after the enlargement. This expansion of expatriate communities was possible on the account of natives of the new Member States who succumbed to the temptation to try their luck in the West. Galgóczi and Leschke treat this post-2004 mobility as a historically new phenomenon in the EU because of its multifaceted character (coexistence of different forms of mobility: commuting, short-term, circular, permanent migration, (bogus) selfemployment and posted workers) and because of dominance of rather highly educated people in this migration. The economic crisis of had not been asymmetric between the Eastern and Western EU Member States. Instead, it had affected a number of countries in both camps of Member States. From the EU12, the most severely hit countries were the three Baltic countries, Hungary, and Slovenia, while among the EU15 Greece, Ireland, Portugal, and Spain. As a result: (1) the overall stock of the foreign EU10 population in EU15 countries has continued to grow, albeit at a slower pace, and mostly on the account of most affected countries; (2) because of economic hardship, return migration has intensified from such earlier popular destination countries as Spain and Ireland, with Poland, Romania, and Lithuania being the main beneficiaries; (3) the decline in Southern EU Member States has led to renewal of south-to-north migration, although the increase of migrants has remained limited compared to the overall unemployed population in Southern Member States. Data show that the new Member States have experienced significant outflows of young and relatively well-trained people. For example, between 2003 and 2012, Lithuania had lost 9.8% of its population and Latvia 7.8% on account of outmigration; of which 3/4 were below age of 35 and 1/3 had tertiary education. Bulgaria and Romania are another 34

35 two major sending countries. The significant outflows of predominantly young and skilled individuals will have a negative long-run impact on these countries economic growth, demographic development, and public finances. The question is to what extent? Will these countries find other ways on how to ensure economic development despite the negative impact resulting from people s outflows and demographic decline? The departure of young people adds to demographic instability in the new Member States. Emigration, in fact, quickens the ageing process which has been fast even without emigration. The only good news is that the population in the new Member States, in general, is younger than in the old Member States; however, as new Member States are less populous and young people dominate among emigrants, the poll of potential emigrants is running low fairy quickly, and, in the case of some new Member States, emigration has achieved epic scale, as they are about to lose a complete generation of young workers who are under 35 today. According to simulation results by Kahanec et al. (2009), 16 due to emigration between 2004 and 2008, the EU8 labour force would shrink by 1.16% leading to a loss of 0.52% of GDP in the short-term and 1.1% in the long-term. Particularly affected appeared to be Poland (1.77, 0.88, and 1.94%, respectively), Slovakia (1.34, 0.53, and 1.51%), and Lithuania (1.14, 0.55, and 1.15%). Kahanec et al. (2009) also found that in the short run GDP per capita would increase by 0.65% in the sending countries due to higher capital endowment per worker, while the long-run effects would be neutral. Moreover, after adjustment of capital stocks, the long-term effect on unemployment and average wages would be insignificant either. These results point to three important things: first, the risk of unbalanced wage increases caused by shortage of labour in the sending countries is very low in the long-term, as investments will adjust to the new labour supply situation. Second, contrary to the expectation that a decrease in labour force will push wages up and higher wages will induce higher labour participation (e.g., see Veselková et al., 2014), this positive effect seems feasible only in the short term at best. Third, countries with high outward migration might see their overall capital stock shrinking unless specific action is taken to enhance technological excellence. Holland et al. (2011) modelling results are even less pleasing. The sending countries may not only loose some growth potential due to emigration, in particular Lithuania (-6%), Romania (-10.6%), Bulgaria (-5.4%), and Latvia (-3.3%), but also the income per capita. This is true, according to Holland et al., for almost all EU10 countries, the only exceptions being Poland, Hungary, and Czech Republic. 16 These simulation results, however, should be treated with some caution, as the calculations were based on official data available at that time, which later turned out to be severely underestimating the true scale of emigration, at least in the case of a few countries (e.g., Latvia), as was revealed by 2011 censuses. 35

