Enlargement: The Challenge of Migration from the New Member States

Ladies and gentlemen,

Enlargement is the final act of the reunification of Europe, which began when the Berlin Wall came down. It is, fourteen years later, and written large on a Continental scale, the reunification of Germany again. As Hans-Werner Sinn points out, the relative sizes of the parts that are being joined are similar in the two cases: the population of the Eastern laender was 26% of the population of West Germany; today the population of the ten CEEC countries – eight of which entered May 1, and two of which remain candidates – is 27% of the population of the European Union before enlargement.

Now, as then, enthusiasm for the end of division is accompanied by concern for the population movements that the removal of barriers will permit. The combination of dramatic differences in standards of living, geographic proximity, and common EU citizenship gives credence to the prospect that these movements will be large.

1. The magnitude of the potential flows

How large? Three recent studies (place cursor here to read note) project potential flows from Central and Eastern Europe into Germany, which has been, and may well remain, the principal recipient of immigration from this region. These studies share a common conceptual framework. Potential flows are related to income and business cycle differences between the sending and the host country. The elasticity of migration to these differences is estimated from patterns observed during the last quarter of the twentieth century. The three studies allow for and find strong evidence of stock adjustment dynamics: after an initial catch-up period, the immigrant population in the receiving country tends to approach an equilibrium level, and net migration falls off.

Though differences in scope and econometric methodology between the studies generate substantial differences in their estimates of total immigration over ten or fifteen years, their projections of flows during the first five years are reasonably similar. The three studies can be interpreted as projecting a cumulative inflow of 1.0-1.5 million immigrants into Germany from the CEEC 10 in the first five years after labor markets are opened ( place cursor here to read note). This implies roughly a cumulative inflow over the same period of 1.5 to 2.5 million immigrants from the CEEC-10 to the EU-15.

These flows are substantial. Arrivals into Germany in the first five years are projected to add up to 1-2% of the host population. Since arrivals will tend to cluster, regions like Bavaria and countries like Austria could well find that the already high shares of immigrants in their total populations might double.

These numbers are not, however, without precedent. They are dwarfed by the influx into West Germany of German nationals from East Germany and Poland before the erection of the Berlin Wall (10.6 million of these refugees stayed in West Germany. More came and moved on.). The Poles, Romanians, Hungarians and others who move to Germany in the coming decade, may be fewer in number than the 2 million ethnic Germans, who were repatriated from former Soviet states, and given automatic citizenship in the first ten years after the fall of the Berlin Wall. Relative to the size of the host country, their number is roughly the same as the number of French citizens repatriated from North Africa at the end of the Algerian War.

2. The Concerns of Policy Makers in the EU-15

Sizeable but within the realm of recent experience, these flows are nonetheless substantial enough to be disruptive if they are not well handled. Policy makers have focused on two economic concerns.

1. Labor markets in the host countries must be flexible enough to assimilate the immigrant workers.
2. To the extent that immigrants receive more benefits than they pay taxes, they are a net burden on the State budget. If these deficits are large and well-known, they may act as welfare magnets and attract further immigration, which will, in turn, magnify the total burden.

2.1. The potential difficulty of integrating new workers in the labor market

This is the classical concern associated with the arrival of low-paid immigrants. The fear, bluntly put, is that "They will steal our jobs." The implicit scenario is one in which standard wage rates are rigid, and the natural mechanisms which would otherwise ensure that new workers find jobs, do not function. Indeed, if wages are rigid, an increase in the supply of labor inevitably generates increased unemployment. Existing workers fear that, by undercutting them, the immigrants will get the jobs, and they will become unemployed.

If real wages are flexible, it is not likely to take much of a wage reduction to absorb a 1% increase in labor supply due to immigration. If the elasticity of employment to reductions in the real wage rate were 0.8 (a number broadly consistent with the observed range of econometric estimates), a real wage reduction of 1.3% would suffice to absorb such an influx. The problem is that if real wages are rigid, a 1% influx of immigrants may indeed lead to a 1% increase in the unemployment rate, say from 10% to 11%. Though the former development would be imperceptible, the latter would constitute a serious deterioration.

2.2 The potential burden on the budget

The possible budgetary costs of a surge of immigration from the new member states elicit a second, different kind of concern. This is the fear that immigrants may draw more in benefits from the State than they contribute in taxes. The possibility that these net benefits may differentially attract immigrants to the host countries with the most generous welfare provisions, and potentially bankrupt those very provisions, adds to the worry. In order to evaluate this concern, it is necessary to examine, one by one, the principal components of the budgetary balance of the immigrant population. I will base my comments on the detailed analysis of this question by Sinn et al. (2003) for the case of Germany.

The social security taxes and transfers from and to immigrant workers clearly constitute a positive, net contribution to consolidated State budgets. By its nature, a new immigrant group consists mostly of young workers paying taxes, and only to a small degree of retirees receiving benefits. On a cash flow basis, the immigrants are positive contributors until their population matures demographically. It has been estimated that in 1989, the German social security budgets received DM 9 Billion more in taxes than they paid in benefits to the resident immigrant population.

