Will the EU-15s current social welfare systems induce migration from the new member states?

Contents

List of Tables and Figures

List of Abbreviations

Prologue

1. Introduction
1.1. The Free Movement of Workers in the EU – Past, Present and Future

2. Migration into the EU-15 – Theory and Empirical Evidence
2.1. The Theories of International Migration.
2.2. Empirical Evidence about Migration into the EU-15
2.3. Is There Migration Diversion after the Eastern Enlargement?

3. The Economic and Fiscal Impact of Immigration on the German State – Boon or Burden?
3.1. Do the EU-15 Really Need Immigration?
3.2. The Economic Impact of Immigrants
3.2.1. Who Wins and Who Loses from International Migration?
3.2.2. Empirical Evidence about the Impact of Immigration on Domestic Wages
3.2.3. Empirical Evidence about the Impact of Immigration on Domestic Unemployment
3.3. The Fiscal Impact of Immigrants
3.3.1. Studies Confirming the “immigrant-as-a-burden” Thesis
3.3.2. Studies Confirming the “immigrant-as-a-net-contributor” Thesis

4. The “Welfare Magnets” Hypothesis – Hollow Threat or Imminent Danger?
4.1. The Different Welfare States in Europe
4.2. The Social Assistance in Germany and the UK – a Head-to-Head Comparison
4.3. The “Welfare Magnets” Hypothesis – Basic Theory
4.4. The “Welfare Magnets” Hypothesis – Empirical Evidence
4.4.1. USA
4.4.2. EU-15
4.4.3. Germany

5. Harmonizing the EU-15’s Social Policies – Panacea or Pandora’s Box?
5.1. Pro-Harmonization Arguments
5.2. Contra-Harmonization Arguments
5.3. Other Migration Policy Options

6. Conclusion

References

Appendix

List of Tables and Figures

illustration not visible in this excerpt

List of Abbreviations

illustration not visible in this excerpt

[Prologue]

Excerpt from “Allgemeine Bevölkerungsumfrage der Sozialwissenschaften“

(ALLBUS) 2004

illustration not visible in this excerpt

Source: ALLBUS 2004

CHAPTER 1^[1]

Introduction

2004 was a momentous year for the European Union (EU). On 1 May the most recent act of Enlargement was finally fulfilled. Ten New Member States from Central and Eastern Europe joined the club and their residents officially became European citizens. The moment was all the more special for eight of the ten newcomers, as they were ruled with an iron fist behind the Iron Curtain for more than forty years. The accession represented the ultimate accolade for their arduous transition from central planned to market economies. The long history of an expanding united Europe thereby reached another peak – regarding cultural integrity, political influence and economic potency. Europhiles praised (lauded, acclaimed???) the EU’s greatest achievement so far.

On the other hand, the EU Enlargement also encountered various feelings of skepticism or even disapproval throughout the population of Western Europe. Many people feared that “fortress Europe” would not hold and the walls of the Western social welfare states would be toppled (breached???) by millions of poor immigrants willing to work for extremely low wages and trying to maraud (infiltrate???) the Western welfare systems. Others were anxious that the Enlargement would induce a massive increase in FDI leading to an outsourcing of Western jobs into the East.

Obviously the clash of two completely contrary positions raises multiple questions. Who is right? How many migrants can be expected to come, once the free movement of workers has been granted to them? How does the EU Enlargement really affect the Western economies and their labor markets in particular? Is there a threat of “welfare shopping” and, if there is any, what measures can the Western countries take in order to protect themselves from the Eastern menace?

In this paper I will deal with these questions by first looking at the theoretical backgrounds and then seeking to support them with empirical findings in order to deduce some answers. In Chapter 2, after a brief overview of the main migration theories, I will examine the migration potential from the New Member States. Chapter 3 aims to provide a deeper insight into the economic and fiscal impact of immigrants on the German state. Chapter 4 deals with the “welfare magnets” hypothesis and provides an answer to the question whether immigrants really pose a threat to the Western welfare states. Finally, in Chapter 5 I argue whether the EU-15’s social policies need to be harmonized and offer a solution to the “welfare magnets” phenomenon.

In this paper I will concentrate on two EU-15 countries – Germany and, to a lesser extent, the UK. There are several reasons for my choice. Although Germany did not consider itself an immigration country until the migration law (Zuwanderungsgesetz) in the year 2002 changed this paradigm, it has received the highest number of immigrants in Western Europe since World War II. Net migration of foreign nationals to Germany included 6.7 million people between 1955 and 2000 and in addition 3.9 million ethnic Germans came^[2]. Table 1.1 displays the latest available numbers of migrants from the CEEC-10 in the EU-15. It clearly documents that geographical distance is a key factor explaining the allocation of migrants from the CEEC-10 across the EU-15: the countries bordering the CEECs, i.e. Austria, Germany, Greece and Italy absorbed more than 80% of the migrants from the accession countries before the Enlargement. In absolute terms Germany is, however, still the largest receiving country. Around two-thirds of the foreigners from the CEECs in the EU reside in Germany. Furthermore, Germany’s high living standards and generous welfare system are an additional incentive when choosing a destination country.

Although Britain was always regarded as an emigration country, positive net immigration flows since 1983^[3] demonstrate that at the end of the twentieth century this trend has been reversed. However, the main reason why I have included the UK in my analysis is its language. Today English is (still) the most important language of the world, with its ubiquity mainly due to the fact that (unlike Chinese) it is relatively easy to learn. Table 1.2 shows that English is easily the most popular foreign language among pupils from the AC-10. And being fluent in the natives’ language is one of the main prerequisites for having success as an immigrant.