36 Yet, the picture may not be as black as presented above. Emigrant workers do remit a large portion of their income back home. According to World Bank data, 17 the biggest beneficiaries in terms of received remittances were Lithuania (4.5% of GDP in 2013), Malta (3.6%), Hungary (3.2%) and Bulgaria (3.1%). Interestingly, among remittances recipients are also countries from the EU15 with equally high shares, namely, Luxembourg (3.0%) and Belgium (2.1%). Kahanec et al. (2009) notes that remittances impact on the economic development so far has been difficult to document, but agree that remittances play a positive role in the economic development of the new Member States inasmuch as they support aggregate demand and investment in education and business activities. Kahanec et al. (2009) also conclude that further investigation is required on how these remittances affect the decisions of those who stay behind, and on their aggregate fiscal impact. Holland et al. (2011) find that while remittances can partially offset the negative effects on growth in the short- to medium-term, they cannot fully address the loss of labour on capacity output in the long-term. The economic crisis of induced a new trend in migration flows, namely, return migration. Many Romanians, Bulgarians, and Lithuanians have decided to return home when their fortunes worsened in the receiving countries like Spain, Italy, and Ireland. 18 In the case of Romania, Lithuania, Latvia, and Portugal, returnees represented more than 60% of total immigrants in these countries in This is a welcome development, as their language skills, expanded professional networks and new skills learned abroad would enhance the productivity of their home economy. However, the overall impact might not be that positive due to two reasons. Firstly, it is noted in several studies that highly-skilled workers are not that prone to return to their home countries. Those who return are mostly low-skilled (e.g., Hazans, 2011). Secondly, even if there are some highly skilled returnees, some evidence suggests that so far they have brought little additional skills to their home countries, because they have lost their skills while working abroad in low-skilled occupations (Kacmarczyk, 2011). This problem of over-education and skills-jobs mismatch in the case of EU10 immigrant employment has been underlined by many studies (e.g., Kahanec et al., 2009; European Commission, 2008; Galgóczi and Leschke, 2012). A study by Kaczmarczyk (2013) reveals that in Poland the return migrants were more exposed to the risk of unemployment than non-migrants. In the 17 The World Bank provides the most complete set of statistics on remittances to the EU countries. Worker s remittances are defined as the sum of three components: (a) workers remittances recorded under the heading current transfers in the current account of the country s balance of payments; (b) compensation of employees which includes wages, salaries, and other benefits of border, seasonal, and other non-resident workers and which are recorded under the income subcategory of the current account; and (c) migrants transfers which are reported under capital transfers in the capital account. 18 Galgóczi and Leschke (2012) attribute these changing fortunes to the peculiarities of employment of the EU10 immigrants in the EU15. Galgóczi and Leschke contend that EU10 migrants were frequently engaged in irregular and non-standard forms of employment and had concentrated in sectors disproportionally affected by the slump, e.g., construction and manufacturing. In consequence, the immigrants from the EU10 were hit harder and they acted as labour market buffers at least to some extent. 36

37 meantime, Hazans (2011) points out that, in Latvia s case, the returnees have better employment performance (see Part VI of this paper). A particular feature of the post-enlargement intra-eu migration is high educational attainment of the EU10 migrant population (Galgóczi and Leschke, 2012). Before the crisis, EU12 workers were significantly overrepresented in the medium-skilled category; during the crisis the shares of both high and low-skilled migrants were rising. Despite the relatively high education level, the EU13 immigrants are concentrated in low-skilled sectors (hospitality and catering, manufacturing, construction, food processing, and agriculture). Allegedly, the foreign language proficiency which is higher among medium and higher education holders is part of the explanation why relatively well-educated prevail in migration flows from the EU13. Another explanation is huge wage difference between the new and old Member States, which makes low-skilled jobs in the Western countries more attractive than high-skilled jobs in the Eastern countries. Anyhow, to the extent that immigrants from the EU12 (now the EU13) possess higher qualifications than they need for the work they undertake, human-capital investment of home countries will have lower returns. According to Bartha (2015) and Galgóczi and Leschke (2012), the post-enlargement East-West labour mobility has not contributed to better human-capital allocation, even worse - it has led to waste of human resources. The problem is not just that of classic brain drain, but also of brain squandering. What is good though is that, following the European Commission s evidence, there is a tendency of increasing enrolment rates for tertiary education in the new EU Member States, which compensate to some extent for the outflow of skilled labour (European Commission, 2008). 37