To an individual immigrant, the expected lifetime balance of his personal, social security account matters more than the magnitude of aggregate, intergenerational flows. For this reason, Sinn et al. (2003) advocate estimating the budgetary balance of the social security system by comparing the present discounted value of lifetime contributions with the present discounted value of lifetime benefits for representative individuals. On this basis also, immigrants are net contributors, for the simple reason that, under present demographic conditions, the internal rate of return of the social security system is substantially below the rate of return for private savings. Like most workers in the system, immigrants never get back, in present discounted value terms, what they contributed. The total estimated positive contribution of each immigrant is less than it appeared on a cash flow basis, but it remains substantial (see below).

By contrast, unemployment compensation payments to immigrants do cost the State more than working immigrants contribute in related taxes, because unemployment rates are higher among immigrant workers. The higher rates are partly a reflection of the youth and relatively low skills of the immigrants. But there is also evidence, in some countries, that particularly generous unemployment provisions have attracted more than the natural share of the unemployment prone. In an econometric study using household survey data from several nations of the EU-15, Boeri et al. (2002) find significant evidence of what they call "excess unemployment dependence" in Denmark, Austria, the Netherlands and France.

Similarly, tax based welfare benefits paid to immigrants also constitute a net cost to the consolidated State budget of the host country. Because they initially work for lower than average wages, immigrants are net recipients of redistributive welfare measures such as supplemental assistance, housing bonuses and allocations to public housing stock. Because they also have more than the average number of children, they benefit disproportionately from child allowances.

Highly progressive benefits of this nature often generate resentment even within a homogeneous community of native born nationals. The visible nature of the differences between immigrant and non-immigrant dependency exacerbates these sensitivities, and often causes the distribution of these benefits to become charged political issues. Notwithstanding the intensity of the sentiments they arouse, neither these tax based welfare payments, nor the unemployment compensation mentioned above are the largest contributors to budgetary imbalance.

In an informative table, Sinn et al (2003) summarize the net positive and negative contributions per individual of the immigrant population to the consolidated German State budget, as of 1997. Table 2 (Table 4.14 in the text cited), decomposes the budgetary balances which can be associated with the immigrant population into their component parts. One sees there the net positive contribution that Germany's immigrants made to social security balances. In 1997, the average immigrant paid 5290 DM in social security taxes, and as a result became entitled to additional benefits whose present discounted value was DM 1778. In present discounted value terms, the average immigrant lost and the State gained the difference between those two numbers -- DM 3512.

Unemployment compensation, a highly politically sensitive element, generated a net cost to the State of only DM 196 per person per year.

The largest net cost to the State was for the provision of public goods and tax financed benefits. Sinn et al. (2003) estimate the cost of these provisions by specifically allocating the benefits which can be allocated, and distributing the rest uniformly over the entire population. The result is a large number, DM 12 337 per immigrant in 1997. The number includes education, fire, police, military expenditures and the services provided by all other government employees. The counterpart is general tax payments. Since the average income of immigrants is lower than the average income of non-immigrants, their tax payments are also lower. In 1997, they were DM 7 576 per person. The difference between these two numbers represents the net cost to the State, per immigrant, of the supply of general public goods and tax financed benefits to the immigrant population. The study does not distinguish between the portion of this cost which is attributable to public goods, and the portion which is attributable to welfare payments, but the relative size of these elements in the overall State budget suggests that public goods are the larger component. In short, the biggest source of the net cost to the budget of the immigrant population was the provision of basic public goods.

3. A Diversity of Policy Proposals

Concerns for labor market disruption and budgetary costs have generated two kinds of policy prescriptions.

Delays and Quotas. The first response has been to seek to delay 100% immediate implementation of the right of new EU members to work anywhere in the Union. The accession agreements, which the Commission negotiated with each candidate country, reserve the right for any of the EU 15 to postpone free access to its labor market for five years, and after review, to ask for an extension for another two years. Any citizen of a new member country who is granted a work permit is to benefit immediately from all of the employment and social rights to which EU citizens are entitled. The delay is optional, not mandatory, and may be applied in part as well as in full. Germany, Austria and France have announced that they are freezing work permits for two years, at the end of which time, they will decide on possible partial or full extension of the freeze. Other countries have decided to offer labor market access during the transition period on a staggered, quota basis, sometimes differentiated by skill, sector or region.

Limiting Benefits. Britain and Ireland, on the other hand, have announced that work permits and residence status will be given to citizens of new member nations immediately (place cursor here to read note). However, both countries propose to depart from the inclusion principle broadly applicable in EU law, in that they will deny public housing and employment related social benefits (the most important of which is unemployment compensation) to immigrants who have not yet worked continuously for a minimum period of months (sometimes said to be 12, sometimes 24). Current EU judicial practice allows member nations to withhold benefits to immigrants from other EU states if they are not working. However, they are deemed to become entitled to the full benefits of the national welfare system, once they have a job. British authorities argue that their benefit limitation is implicitly consistent with the accession treaties framework, because the later permit exclusion from employment, and therefore benefits, during the transition period.