After outlining the plan for the paper, in the next section I present an overview of the free movement of labor in the EU, as this fundamental freedom constitutes an indispensable part of European migration.

1.1 The Free Movement of Workers in the EU – Past, Present and Future

The foundation of a Common Market in Europe was based on the idea that integration of all four markets, that is, the markets for goods, services, capital, and labor would enhance welfare of all members of the Community. The Treaty of Rome, which established the then European Economic Community (EEC) in 1957, defined the free movement of workers, consequently, as one of the four fundamental freedoms of the “Single Market”^[4]. It took, however, another ten years until the free movement was granted to the citizens of the Community. Table 1.3 summarizes the most important events concerning the immigration policy of the EU for the period 1957 – 2004.

The admission of ten new members on May 1, 2004 differed substantially from previous enlargement rounds (see Table 1.4). Never before have so many countries and persons entered the EU at the same time. This has fuelled fears that a large number of new EU citizens will move to the EU-15. Furthermore, the collective GDP of the AC-10 amounts only to 4.8% of the EU-15 GDP. This is less than for all preceding enlargements, except the accession of Greece in 1981. Whereas the current enlargement in economic terms is smaller than previous enlargement rounds, it also results in greater differences between rich and poor EU member states. Comparatively less wealth in acceding countries is seen as a push factor for migration and the higher wealth of older member states as a pull factor. On average the GDP in the AC-10 is only a quarter of the EU-15 average. Even by measuring the GDP by purchasing power parities (PPP), significant gaps remain between the old and the new member states. All of the acceding countries had levels of PPP-GDP per capita below the EU-15 average in 2003^[5] (see Figure 1.1). Eight of the AC-10 had lower income levels than Portugal (the EU-15 country with the lowest PPP-GDP income) and four of the new member states, namely Estonia, Poland, Lithuania and Latvia could not even reach 50% of the average income level of the EU-15. Another important difference is that labor in the AC-10 is comparatively cheap. In the field of manufacturing the average hourly wage in the AC-10 reached a value of 22% of the comparable western European wage in 2000 (see Figure 1.2). Concerning the AC-10 endowment with physical capital it can be stated that the book values of physical capital stocks amounted to 10% of capital stocks in the EU-15 member states^[6].

In the face of these disparities, the EU-15 member states opted for the introduction of transitional periods in the area of the free flow of persons for the citizens of eight new member states (Cyprus and Malta were not included). The old member states were given the right to introduce the transitory provisions for two, five or maximally seven years in order to limit the access to their labor markets for workers originating from the new member states^[7]. The key element of the transitory provisions was the possibility to apply national measures and those resulting from the previous bilateral agreements in lieu of the Community law, which normally would have had to be applied.

The decision whether to introduce the transition periods was left to the respective member states. After the first two years following the accession (i.e. in 2006), the EU Council is expected to review the situation, but the decision whether to discard or to maintain the restrictions will be again left to the states. Another three years later (i.e. in 2009) all the restrictions on the free flow of workers should be lifted, with the exception of the countries where there are serious disturbances on the labor market or a threat thereof. Such countries would be eligible to extend the application of the transitory measures for the subsequent two years. In any case, after 2011 no Member State will be allowed to require work permits anymore^[8]. Table 1.5 contains the transitional regulations in the different EU-15 Member States.

What does the future hold for the EU? Bulgaria and Romania, the two candidate countries that were not accepted with the first wave in 2004, have signed their Accession Treaties in Luxembourg on 25 April, 2005; they hope to join the EU in 2007. Turkey, a very controversial candidate, is set to open negotiations on 3 October, 2005^[9]. But there are more countries than will be knocking on the EU gates sooner or later (see Figure 1.3). Not letting them in would not only be unfair to them, but would also conflict with the set of political values the EU has been trying to identify with ever since its foundation.

CHAPTER 2 Migration into the EU-15 – Theoryand Empirical Evidence

What causes a person to move to another country and what are the economic theories that try to explain such a person’s behavior? The purpose of this chapter is to shed some light on these and other similar questions. Section 2.1 offers a brief overview of the main migration theories according to economic literature. Section 2.2 focuses on the empirical studies that predict the migration potential into the EU-15 following the accession of the AC-10 on May 1, 2004. Furthermore, it demonstrates that the now ubiquitous fear that an army of poor East-Europeans will flood the Western labor markets is ill-founded and implausible. Finally, Section 2.3 shows whether the studies’ forecasts hold true by examining the actual migration flows after the Enlargement.

2.1. The Theories of International Migration

Ravenstein was one of the early migration analysts, who lived in the late 19th century. The basic assumption of his model^[10] is that individuals maximize their utility subject to a budget constraint. Migration occurs because of the geographical differences in the demand and supply of labor markets. Regions with a shortage of labor relative to capital are characterized by a high equilibrium wage, whereas regions with a large supply labor relative to capital are facing low equilibrium wages. This wage differential causes a migration flow from low wage to high wage regions. In response to the additional workers, the supply of labor in the high wage region increases, thus depressing the wage there. Similarly, the reduced number of workers in the low wage region decreases the supply of labor there, pushing the wage rate upwards. The migration flow ends as soon as the wage differential between the two regions reflects the costs of moving from the low wage to the high wage region. The model demonstrates that labor migration emerges from the actual wage differentials between the regions, i.e. the larger the wage differential, the larger the migration flow.

Most modern studies of migration decisions seem to agree with this first model; they are based on the hypothesis formulated by Hicks which states that “differences in net economic advantages, chiefly differences in wages, are the main causes of migration”^[11]. According to Hicks, the prime determinant of the decision to migrate is the difference in wages (as well as other sources of income) between the receiving and the sending countries, net of migration costs.

