Monday, May 18, 2009

Immigration Is Economically Intractable

Last year over 191 million migrants were on the move worldwide (Transatlantic Trends: Immigration 2008). In 2009 “…immigration issues will again be atop the political agenda. The United States will look to a new president to pass much-needed immigration reform, and in Europe, the [former] French EU Presidency has made it clear that a common EU immigration policy is a priority. All this comes at a time when the United Kingdom is adjusting to its new points-based immigration system. Germany has put its new citizenship test into use, Poland is transitioning from a country of emigration to one of immigration, and Italy is adopting more restrictionist [sic] policies than ever before” (from the report cited above).

Intra-Europe migrations

European Union enlargement from 15 member states to 25 (2004) and then 27 (2007), and the EU emphasis on “free movement” of labour – as well as goods, services and capital – raised fears of invasion by job-seekers and “welfare-scroungers”. Only the UK, Sweden and Ireland did not take advantage of the possibility of suspending full access for a transitional period of up to 7 years.

The UK saw the largest inflow of foreign labour ever recorded, hugely exceeding all predictions; Ireland followed suit. In Poland between 2004 and 2007 the number of citizen “temporarily residing” in another EU country more than doubled, reaching almost 2 million; the Poles’ preferred destination switched from Germany to the UK (where 27% of Polish immigrants in 2007 had a University degree, up from 22% in 2003). This was reflected in Polish remittances in 2007 being 60% higher than in 2003; while in Latvia remittances were nearly three times the level of 2003. (see B. Galgòczi, J. Leshke and A. Watt, Intra-EU labour migration – Flows, effects and policy responses, ETUI-European Trade Union Institute, Working Paper 2009.03, March 2009).

Other ‘sending’ (Hungary) and ‘receiving’ (Sweden) countries experienced much less dramatic flows. Clearly the pattern of migration was distorted by transitional restrictions, with the three more open countries raising their share of Polish immigrants from 12.1% to 42.4%; only Hungarian immigrants stuck to their favourite destinations of Germany and Austria. Some of the sending countries also experienced inward migration, notably Hungary from Romania and Slovakia (Ibidem).

The key macroeconomic drivers of immigration were the wage gap and relative unemployment (especially youth unemployment). Prior to enlargement the wage gap between the EU-15 and accession countries was very large. In 2003 Latvia had an average wage, at Purchasing Power Parity exchange rates, equal to 12.9% of the EU-15 average, Poland 21.5%, Hungary 29.0%; gaps were much higher at actual exchange rates, due to relative undervaluation of eastern currencies. By 2007 the gap had shrunk but was still substantial: 18.2% for Latvia, 25.4% for Poland and 31.1% in Hungary. Unemployment rates in 2003 were 17.9% in Poland, 10.5% in Latvia, 5.9% in Hungary. By 2007 unemployment had risen to 7.4% in Hungary but fallen respectively to 9.6% and 6.0% in Poland and Latvia – against a EU unemployment rate of 7.9 in 2003 and 7.0% in 2007. Relative employment opportunities in the destination country were a very important driver. (Ibidem, pp. 14-25).

In the receiving countries the rise in labour supply due to immigration was accompanied by higher aggregate output and labour demand. Even lacking a counterfactual alternative for a comparison, the ETUI study concludes that in Austria, Germany and Sweden “the overall unemployment trend … is inconsistent with the idea of post-accession pushing up unemployment: it rose slowly until 2005; and subsequently fell. However, the unemployment trend in the UK, the country with the largest relative influx, does appear at first sight consistent with the idea of unemployment-creating immigration: the jobless rate bottomed out in 2004 and has been slowly but inexorably rising ever since”. The evidence on sectoral wage trends in the presence of high or low incidence of immigration is mixed. (Ibidem p. 25-27). The ETUI underlines the importance of welfare provisions trends in the sending countries as a determining factor of emigration, but is silent on the vexed question of the attraction of relative welfare provisions with respect to and among the countries of destination.

Economic and non-economic obstacles

Immigration raises deep emotional issues of race, ethnicity, religion, culture, language, customs – generating barriers and conflicts. Apparently 50% of Americans and 47% of Europeans in the public opinion poll conducted by Transatlantic Trends (cited) “perceived immigration to be more of a problem than an opportunity, but majorities in France and the Netherlands, as well as sizable minorities in other countries considered it to be more of an opportunity”. Interestingly, on average 7% of those respondents who were asked this question twice during the interview, at the beginning and at the end, consistently switched their view of immigration from problem to opportunity – presumably on reflection or out of concern for their image.