The purpose of the British and Irish policies is clearly to discourage migration in the pursuit of welfare benefits. In the words of U.K. Home Secretary David Blunket, "Hard working immigrants are welcome. Benefit tourists are not." (place cursor here to read note)

Hans-Werner Sinn, who advocates a similar policy of no job restrictions combined with benefit limitations for Germany, argues that the transitional period of exclusion from benefits should be five to seven years.

The likelihood that benefit exclusions of any significant duration will be challenged as contrary to nondiscrimination provisions in EU law will increase substantially if the new EU constitution is adopted and ratified in its present form. As is widely known, the draft constitution elevates the provisions of the Social Chapter (which includes entitlements to many welfare benefits) to the status of inalienable constitutional rights (place cursor here to read note). The ability of the British and Irish to sustain their justification of exclusion before the European Court will have a major impact on the character of immigration policies and the course of migration flows. If the new constitution further entrenches "social rights" liberal policy makers will be able to draw temporary comfort from the delays which will likely accompany ratification. However, such comfort would only be temporary, and would be unlikely to last long enough to be of any avail when it comes time to admit the next group of new members - including most likely Romania, Bulgaria and Croatia.

4. An appraisal

Any appraisal of polices toward migration from the new members states must start with a general evaluation of the benefits and costs of the migration itself. Clearly, from the point of view of the receiving countries, the potential benefits far outweigh the likely costs.

The wave of migration which the EU 15 stand ready to receive during the coming decade is unique. A million or more young, disciplined workers, who share a common, cultural and historical heritage, and who are not marked by any of the ethnic "color lines" that can make social integration more difficult, are preparing to join the EU 15 labor force. They will bring with them, a level of educational attainment, which though it may be low by Western European standards is higher than the average educational attainment of immigrants from North Africa, Sub Saharan Africa and the Carribbean (place cursor here to read note).

This dramatic increase in labor supply is even more significant than its numbers, suggest, because the new comers, like all migrants, will be more mobile and flexible than the existing population. In a prescient book on the importance of labor supply to Europe's PostWar Growth, C.P. Kindleberger noted, almost forty years ago, "Having already cut his roots to his native land, the foreign worker is peculiarly mobile. ..Foreign labor is highly mobile. ..It can transfer readily from firm to firm, occupation to occupation, and region to region within a country and between countries and between abroad and home (place cursor here to read note). "Europe's most conclusive example of the growth dividend of a large influx of skilled labor is the German miracle itself, and the depth of its debt to the repatriation of German refugees from the East.

Not surprisingly, the greater horizontal mobility of immigrants also correlates with greater vertical mobility, over time. The social and economic upward mobility of immigrant populations tends to exceed that of long time resident populations. Though recent immigrants tend to work for lower average pay levels than long-time residents, after twenty years in a country, established immigrants, tend to work for higher average pay levels than their native counterparts. Sinn et al. (2003) note this pattern in the life earnings profiles of the cross section of immigrants working in Germany in 1997 (place cursor here to read note). If anything, this pattern may become more marked in the future than it has been in the past, as intensifying technical change and global competition increase the sectoral variability of growth.

In short, admitting a large supply of qualified workers entails on the part of the host country an initial investment. Schools, hospitals and housing must be built, and public services expanded. This is much of what the initial fiscal deficit, calculated by Sinn et al. (2003) pays for. But in time, the addition to the work force, because it is flexible and motivated, produces a growth dividend which raises living standards for all of the population, residents and immigrants alike.

6. In favor of modest, qualified restraints

If immigration from the new member states has the potential to be as beneficial in the long term as I have argued, care should be taken not to obstruct it. Delaying the arrivals for seven years would clearly be a mistake. Both of the concerns that have been elicited by the immediate prospect of new arrivals, would be better addressed by other measures than delay.

If labor market rigidities threaten to channel a large share of the immigration into unemployment, the correct response is to dismantle the rigidities, not stop the immigration.

If some tax financed welfare benefits distort location incentives and the incentive to work, those distortions should be corrected. The inefficiencies they generate have the potential to be as great for the native population, and for non-nationals coming from the EU-15 as they are for immigrants from the CEEC 10. In short, proposals to transform some welfare provisions into workfare provisions are more promising candidates for the agenda for reform than extended immigrant exclusions.

This being said, the frictional costs of adjustment can be significant, and phasing in what would otherwise be a large surge of immigration and large demands on the budget, may have merit. Smoothing the potential arrivals of the first three years over a five year period could, while keeping the pressure on, allow a little time for reforms. Similarly, the British decision to withhold unemployment compensation and public housing for at least a year seems to provide a minimal, reasonable safeguard against what Home Secretary Blunket has called "welfare tourism". In both cases, however, gradualism should not become an excuse for avoiding reforms. The objective should be flexible procedures of admission, not damns to stop the flood.

Thank you.