Roy’s^[12] hilarious story of hunters and fishermen laid the foundations for the well-known model of self-selection. This model describes how workers sort themselves among employment opportunities. Several decades later, Borjas developed Roy’s model further in order to tackle the question of how immigration policy might affect the composition of migrants. Under the assumption that the costs of migration are a constant share of wages Borjas^[13] showed that the self-selection of migrants is determined by the distribution of incomes in the countries of origin relative to the countries of destination. He found out that, for a given structure of incomes in the host country, the net benefits for migration are higher for individuals with high abilities and skills if the income distribution is more equal in the country of origin than in the country of destination and vice versa. Thus, removing the barriers to immigration vis-à-vis a country with a lower inequality of income would lead to a more “favorable” skill composition of migrants, while the removal of barriers to immigration vis-à-vis a country of a higher inequality of earnings yields a less “favorable” self-selection of migrants.

The early approach of Ravenstein was extended in various ways, most notably by the contribution of Harris and Todaro^[14]. Their original model describes migration from rural areas to the city in a developing economy but can also be applied more generally to migration between countries. They have dropped the neoclassical assumption of full employment in the sending and the receiving region and included the consideration of the probability of employment in the destination region by migrants. In this model, the migration decision is determined by the expected rather than the actual earnings differentials: people migrate if the expected wage is higher than at home. The model’s equilibrium is again brought by labor mobility: migration between countries continues until the expected earnings are the same in each country. The main conclusions are that high unemployment and lower wages lead to lower immigration into the country of origin. One serious shortcoming of this model is the fact that the expected wage (after adjustment for migration costs) is the same in all countries in the long run. This is not very realistic, considering the fact that even between the EU-15 Member States there are marked differences. For example, the net hourly wage in the Netherlands is over 50% higher than in Portugal, while unemployment in the Netherlands is lower^[15].

Another important school of thought regards migration from a human capital perspective. Sjaastad is generally regarded as its founder. His model, which probably has become the most influential and widely spread approach, treats migration as “an investment increasing the productivity of human resources, an investment which has costs and also renders returns”^[16]. Depending on their skill levels, all individuals calculate the present discounted value of expected returns of their human capital in every region, including the home location. Migration will occur, if the returns, net of the discounted costs of movement, are larger in a potential destination region than the returns in the country of origin. Sjaastad breaks down the migration costs into money and non-money costs. The former include the out-of-pocket expenses of movement (e.g. the transportation costs for migrants and their belongings), while the latter include foregone earnings (e.g. earnings foregone while looking for a new job in the destination country) and “psychic” costs of changing one’s environment (e.g. leaving friends and family behind). It should be mentioned as well that every individual evaluates the returns and costs in a different way, depending on personal characteristics such as age, gender, schooling, etc. A number of important conclusions can be drawn from the human capital model:

- The likelihood of migration decreases with the person’s age, reflecting the smaller expected lifetime gain from moving for older people.
- Individuals with higher education exhibit a higher migration probability, because an individual’s greater ability to collect and process information, gained through higher education, reduces the risks of migration.
- The risks and costs of movements are expected to rise with distance, because more information about labor market conditions is available for closer locations.

Finally, the network approach^[17] offers a dynamic view of labor migration. According to this model, migration may become a self-perpetuating process, because the costs and risks of migration are lowered by social and information networks. Due to a lack of information about the labor market in the destination country, the first person moving faces high costs and risks. After the migration of the first individual, the monetary and psychological costs of migration are substantially lowered for the relatives and friends of this individual in his home country. Furthermore, existing networks can for example help newly arrived people to find a job or a house or to obtain a work permit. This reduction of costs and risks leads to a higher migration probability. A new migrant raises the number of persons in the destination country, which results in a self-perpetuating migration process. However, the entire population in the home country will not be affected, hence this process will eventually stop. The rising wages in the home country and the falling wages in the destination country are another factor which lowers the possible benefits of migration. These diminishing effects are very important for the stability of the model, because it would otherwise predict the migration of whole countries.

To sum up, most migration theories predict substantial migration into the EU-15, considering the fact that the real wages in the CEECs are around 22% of the respective levels in the EU-15 (see again Figure 1.2). It would appear that the west European countries’ fear of the “East European army of labor migrants” is justified. The next section introduces the most important empirical studies that deal with this problematic.

2.2. Empirical Evidence about Migration into the EU-15

It is very difficult to accurately forecast the migration flows from East to West Europe. A large number of studies have been conducted in order to test the predictions from migration theories with respect as to the migration flows into the EU-15. These empirical studies differ widely. Some examine interregional mobility within one country, while others examine regional mobility between countries; some studies consider the migration decision from the perspective of the individual, while others concentrate more on the implications of network effects in the destination countries. The diversity of approaches makes it difficult to compare the results, as various studies arrive at different conclusions.

Table 2.1 presents an overview of the most renowned empirical studies assessing the size of post-enlargement migration flows. It starts with the seminal work of Layard et al. in 1992, which is widely considered the first attempt to estimate the East-West migration potential. The table is organized chronologically, with the most recent studies coming at the end. The majority of studies examine the possible migration potential from the ten CEECs into the EU-15, three of them concentrate on Germany and one deals with migration into the UK. All studies can be clustered into three groups, depending on the methodological approach they use: opinion polls, extrapolations from South-North to East-West migration and forecasts based on econometric estimates of macro-migration models.