38 Part VI. The case of Latvia The economic impact of migration on Latvia s economy has not bee studied extensively so far. However, academic researchers have started to gather data and to construct knowledge on the patterns of people s migration from and to Latvia. The Latvian government has become more attentive to the problem of outmigration and eventual shortage of labour after the 2011 national census. In response to shocking figures of emigration and demographic decline, revealed by the census (see Part I on statistics in this paper), the government set among its priorities strengthening the capacity of public administration to communicate with emigrees and support to the Latvian diaspora communities abroad. Moreover, government has supported the establishment of the Centre for Diaspora and Migration Research under the University of Latvia, which became functional in The government has also commissioned a wide-ranging scientific project titled The Latvian emigrees communities: national identity, transnational relations and diaspora policy. 19 During this project, which was started in January 2014 and is still ongoing, people of Latvian decent were interviewed in 118 countries, and it aims at delivering comprehensive and generalizable data on Latvian emigrant communities. The scale of emigration from Latvia and its impact on the age structure of Latvia s population From 2000 until today, around 332 thousand people have emigrated from Latvia, while only 94 thousand have immigrated to Latvia (see Figures 6 and 7). Net migration has thus left Latvia with a deficit of 248 thousand people that corresponds to more than 12% of the Latvian population and to at least 14% of economically active people (15-74 years old). Among those who have emigrated, 3/4 were under 35 at the moment of departure. According to pioneering research by Professor Mihails Hazans on the Latvian emigration of modern-times, the patterns of emigration have changed between 2000 and 2010, and it is possible to discern three emigration waves. The first wave of around 50 thousand people took place in ; it had wide geographic dispersion and was dominated by highly educated and entrepreneurial people. The second wave happened after Latvia s accession to the EU, involved 70 thousand people, and was concentrated on Ireland, the United Kingdom, and Sweden. This wave had a larger share of people with lower educational attainment, as the costs of emigration had substantially diminished compared to the pre-accession period. The third wave was linked to the economic crisis in Latvia in and high unemployment (above 30% among low-skilled workers in 2009). Today emigration still continues, albeit at a much lower rate than in 2008 and In contrast to earlier emigrations, the emigrants of 2008 and later years are dominated by 19 For more details see 38

39 high-skilled ones, they have manifest tendency to leave Latvia forever, and they depart together with their families (Hazans, 2011). At the moment of the accession to the EU, the population of Latvia was considerably younger in comparison to the old EU Member States. However, due to low fertility rate in the 1990 s the generational replacement has become much slower in Latvia than in many other parts of the EU, and, compounded with emigration of young people, Latvia s population is not only quickly ageing, but also shrinking. From 2004 to 2013, the number of people in Latvia has decreased from 2.28 million in 2004 to 2.00 million in 2013, and the projections show that it will diminish further to 1.63 million until On the regional level, the trends are even more disquieting. The Riga (capital) region comprises more than 1/3 of Latvia s population and it has been a traditional destination for young people from other parts of Latvia. Consequently, because of external and internal migration and demographic decline, between 2004 and 2013, the number of people has decreased by 18% in Latgale region, 15% in Vidzeme region and 14% in Kurzeme region. At the same time the old-age dependency ratio has become much higher in these regions, and there are already a few counties where retired people comprise 1/3 of local population. The emigration potential from Latvia remains very high. Although gradually diminishing, the gap in income levels between Latvia and the EU15 countries remains considerable. Thus, in 2014, the minimum statutory wage in Latvia was 360 euro, while in the EU15 it was around 1400 euro. Moreover, almost every Latvian household has a family member or a good friend abroad, which makes departure for those who stayed at home much easier, as family members and friends can help them to find appropriate job abroad. Hazans (2011) describes this as an effective migrant network at work that will continue to exert its influence for years to come. The impact of migration on the labour market in Latvia, skills formation and equality Latvia s economic fortune after accession to the EU has been very unstable. During the first four years, from 2004 to 2007, Latvia s economy was booming, with economy growing at a speed of close to 30% in nominal terms shortly before crush. This boom was fuelled my foreign capital inflows, and emigration and foreign labour remittances were a significant factor adding to growing internal and external economic disbalance. During the boom years, unemployment dropped to 5.3%, a record low level for Latvia. Later, from 2008 to 2010, Latvia suffered from steep economic slump. The global financial crush of 2008 had extremely nasty repercussions on Latvia. The bubble burst, turning Latvia from a growth champion into a nose-diving Icarus. In two year s time, between the 3 rd quarter of 2008 and 2010, from peak to trough, Latvia lost 28% of its economy, representing the deepest crush ever experienced by a developed economy during peace 39