But even if all these controversial issues did not arise, even if immigrant assimilation was costless and instantaneous, immigration would still be economically intractable. Because even in this ideal world:

1) the redistribution of gains and losses from immigration, so that its net benefits should affect everybody positively, is either impossible or undesirable or both; and

2) immigration, like any population increase, involves the appropriation by the newcomers of some social capital. Losers in the host country are usually happy to bear this cost for the future generations of the nation’s children (and not just of their own prospective children), but resist social capital dilution for the benefit of outsiders.

Free movement of labour was remarkably trouble-free and successful in the re-unification of Germany, but only thanks to exceptional factors. The non-economic obstacles mentioned above – race, ethnicity, religion, culture, language, customs – were totally absent; they were all tall, blond, blue-eyed, and one as good or bad as the other. Massive re-distribution of gains and losses from (internal) migrations were funded by the federal government within a single state; social capital was left in the hands of the entire nation; and the economic disruption involved by internal migrations was compensated for by the joys of re-unification, national independence and the return of democracy. These are all non-repeatable factors – except for the prospective reunification of Korea.

Impossible or undesirable redistribution of gains and losses

There is no doubt that immigration is efficient, for it involves a potential net increase in universal economic welfare. But this potential benefit is the net result of gains and losses accruing to different groups of individuals, and even accruing to the same individuals in different proportions in different capacities. A potential net benefit is not enough to assert the superiority of any economic change, including the opening of borders to immigration. The losers must actually be over-compensated by winners, potential over-compensation not being sufficient to assert an improvement. When it comes to immigration – as is often the case with opening international trade and factor movements – redistribution of winners’ gains to losers is almost invariably impossible, or undesirable or both.

Imagine two countries, North and South, both – for simplicity but without loss of generality – with full employment of labour at a Northern wage higher than in the South and closed borders. Suppose borders are opened so as to allow migration from South to North. With unrestricted migration, after the time necessary to reach a new equilibrium, and neglecting travel and resettlement costs, a uniform wage level (within bounds) should prevail. Southern winners are the non-migrant workers whose wage will increase due to the fall in labour supply. Southern losers are firms that now pay higher wages and –starting from full employment – also have to reduce employment and production. Northern winners are firms that, thanks to the increase in labour supply due to migrants, now pay lower wages than they would have had to pay otherwise, and thanks to immigration can raise employment and production. Northern losers are the national workers, who get a lower wage (not necessarily lower than before, but lower than it would have been without immigration). Immigrants gain directly from being employed at higher wages.

Some of the gains and losses are national internal transfers and cancel out; all of the gain for remaining Southern workers is a loss for Southern firms; all of the loss of Northern workers is a gain for Northern firms. There is an additional loss for Southern firms, which is the surplus that they used to obtain from employing, at the lower pre-migration wage, those who now have migrated; that loss is equivalent to some (half, with linear demand and supply curves for labour) of the net gain that migrants obtain from the higher wages in the North. Finally, there is an additional gain by Northern firms, which is the surplus produced by migrants over and above their wages (except for the marginal migrant who will yield no surplus in equilibrium).

In conclusion, there is a net gain from immigration, and it accrues entirely in the North (if immigrants are considered part of the North): it consists of half the extra wage of migrants (the other half is matched by Northern firms losses) and the additional surplus obtained by Northern firms on migrants’ employment.

In principle, gains and losses could be redistributed from winners to losers and make everybody better off. But this is 1) impossible and/or 2) undesirable. If this were done nationally, in the South a transfer would have to be made from the poor non-migrants to the richer owners of Southern firms. In the North, the relatively rich owners of Northern firms could compensate Northern workers for their losses. But profit is seen as the reward of entrepreneurship and risk-taking, and the extra-profit on emigration is difficult to tap; the extra tax (if any) levied on that extra-profit is indistinguishable from other government revenues in the budgetary melting pot. Therefore the Northern extra-profit is not available for compensating Northern workers – let alone to compensate the owners of Southern firms who are also losers. And it would be undesirable to tax migrants, less poor than before but still relatively poor, in order to compensate the losses of Northern workers, let alone those of the owners of Southern firms.

International transfers are anyway unthinkable; they would require taxation and expenditure by a non-existent supranational agency, or bilateral agreements very hard to negotiate and agree between the two governments.

Therefore: no transfers internationally; no transfers from the poor to the rich which would be unfair and unjust; no transfers involving firms' profits other than through ordinary taxation. Moreover, any transfers to compensate losers who started from a monopolistic position broken by immigration is also undesirable, for it perpetuates an unfair and inefficient monopoly advantage – of capital in the South, of labour in the North.