Opinion polls allow deep insights into migration intentions and the human capital characteristics of potential migrants. However, there are numerous problems which make it almost impossible to derive accurate quantitative forecasts for the migration potential from such polls. An important limitation of the survey-based studies is the gap between observed intentions to migrate and actual migration behavior; it is not known whether somebody who reveals a general propensity to migrate in an opinion poll has indeed serious intentions to move, as the survey answers are not binding. Secondly, representative surveys provide information on the supply side only, that is, on the propensity of workers to migrate, but not on the demand side, i.e., on the ability of labor markets to absorb an additional labor demand or on the availability of housing. Thirdly, opinion polls cannot mirror the temporary dimension of migration appropriately: as migration from the 10-CEECs is largely a temporary phenomenon, the share of the population that will move to another country and perhaps return within a certain period of time is much higher than the share of the population which will live in a country at a given point of time.

The second approach to predict migration flows is to extrapolate the number of South-North migrants in the 1960s and early 1970s to East-West migration. However, this approach also suffers from serious shortcomings^[18]. On the one hand, income differences were much smaller during the Southern Enlargement. At that time, the average wage income of Spaniards and Portuguese was close to 50% of West German wage income. This is very different from the 15% of West German wages currently existing in the AC-10. On the other hand, in the Southern Enlargement case, migration had already occurred before EU accession, where there has only been relatively little legal immigration from Eastern Europe. In the period from 1960 to 1974-75, until the end of the Franco and Salazar dictatorships, there had already been mass emigration from Spain and Portugal. Although both countries had also been experiencing very high immigration from their former colonies at that time, net emigration of the Iberian population had already amounted to 5.5% p.a. during that period. The possibility of emigration distinguished the two dictatorships from those in Eastern Europe. Whereas Franco and Salazar had allowed the free movement of people, the Soviet Union had closed off its territory with an Iron Curtain, thus preventing any emigration. This fact should not be overlooked when searching for parallels.

A final method to evaluate the potential migration flows from Eastern Europe is to rely on an econometric analysis of the determinants of internal migration within the EU-15 and to use the results in order to simulate the potential future migration flows from East to West. The main economic variables that explain the migration flows and future migration stocks are income differential, the (un)employment rate, geographical distance, the presence of networks, etc. Although most studies employ the same set of explanatory variables, the estimates of parameters, and hence, of migration potentials differ considerably in the economic literature. The main problem of the econometric studies can be traced back to the fact that they have to transfer parameter values which are estimated in another historical context and in another country to project the migration potential from CEEC-10. The CEECs could not be included in the original sample, since the Iron Curtain prevented migration. Thus, the projections rely implicitly on the assumption that the estimated parameter values remain constant not only across time, but also across space. Unfortunately, this assumption isn’t realistic, as migration behavior differs largely across countries due to differences in geography, language, culture, etc.

Bearing in mind these caveats, the migration potentials presented in Table 2.1 should be evaluated with caution. All authors more or less obviously hint that their results should be interpreted as possible migration trends, rather than taken at face value. The main results can be summarized as follows:

- Most studies forecast an initial migration flow into the EU-15 of 100,000 – 340,000 p.a.
- After an initial peak (usually immediately after the introduction of free movement or several years afterwards) migration flows gradually decrease.
- Some studies assume, that the long-term migration stock would be reached as soon as 2015 with figures ranging from 2,200,000 – 5,000,000 persons; other studies assume, that it would be reached in 2030; the spectrum here shrinks to 2,300,000 – 3,900,000 persons.

A major deficiency, common among all but two studies, is the neglect of transitional periods. Chapter 1 showed that most EU-15 countries have opted for the implementation of transitional periods of up to seven years. The omission of this very important proviso is mainly due to the fact that only until recently it was not known if and for how long transitional periods would be implemented. The two studies that have considered transitional periods in their calculations conclude that the delay of free movement will not cause a significant increase of migration flows towards the EU-15. Zaiceva states in her paper, that “Legal introduction of free movement of workers in 2011 will not lead to substantial jump in the number of migrants”^[19]. Alvarez-Plata et al.^[20] arrive at similar conclusions: postponing free movement until 2011 yields only a marginal reduction in the net increase of migrants (some 3,000 p.a.) after free movement has been introduced. This decrease in migration is due to income convergence (due to trade liberalization, openness to FDI, structural funds from the EU and remittances back home). The migration stocks tend to converge relatively rapidly into their long-run stocks. Altogether, transitional periods tend to postpone the migration influx, but they are only going to have a marginal impact on the size of the inflows and, more importantly, on the long-run stocks of the migrant population.

2.3. Is There Migration Diversion After the Eastern Enlargement?

At present, there is very limited information on migration trends since May 1, 2004. Available evidence suggests that the transitional periods resulted in diversion as well as in reduction of migration flows relative to the case of free labor mobility. First, the Home office in the UK reported that more than 130,000 nationals from the AC-10 had registered for work between May and December 2004, of which 40% had already been in the UK prior to the Enlargement. If these figures imply that 80,000 persons from the AC-10 migrated to the UK in 2004, this would be more than five times the migration potential predicted by Dustmann et al.^[22]. However, the figures published by the Home Office covered also temporary migrants such as seasonal workers, and not all individuals who had registered did actually take up jobs later, in which case, the actual immigration might be lower. Yet given that the official number of nationals from the CEECs was below 50,000 persons at the beginning of the 1990s, the figures published by the UK Home Office pointed to a substantial increase in immigration.^[21]

For Ireland, a country which pursued the same immigration policies as the UK during the transitional period, there is contradictory information. On the one hand, 7,500 work permits were issued to AC-10 nationals from January 1 to October 31, 2004, down from 20,000 in the year 2003. On the other hand, the Irish government reported that 31,000 personal public service numbers (certificates which are necessary for a work permit) were issued to AC-10 nationals in the five months from May 1 to October 31, 2004 pointing to a substantial increase in the flows form CEECs relative to 2003. Thus, it is possible that migration flows into Ireland were several times higher than predicted.