40 time. Consequently, unemployment soared, reaching its apogee at the level of 21.3% in the first quarter of Unemployment was particularly high 41% among males of age Without emigration, the economic hardship could have been even more severe. Emigration acted as a safety valve for many who had lost their job and income, but had liabilities against credit institutions or had difficulty to sustain their families. Hazans (2011) notes an interesting fact a significant majority of those who left Latvia before and after the crisis were employed before departure, 87% and 84%, respectively. In later years, in 2011, this rate dropped to 52% (Hazans, 2013). This observation confirms the assumption that wage difference matters more than difficulties of employment when people take decision on departure. The good news is that the departure of economically active people opened employment opportunities for those, who were inactive this far, thus fostering labour participation. The educational attainment of the Latvian population has been increasing despite emigration. This development seems to confirm the hypothesis on emigration s positive impact on educational attainment of those who stay behind, which is explained by a mere fact that chances of better paid job abroad motivate people to advance their education. Thus, according to statistics, the share of Latvian people with upper secondary or tertiary education attainment increased from 75.9% in 2000 to 83.4% in 2013 among the years old, while the share of those with higher education has increased from 15% to 27% in the same period. The number of students enrolled in higher educational institutions was increasing between 2000 and During the crisis and later this enrolment rate has been declining though. In 2014, according to labour force survey, 19% of Latvian emigrants had higher educational attainment. The research of Hazans (2011) puts this share above 30%. What is more, Hazans reports on high shares of those emigrants who do not have employment abroad according to their skills. Among low-skilled emigrants these incidents are very frequent, close to 80%, while in the case of medium-skilled emigrants it is around 50% and high-skilled 40%. These findings broadly match the results from the ongoing Latvian emigrants research project (Mieriņa, 2015). Latvia s Ministry of Economy (Ekonomikas ministrija, 2015) assesses that, in order to sustain economic growth in margins of 4 6% annually, Latvia will have to find extra high-skilled and middle-skilled labour (43.2 and 15.2 thousands, respectively, until 2020) and shed spare low-skilled labour (9.3 thousands until 2020). The areas where most of extra labour demand will come from are processing industry, trade, and commercial services. The Ministry does not see problems with labour supply demand mismatch, as according to its calculations labour supply will exceed labour demand by 6% in 2020 and by 5% in 2030 the shrinking population would be compensated through substantial return migration and larger labour participation. At the same time, the 40