No over-compensation of losers means stern, economically rational, opposition to immigration by national workers.

There are small oversimplifications in this picture: suppose workers have two degrees of skill, higher and lower. If the skilled migrate the unskilled wage will not rise but fall in the South, and viceversa in the North. The presence of unemployment in the South and/or the North alters the pattern of re-distribution necessary to make everybody better off in both North and South, but without making it any easier, for now Northern workers are likely to lose employment opportunities as well as wage levels if employed. Re-distribution from migrations will include remittances. As a result of immigration prices might rise in the South and fall in the North relatively to what they would have been otherwise, with people gaining or losing as consumers as well as producers. The general problem of impossible and/or undesirable re-distribution does not change.

Social capital appropriation

Every country is endowed with a certain amount of what we could loosely call “social capital”: a physical infrastructure of transport routes, schools, hospitals; public facilities like social housing, swimming pools, libraries, museums; and all the other provisions that a society makes for the future of its children. When the children arrive, they dilute that social capital, they appropriate it, but this is what it was provided for, society takes pride in providing it. Newcomers from outside have the same effect of diluting and appropriating social capital, but are regarded as intruding usurpers. Unless – as in the new worlds of the 19-th century – immigrants are needed to build up that social capital and to exploit economies of scale in under-populated territories, which is not the case today in the advanced countries that are on the receiving end of large-scale migration.

This is another source of resistance to immigration: the refusal to allow current and future generations of outsiders to take part of the social capital that has been provided and paid for by society for the benefit of future generations of national children.

Economically Intractable

These are the purely economic reasons why immigration is intractable. These reasons should be addressed as such and not mixed with less rational reasons, although all of them lead to the recommendation and adoption of unpleasant illiberal means to discourage or stop immigration.

The double-pronged resistance to immigration can be overcome either by the economic benefits of immigration being or becoming so overwhelming that opposition comes only from a silent minority, or by widespread sentiments of human solidarity prevailing over self-interest.

Immigration pressure can be reduced by a less unequal worldwide distribution of income and wealth, but this is the opposite of what has been happening in the last twenty years. Otherwise unpleasant illiberal means of dubious legality might be used to discourage it or stop it, like the “respingimento” (rejection, forced repatriation) policies of the current Italian Minister of the Interior.

9 comments:

D. Mario Nuti said...

A reader comments by e-mail: “Very good – and also very dismal. Except that: a) there are jobs that in the North nobody wants to do (dustmen in Germany, “badanti” [carers] and tomato pickers in Italy …) and for which immigration is welfare enhancing; b) social capital means something else (Putnam)”.

Thanks. Unfortunately Economics IS the dismal science par excellence. I broadly agree on both points, however noting that:

a) In the North those jobs have no takers at the low wages that immigrants are prepared to accept, but without immigration the same jobs would command higher wages and there would be willing takers and a lower demand (if any) for those services. General net welfare enhancement after re-distribution would be positive as in all other cases of immigration.

b) Robert D. Putman’s social capital is an immaterial good, made of "social networks and the norms of reciprocity and trustworthiness that arise from them", of "civic virtue ... embedded in a network of reciprocal social relations". I was referring to a different but common and older alternative definition as "social capital infrastructure". But even Putman’s social capital, just like social infrastructure, can be diluted by immigration.

Of course immigrants to their ex-colonial empires have already paid part or all or probably more than the cost of the social capital which they appropriate, through the colonial exploitation of their forbears. Unfortunately in economics bygones are bygones, but this factor should encourage sentiments of solidarity in the receiving countries.

Jon said...

This is an interesting and useful blog.

In terms of analysis, I think it helpful to divide between skilled and semi/unskilled workers as the former seem to be complements to capital while latter are substitutes. That can be important in recessions sufficiently strong as to affect investment and scrapping rates ove a 5-year or longer period. The skill dimension is also important in the political economy dimension. In the UK - and I think elsewhere in the EU - the main employee and union protests against immigrant workers have been from relatively skilled workers. They fear being undercut by lower paid (short-term?) immigrants often employed on short-term contracts. Both Polish and Italian skilled construction workers have been so targeted. (See Financial Times story of 24 May and previous similar reports).

Perhaps the most interesting issue is whether past net flows from CEE countries into 'old' EU member states will continue. When the recession initially hit the UK, there was much discussion of Polish workers going back home as unemployment was still low there, wages were reasonable and the £ had fallen relative to the Euro. But, as the recession has now hit CEE countries pretty hard, this no longer holds. Further, even if Poland withstands the recession well, the Baltic States and some others won't. Hence, it is an open question as to whether the economic incentives are for higher or lower net migration from CEE countries. The one clearcut point is that vacancies for unskilled 'old' EU country jobs will be higher and competition from the native-born rather stronger.