Some diversion of migration flows from CEECs was also observed in the Nordic countries. In Sweden, the only EU country without transitional arrangements, the number of work permits doubled from 2,097 in 2003 to 3,966 in 2004. In Norway, which partially opened its labor market and is booming because of the oil price hike, the number of released work permits increased from 18,170 in 2003 to 25,325 in 2004. Meanwhile the Nordic countries tightly restricting migration from the CEECs experienced modest or declining migration flows. In Denmark, 2,048 work permits were issued in 2004. Comparable figures for 2003 are not available here. However, the number is reasonably low relative to the predicted inflow of 3,000 persons. In Finland, work permits dropped from 6,747 in 2003 to 2,169 in 2004.

No information is available as yet on migration to the traditional destinations of migrants from the CEECs, namely Austria, Germany and Italy, although according to statements of Government officials it would seem that migration from the CEECs has been stable after the Enlargement.

Overall, the scattered information available at the time of writing points to some diversion of flows from countries tightly closing borders to countries with more liberal rules with respect to migration from the AC-10. This is particularly true for English speaking countries, where migration figures exceed by far those of the migration projections. The Eastern Enlargement episode so far suggests that asymmetries in migration restrictions affect the geographical orientation of migration flows. Moreover, these diversion effects may become over time more important as networks of citizens from CEECs are established in the new destinations.

CHAPTER 3 The Economic and Fiscal Impact of Immigration on the German State – Boon or Burden?

In the previous chapter I analyzed the current migration trends in the CEECs and came to the conclusion that migration pressure has indeed been accumulating in the New Member States, albeit to not unendurable levels. In this chapter we turn our attention to the EU-15 and in particular to Germany. Section 3.1 provides an answer to the question whether the EU-15 Member States need immigration at all. Section 3.2 begins with a simple theoretical model that identifies the winners and losers from international migration in the host country. Afterwards it focuses on the consequences that arise for the German labor market, i.e. whether immigrants depress the wages of the natives and/or force them into unemployment. Section 3.3 addresses the fiscal aspects of immigration; it tackles the question whether immigrants are net fiscal beneficiaries or contributors to the German state.

3.1. Do the EU-15 Really Need Immigration?

In 1957 the founding fathers of the EU knew that the road towards an “ever closer union” was not going to be easy, but they certainly did not expect it to be this arduous. Today the EU is mired in a myriad of contentious issues, ranging from budget squabbles to the identity crisis caused by the rejection of the EU Constitution. Small wonder, that a new looming threat, one of the most insidious problems of modern-day rich societies, passes largely unnoticed – the ageing population of the EU. This problem is two-fold: on the one hand, life expectancy has been rising for nearly 300 years, and it rose particularly quickly in the second half of the 20th century. Between 1950 and 1995, average life expectancy in Britain increased from 69.2 to 76.2 years and in France from 66.5 to 77.1^[23]. On the other hand, the ageing of the EU population is being combined with falling birth rates, a relatively new phenomenon. Between 1960 and 1965 the overall birth rate in the EU was 2.7 children per woman, comfortably above the rate of 2.1 required to maintain the size of the population. By 1995, however, this rate had fallen to 1.5 children per woman, in some countries (e.g. Italy) this figure is even lower.

These problems are endemic to Germany as well. Möller^[24] argues that, in the next 50 years, 344,000 persons p.a. should immigrate to Germany in order to keep the population constant and 486,000 persons p.a. to keep the working-age population constant. However, his estimations are rather modest if compared with those of Brücker and Kohlhaas^[25]. Their forecasts are presented in Table 3.1. In their first scenario they depict the demographic development without immigration: the German population is set to fall from 83,000,000 persons in 2002 to 59,000,000 in 2050. Only a net migration inflow of 710,000 p.a. ensures a constant working-age population at the present volume of 45,000,000.

All these figures show unambiguously that the EU-15 will be in acute need of immigrants in the years to come. It is still a moot point among the economists, though, whether the majority of migrants should come from the CEECs, as they have shrinking population as well. Furthermore, migration doesn’t seem to be the ultimate solution to the ageing problem of the EU-15, since its main cause, the fertility decline, is not likely to be reversed due to typically observed adjustment of migrant fertility to the standards of natives^[26].

3.2. The Economic Impact of Immigrants

3.2.1. Who Wins and Who Loses from International Migration?

Because natives often fear the competition from foreign labor, public concerns about the wage and unemployment impact of immigrants have become a highly controversial issue and a hot political commodity. The following simple comparative-static model^[27] derives the main labor market effects of immigration, although it does not give a clear-cut answer as to whether immigration is beneficial or detrimental to the natives.

Assume that the receiving country produces a single good by means of capital and homogenous labor. Prior to immigration, the labor market is in equilibrium at point b with wage w1 and employment L1 (Figure 3.1).

Figure 3.1 Labor Market Effects of Immigration

illustration not visible in this excerpt

Source: Bauer and Zimmermann (1999) p. 49

Immigration leads to an outward shift of the labor supply curve from S1 to S2 . This labor supply effect of immigration leads to a new equilibrium at point c, which is characterized by a higher total employment L2 and a lower wage w2 . The employment level of natives has decreased from L1 to L3, while the employment level of foreigners is given by the distance L2 to L3 . According to this model, immigration leads to a lower employment of natives and reduces their wages. How much the employment and wages of natives decrease depends on the elasticity of the supply and demand curves.