41 Ministry points to developing substantial labour-skills imbalances. Thus, by 2020, there will be 20 thousand specialists in excess in humanitarian and social fields, and shortage of 20 thousand engineers and specialists of IT and natural sciences. The Ministry also finds that a too large share of young people in Latvia have no specific profession or are low-skilled, which is worrisome as demand for such labour will substantially diminish. Unfortunately, in the case of Latvia, emigration has not much contributed to greater equality of income distribution between different groups of society. 20 Initially, between 2004 and 2008, the GINI coefficient fell from 51.2 to 47.2 indicating improvement in income equality. However, during the crisis the improvement withered away and GINI jumped to 52.3 in Henceforth, the situation has slightly improved, as GINI is gradually going down and was 50.3 in In fact, Latvia along with Bulgaria, Greece, and Romania are the most unequal countries in the EU, with the income level of the richest quintile of population exceeding the income level of the poorest quintile by more than seven times. Immigration to Latvia has been very low so far, in the margins of 0.4% of Latvia s population. Among immigrants a considerable majority are Latvian nationals returning home from the Western countries (83.8% in 2012 and 65% in 2013). Nationals from other EU28 countries make only 4% of immigrants to Latvia and, most likely, are from the neighbouring Estonia and Lithuania. As far as the employment prospects of returnees are concerned, Hazans (2011) reports that the employment rate among the returnees is much higher (around 66%) than among Latvian population without foreign experience. Likewise, the share of business people and the self-employed is higher among returnees than among the Latvian population without foreign experience (6 and 4% respectively). The preliminary data of the research project, The Latvian emigrees communities: national identity, transnational relations and diaspora policy, point to following patterns (Mieriņa, 2015): 1) The Latvian emigrants, on average, appeared to be more satisfied with life than people in Latvia and in their destination countries; 2) Before departure, around 80% had experienced some sort of financial difficulties, and around 45% were under severe financial stress; 3) Close to 70% of Latvian emigrants in 2014 were employed or self-employed during the first year of their stay abroad; 4) On average, 39% of Latvian emigrants had a job abroad matching their skills. This rate was higher among the emigrants with tertiary education 57%. 20 Theoretically, emigration can lead to more equal distribution of income if those who emigrate are low-skilled., as their departure puts downward pressure on the wages of those high-skilled staying behind. Plus, earnings abroad increase the household budget of departing low-skilled people. 41

42 Compared to their last employment in Latvia, 24% had a job at lower skill level, while 7% a job with higher skill level; 5) 7% of emigrees had increased their skill level while living abroad; 6) Most of the Latvian emigrants had kept close ties with Latvia: 40% have real property in Latvia and another 40% admit that they could buy one in some future. Moreover, 25% of emigrees think that they could establish business links with Latvia after a while. 7) 18% of surveyed emigrees had a shared residence between Latvia and their destination country, and 24% felt that they were not belonging to a particular place, but were feeling like citizens of the World; 8) From those who have emigrated, only 6% planned to return to Latvia in 6 months time. 12% thought that they will return in 5 years time, and another 14% after their retirement. 40% admitted that they could consider return, while 30% declared that they would never return; 9) Those, who planned to return, were either under 30 years, or males with low level of skills, or with a job for which they were overqualified, or used to live in Latvia s countryside or small towns before departure, or had left family in Latvia; 10) Friends, relatives and longing for home were given as most significant motives for return; 11) On the other hand, the lack of appropriate work in Latvia, low level of social security, and shortage of professional opportunities were most frequently cited obstacles for return to Latvia; 12) 40% of those who returned to Latvia had emigrated again. The impact on Latvia s economic growth and technological development Emigration did act as a catalyser of economic overheating of Latvia s economy in During the crisis of , on the contrary, emigration acted as a safety valve both for people and the government. For individuals it was a chance to sustain their income level despite a deep economic slump in Latvia. For the government, emigration eased public tensions and helped to avoid a major popular revolt against the political elite, and it also helped to save money on unemployment and other social benefits, thus reducing pressure on already heavily stressed fiscal situation. Moreover, labour remittances were considerable and boosted aggregate demand during slump. The Bank of Latvia puts the total amount of these remittances between 2 to 3% of Latvia s GDP. There is no specific scheme on how to channel these remittances into productive sectors of Latvia s economy; therefore, one can assume that a large part of this money is spent on consumption and that government benefits mostly through larger intakes from the valueadded tax. The awareness of the risks stemming from emigration is not equally shared among the Latvian economists, and views differ on emigration s impact on Latvia s future economic 42