The income and distribution effects can also be derived from Figure 3.1. Prior to immigration, the receiving country produced an output described by the area abL10 . The income of the native workers was described by the area w1bL10 and the income of the capital owners by the area abw1 . Immigration leads to an increase in the total output of the country from abL10 to acL20 . The income of the immigrants is described by the area ecL2L3 . Because of immigration, the income of native workers decreases to w2eL30 and the income of the capital owners increases to acw2 . The triangle bcd is the so-called immigration surplus, “the increase in national income accruing to natives”^[28]. To summarize, immigration leads to an increased output and to a redistribution of income from native workers to the owners of capital.

This model overemphasizes the negative effects of immigration for native workers, because indirect labor demands are neglected. Immigrants are consuming goods, thus increasing the demand for goods. This increased demand leads to an increase in the demand for labor (a shift of the labor demand curve to the right). Due to this shift, the employment of natives as well as their wages increase. Whether this indirect effect is able to compensate the direct effects of immigration depends on how much the labor demand curve shifts to the right. If the increase in labor demand is substantial enough (e.g. D3), immigration could lead to an increase of wages and the employment of natives.

The main results from this model could be used for a slightly more complicated model with heterogeneous labor^[29] that describes more realistically the current European labor markets. The main assumptions of this model are that the previously homogeneous labor is now split into low-skilled and high-skilled workers; capital and both types of labor are complements, while high-skilled and low-skilled workers are substitutes; immigrants are either perfect substitutes to low-skilled or to high-skilled natives; immigrants do not carry any capital with them and have no effect on the demand side of the economy; the labor market clears, i.e. wages are flexible.

An additional supply of low-skilled immigrants does not increase the unemployment rate of low-skilled natives, because the wage for low-skilled workers falls; the income of capital owners and high-skilled natives increases as a consequence of the complementarity to the low-skilled workers.

In the case of high-skilled immigration, we can observe exactly the opposite results: the wage rate of high-skilled natives falls, while returns to capital and the income of low-skilled natives increases.

To sum up, the main insights from the two models are that although the total output in the receiving country increases, there will always be opponents of immigration because of the income distribution effects between the various social classes.

3.2.2. Empirical Evidence about the Impact of Immigration on Domestic Wages

The second model in the last subsection demonstrated that natives who are complementaries to the immigrants benefit (i.e. their wages rise) and natives who are substitutes to the immigrants lose (i.e. their wages fall) from international migration. A number of papers assess empirically this conformity between immigrants and domestic wages. Most of these studies rely either on a cross-section of regions, or branches, and they use variations in the immigrant density in order to identify the impact of migration on wages or other variables of interest, e.g. employment opportunities. All these studies face two basic methodological problems^[30]. First, the share of foreigners in a region or industry cannot be treated as an exogenous variable: immigrants usually select themselves into prosperous regions or branches, such that simple regressions between immigrant density and wages yield spurious results. Many studies rely therefore on the instrumental variable technique in order to control for the endogeneity of migration decisions. However, the choice of suitable instruments is controversial and the findings of several studies have demonstrated that quantitative results are strongly affected by the use of instrumental variables and their specification. The second problem is that in open economies an increased labor supply through migration to an industry or region may spill over into other sectors of the economy and affect relative wages at the aggregate level rather than inter-industrial wage differentials.

Table 3.2 contains a summary of the most important European studies about the effects of immigration on domestic wages. The last column shows the effect (in %) of 1% increase of immigration on the domestic wages. The sign in front of the coefficient indicates whether the effect is positive (i.e. wages rise as a consequence of immigration) or negative (i.e. wages fall). The relatively big variation between the coefficients is due to different assumptions about the models and different methodologies used.

To sum up, the majority of empirical studies (with a few exceptions) confirm the theoretical results that immigration affects negatively only those native workers it substitutes, and has a positive impact on the workers it complements. The wage effect in most studies is rather small or neutral. There are several factors that may explain the small wage effects^[31]. First, immigration may cause natives to move out of the areas with a high density of immigrants into areas with a low proportion of foreigners. This would result in a decreased supply of native labor in those areas with a high proportion of foreigners and an increased supply of native labor in regions with a low foreigner share. Overall, natives’ wages would fall, but far less when compared to a situation in which natives do not move in response to immigration. Secondly, binding wage floors (i.e. minimum wages or social security levels) might prevent natives’ wages from falling. In this case, immigration would lead to increased unemployment of natives without reducing their wages. The next subsection deals with the empirical studies tackling this problematic.

3.2.3. Empirical Evidence about the Impact of Immigration on Domestic Unemployment

As explained above, the existence of rigid wages may be one possible reason for the small wage effect found in the empirical studies in Table 3.2. If there are binding wage floors in the EU-15 labor market, it might be the case that immigration would actually increase the unemployment of natives without affecting their wages. According to the theoretical framework presented above, one plausible assumption would be that the higher the substitutability of immigrants for native workers, the more likely an increase in immigration would lead to an increase in unemployment, if the wages in the destination country are inflexible. Once again using the reasonable simplification that high-skilled and low-skilled workers are complements and that the immigrants are substitutes for low-skilled workers and complements for high-skilled workers, an increase of the immigration stock would push the blue collar workers into unemployment, while inducing the reverse effect on the white collar workers. On the other hand, assuming that the labor market for high-skilled workers is flexible, foreign white collar workers would decrease unemployment of low-skilled natives without having an employment effect on the high-skilled natives.

The empirical studies on the impact of immigration on unemployment are less numerous than the ones on wages. In the previous subsection I had the “luxury” of selecting only the most appropriate studies for my paper (i.e. the studies on Germany and the UK) out of a very voluminous range of papers. The motley group of studies presented in Table 3.3 incorporates the empirical evidence on various EU-15 countries. Overall, the empirical studies seem to confirm the premise that immigration has a slightly negative effect on domestic employment. The only study that differs from this trend is the one by Gross^[32], who investigates the effects of immigration flows into the French labor market. He arrives at the somewhat counterintuitive conclusion that, in the long run, the immigration rate and the unemployment rate are negatively correlated. In the short run, however, immigration slightly increases domestic unemployment. His explanation is that, independently of their participation in the labor market, immigrants create additional jobs by way of their demand for goods and services in the long run. The findings from the other studies can be best summarized by Dustmann et al.: “The main result of the empirical analysis is that there is no strong evidence of large adverse effects of immigration on unemployment [or wages] of existing workers”^[33].

3.3. The Fiscal Impact of Immigrants

Migrants, to varying degrees, pay taxes, claim benefits, and consume government-provided goods and services for the entire time they live in the host country. Through their participation in all these activities they have a direct impact on government expenditure and revenue; the extent to which they are involved in each of them determines the direction and magnitude of this effect. If migrants pay more in taxes than they consume in benefits and state services they are said to make a positive net fiscal contribution. This represents a transfer from migrants to natives. If migrants have a negative fiscal impact on the host economy the value of the benefits and public services they receive exceeds the value of the taxes they pay, with a corresponding transfer of wealth from the natives to immigrants. In the remainder of this chapter I will first present some papers that conclude that migration has a negative fiscal impact on the public finances; afterwards I turn the tables and discuss studies with the opposite results – that immigrants are positive net contributors to the public coffers.

3.3.1. Studies Confirming the “immigrant-as-a-burden” Thesis

In an extensive study based on the socio-economic panel, the Ifo Institute tried to estimate the impact of immigration on the public finances in Germany in 1997^[34]. In a nutshell, Sinn et al. balanced the immigrants’ received transfers from the state against the net taxes they paid on an annual basis. The authors have taken account of taxes, contributions, pensions, welfare benefits as well as all the indirect benefits from public goods. Public goods include, for example, roads, bridges, parks, environmental protection, the courts, the administration, the police, fire fighters, etc. The findings of this “snapshot” study are presented in Table 3.4. It shows that immigrants paid less into health insurance than they received from the state, but made high net payments into the pension insurance system, because the discounted value of their payments exceeded the pension claims established (which is not surprising, as the average migrant is 5 years younger than the average native)^[35]. Unemployment insurance profited from those immigrants who had lived in Germany for less than 25 years and lost from those immigrants who had been in Germany longer than this. As the latter were not very numerous, unemployment insurance gained on balance. But the immigrants paid less in taxes than they received in the form of tax-financed welfare benefits and public infrastructure services. In these areas the state experienced a big deficit. Immigrants who had been in Germany for less than ten years could realize, on balance, a net gain from redistribution of annually € 2,367 per head. This net gain can be interpreted as a migration premium.

However, this fiscal loss could be reduced by taking into account the implicit taxes paid by the children of immigrants (the last two rows in Table 3.4.). The reason for this is that the immigrants’ pension will be financed by the children of the immigrants themselves. If all immigrants would leave their offspring in the pension system, there would be an additional benefit to the pension system of € 1,126 per immigrant. This effect reduces the absolute value of the negative net fiscal externality, but does not change its sign for the group of immigrants who stay for less than ten years. The negative fiscal externality in Germany is € 1,241 per person and year in that case.

If immigrants stay longer, they will also succeed in integrating themselves better into working life. Occupational skills and language proficiency will improve and wages will rise with productivity. With higher wages they will have to pay higher taxes, and the redistribution loss of the state will decrease. Migrants residing in Germany for less than 25 years but for more than ten years received during that period of their stay, on balance, only € 1,330 per year from the state and with a full child effect they received about as much as they and their descendants contributed – the net fiscal balance in that case was only -€ 16. On average, migrants who had resided in Germany for more than 25 years made net payments to the state of € 853 per head during that period of their stay without the child effect and € 2,610 with the full child effect. Unfortunately, as a rule, the immigrants did not stay long enough in Germany to become net payers. About 60% of the immigrants surveyed had returned home after ten years, and only 15% of them participated in the official labor market. After 25 years, more than 80% had either died or had returned to their home country. Overall, the authors arrive at the conclusion that immigration involves quite substantial fiscal losses for the state, at least in the short- and middle-term.

Storesletten^[36] employs an overlapping-generations model in order to assess the fiscal impact of immigration on Sweden. According to his results, the discounted net government gain from admitting a 20-30-year-old immigrant is roughly SEK 200,000 (€ 21,400^[37] ). However, the costs of old and very young immigrants are substantial – up to 1.5 million for a 66-year-old immigrant. These findings are due to the fact that the pension system, healthcare, childcare and schooling are very expensive programs. On average, the net government loss per new immigrant is SEK 175,000 (€ 18,725). These results suggest that immigrants to a typical welfare state such as Sweden impose, on average, a substantial fiscal burden.