43 development. For example, according to Hazans (2013) and other academic researchers, emigration has reached levels that pose threat to the reproduction of the Latvian population, the country s economic development and sustainability of its social security system. Meanwhile, the monetary economists from the Bank of Latvia claim (Melihovs, 2014) that net migration will be irrelevant for Latvia s population dynamics in the medium term, as emigration will recede with the improvements in the economic performance of Latvia and with convergence of incomes with the more developed EU Member States. It seems that Latvia s monetary economists underestimate the risks of shrinking population and market. The literature on people s migration underlines that the migratory process should be analyzed as a long-term social process with its own dynamics (Carrera et al., 2015). As such, migration will not succumb that easily to dynamics of economic development. Indeed, largely permanent departure of the younger and more educated workers may indeed be costly for those who stay (Blanchard, 2013), as emigration entails loss of resources invested in upbringing these people. The Ministry of Economy of Latvia predicts that the demographic burden will rise by 21% from now on until 2030, from 550 to 665 per thousand employed. All in all, Latvia s economy is suffering from emigration which has caused a considerable shrinkage of labour force, and labour remittances can only partially compensate the negative impact on growth, finds Hazans (2013). According to Holland et al. (2011), based on numbers of emigrants until 2009, Latvia will lose 3.3% from its growth potential because of emigration. The GDP per capita will also diminish, by 0.7%. Other countries destined to substantially lose from emigration are Latvian neighbours Lithuania (-6%) and Estonia (-3.0%). Hazans (2013) is even more pessimistic, and estimates that the impact has been larger taking into account the scale, age and educational level of departing people from Latvia, and puts this impact at -9%. Future technological advancement of Latvia is a teasing issue. Will Latvia be able to converge with the more developed EU Member States or is she doomed to permanent convergence? Indeed causing shrinkage of population and hence domestic demand and labour pool, emigration in combination with demographic decline discourages investment, both foreign and domestic. Moreover, today, at least hypothetically, it is much cheaper to move people to the West than to build a new factory in the East. The statistics indicate that in Latvia gross capital formation has fluctuated around 23% of GDP between 2011 and 2014 (IMF, 2014). As far as foreign direct investment is concerned, the inflows are diminishing, from 5.2% of GDP in 2011 to 1.5% in In the meantime, the accumulated stock of foreign direct investment has increased though, from 45% of GDP in 2010 to 49.9% in 2014 (see Figure 8). Latvia became the Member State of Eurozone as of 2014, and inflows of FDI were seen as one of the attractions for the euro adoption by the Latvian public. However, these expectations have not yet 43

44 materialised. Certainly, Latvia s proximity to Russia, the Ukrainian conflict and sanctions against Russia may have reduced the investors propensity to invest in Latvia. However, the lack of business opportunities due to shrinking domestic market may have had a role, too. In this situation the technological advancement becomes crucial. In order to survive, businesses in Latvia will have to adjust to higher labour costs through innovation and productivity enhancement measures (see Figure 9). There are already some good signs. The share of people with tertiary education and employed in science and technology is increasing among the economically active population in Latvia, and has reached 18.7% (in Lithuania 23.7 and Estonia 20.8%). Also the share of high-tech exports is growing from 4.6% in 2007 to 8% in 2013 of total exports. Meanwhile, the number of patents has been decreasing since 2008 and stood at patents per million people in 2012 (the corresponding EU28 reference was patents in 2012). It is quite possible that the economic volatility inherent to small economies discourages companies from investing in research and development activities in these countries, and companies outsource new product development to the centres of excellence abroad (Kattel, 2010). If this is true, it will not be possible to resolve the problem of capital deepening at domestic level. Instead, action at the EU level will have to be taken, ensuring fair division of research work across all EU Member States. 44

45 Figure 6. Population by labour status in Latvia, thds Figure 7. Long-term migration in Latvia, thds Source: Central Statistical Bureau of Latvia [NB0010] Source: Central Statistical Bureau of Latvia [IB0010] Figure 8. FDI and GDP in Latvia, per capita, thds Figure 9. Labour productivity and ULC index in Latvia (2010=100) Source: Eurostat [nama_gdp_c], [bop_ext_intpos] and Bank of Latvia (2015) Source: Eurostat [nama_10_lp_ulc] 45