3.3.2. Studies Confirming the “immigrant-as-a-net-contributor” Thesis

Using estimates of migrants’ shares in various public expenditure and taxation categories Gott et al.^[38] calculate the net annual fiscal impact of immigrants on the UK. They estimate that in 1999/2000 migrants in the UK contributed ₤ 31.2 billion in taxes, and increased public expenditure by ₤ 28.8 billion through their receipt of public goods and services, resulting in an estimated net fiscal contribution of around ₤ 2.5 billion. In other words, in 1999/2000 migrants in the UK reduced the amount that the existing population paid in taxes, or increased the amount the natives received in welfare and public services, by ₤ 2.5 billion. The paper concludes that, overall, the current immigration population in the UK performs well economically and makes a net fiscal contribution. However, immigrants are heterogeneous. Though some are found to make a positive net fiscal contribution, some do less well economically, and are likely to have a negative fiscal impact. Domestic policies aimed at improving access to the labor market and tackling social exclusion can help address this.

[...]

---

^[1] I would like to express my utmost gratitude to Robert Gralley for proof-reading this paper as well as for valuable comments.

^[2] Dietz, B. „East West Migration Patterns in an Enlarging Europe: The German Case“ (2002) p. 29

^[3] Hatton, T. et al. „Migration, Migrants and Policy in the United Kingdom“ (2005) p. 116

^[4] Art. 48-51, EEC Treaty

^[5] van Suntum, U. „The Impact of EU’s Eastern Enlargement on Western European Labour Markets“ (2005) p. 3

^[6] Brücker, H. „Werden unsere Löhne künftig in Warschau festgesetzt?“ (2001) p. 76

^[7] Bijak, J. et al. „Long-term International Migration Scenarios for Europe, 2002-2052“ (2004) p. 3

^[8] European Commission „Free Movement for Persons – a Practical Guide for an Enlarged European Union“ (2002) p. 4

^[9] Le Monde, „Un an après, l’élargissement continue d’inspirer des craintes“ (2005)

^[10] Ravenstein, E. “The Laws of Migration“ (1889)

^[11] Hicks, J. “The Theory of Wages“ (1932) p. 76

^[12] Roy, A. “Some Thoughts on the Distribution on Earnings“ (1951)

^[13] Borjas, G. “Self-Selection and the Earnings of Immigrants“ (1987)

^[14] Harris, J. Todaro M. “Migration, Unemployment and Development: A Two-Sector Analysis“ (1970)

^[15] Ederveen, S. et al. “Destination Europe. Immigration and Integration in the European Union “ (2004)

p. 44

^[16] Sjaastad, L. “The Costs and Returns of Human Migration“ (1962) p. 83

^[17] Carrington, W. et al. “Migration With Endogenous Moving Costs“ (1996)

^[18] Sinn, H.-W. “EU Enlargement, Migration and the New Constitution“ (2004) p. 4

^[19] Zaiceva, A. “Implications of EU Accession for International Migration: an Assessment of Potential Migration Pressure“ (2004) p. 15

^[20] Alvarez-Plata, P. et al. “Potential Migration from Central and Eastern Europe into the EU-15 – An Update“ (2003) p. 40-43

^[21] This section draws heavily on Boeri, T. Brücker, H. “Migration, Co-ordination, Failures and EU Enlargement“ (2005) p. 15-16, as, to the best of my knowledge, this is the only paper so far that gives information about the current migration flows after the Enlargement

^[22] Dustmann, C. et al. “The Impact of EU Enlargement on Migration Flows“ (2003a). The forecast for the UK should be interpreted with great caution, as the methodology of this study was criticized severely, see Stone, M. “Prediction of Future Migration Flows to the UK and Germany” (2003)

^[23] The Economist, “The Crumbling Pillars of Old Age“ (2003) p. 72

^[24] Möller, D. “Migration und ihre Arbeitsmarkteffekte in Deutschland“ (2002) p. 154

^[25] Brücker, H. Kohlhaas, M. “Möglichkeiten der quantitativen und qualitativen Ermittlung von Zuwanderungsbedarf in Teilarbeitsmärkten in Deutschland“ (2004) p. 4ff

^[26] Zimmermann, K. “European Labour Mobility: Challenges and Potentials“ (2004) p. 22

^[27] See Bauer, T. Zimmermann, K. “Assessment of Possible Migration Pressure and its Labour Market Impact Following EU Enlargement to Central and Eastern Europe“ (1999) p. 48ff

^[28] Borjas, G. “The Economic Benefits from Immigration“ (1995) p. 6

^[29] See Bauer, T. “Arbeitsmarkteffekte der Migration und Einwanderungspolitik“ (1998) p. 40ff for the complete model with all specifications

^[30] Boeri, T., Brücker, H. “The Impact of Eastern Enlargement on Employment and Labour Markets in the EU Member States“ (2001) p. 78

^[31] Bauer, T. “German Migration: Development, Assimilation, and Labour Market Effects“ (2005) p. 249

^[32] Gross, D. “Three Million Foreigners, Three Million Unemployed? Immigration and the French Labor Market“ (1999)

^[33] Dustmann, C. et al. “The Local Labour Market Effects of Immigration in the UK“ (2003b) p. 48

^[34] Sinn, H.-W. et al. “EU-Erweiterung und Arbeitskräftemigration. Wege zu einer schrittweisen Annäherung der Arbeitsmärkte“ (2001) p. 159-237; see also Sinn, H.-W. “EU Enlargement, Migration and the New Constitution“ (2004)

^[35] Sinn, H.-W. et al. (2001) p. 185

^[36] Storesletten, K. “Fiscal Implications of Immigration – A Net Present Value Calculation“ (2003)

^[37] 1 EUR = 9.3545 SEK (10.08.2005) http://economist.com/markets/currency/md_conv.cfm

^[38] Gott, C. et al. “The Migrant Population in the UK: Fiscal Effects“ (2002)