In the UK, 2007 data indicate lower net migration from CEE countries but with both inflows and outflows rather lower. The result was net immigration of 198,000 in 2007 compared to 204,000 in 2006 (FT, 24 May). However, these are just aggregates compiled from imperfeectly measured and very heterogeneous flow data. Some disggregated data is very necessary to get behind these.

So, where I live, the local supermarket continues to offer Polish goods and the library maintains its Polish section for the (mainly) Polish and Baltic State fruit and vegetable pickers. Whether or not the construction workers have stayed is more doubtful but I suspect that the care workers for th elderly have stayed - and will continue to stay.

D. Mario Nuti said...

Thank you Jon. I agree with the importance of the distinction between skilled and semi/unskilled immigrants, I had mentioned it briefly as one of the qualifications needed by my oversimplified picture. But I believe such a distinction does not alter the general picture of economic intractability, due to the difficulty or undesirability of redistribution of gains and losses from immigration. If anything, the pattern of required re-distribution needed to make everybody better off becomes more complicated.

You are right about the recession, now hitting Eastern Europe hard with a lag, re-establishing the attraction to emigrate for Polish nationals and now especially for the Baltic states. In the first quarter of 2009 Latvia’s economy shrunk by 18 per cent year on year and by a massive 28.7 per cent compared to the previous quarter . Lithuania contracted by 12.6 per cent year on year, Estonia by about 15 per cent – due to their fixed/hyper-fixed exchange rate regimes and the credit squeeze by foreign banks - declining at a faster rate than during their post-Soviet recession of the early 1990s. While the incentive to emigrate therefore is still there, especially to the more open countries like the UK, the actual gross and net flows are hard to predict.

ugopagano@gmail.com said...

Dear Mario, doesn't it follow from what you say that having capital (including capitalists and managers) emigrating is better than having workers immigrating?
Their demand on the social capital of the host country is negligible and, sometimes, they may even increase it!
Perhaps, what I am saying is not "politically correct" but, given that you have raised this very important political issue, we should not miss the opportunity to discuss it........

D. Mario Nuti said...

Thanks for an excellent question, Ugo. I believe that from a purely formal viewpoint capital migrations raise the same issues as labour migrations. Both involve net benefits arising from gross benefits and gross losses, with inter-national and inter-class compensations needed to make everybody better off, the difficulty or undesirability of compensatory payments creating obstacles to migrations.

But capital and labour migrations are not symmetrical:

1. Immigrant capital does not lay a claim to social capital infrastructure; indeed, as you rightly suggest, it might contribute to it;

2. Capital has no ethnic or racial connotation and therefore does not meet resistance on that ground. There may be cultural differences between national and foreign capital, such as the emphasis on corporate identity placed by Japanese foreign investors, but these differences are not paramount.

3. Foreign capital is largely welcome, both by governments and by the populations at large (except sometimes when it takes over existing national companies); national capitalists may oppose immigrant capital but they are a narrow segment of society with negligible voting power (though their political influence can still be important);

4. The case for compensating national capitalists’ losses from immigrant capital is weak, because: 4.1) capital is international by nature and it is very hard, in most cases, to link international capital to a particular nationality; 4.2) risk-taking by capitalists includes the risk of immigrant capital lowering profit rates, and there is no case for ex-post coverage of such a risk by the government; 4.3) unlike workers, capitalists can reduce their exposure to the risk of immigrant capital by diversification; 4.4) the standard of living of national capitalists is not as adversely affected by competition from immigrant capital as that of national workers is adversely affected by the competition of immigrant labour.

These asymmetries imply that capital immigration is “better” than labour immigration, as Ugo Pagano suggests, not in a normative welfare sense but only in the very limited sense that international mobility of capital meets much less resistance than that of labour. This is a factor that automatically favours Foreign Direct Investment – the form that capital migrations take (portfolio investment is different). Further encouragement of FDI is a matter for national policy, not for normative prescriptions.

Foreign Direct Investment is in any case contained by the needs of investment diversification. National risk factors stop FDI – capital migrations - well before rates of return can be equalised on a global scale. Wage equalisation world wide is stopped by various factors (travel cost, displacement cost, risk of unemployment on arrival, etcetera) as well as the hostility with which immigrant labour, unlike immigrant capital, is met in the host countries.

ugo pagano said...

Dear Mario, I would like to add that there is an externality problem. In the case of capital and manager migration the firm has to pay for the entire cost of its investment including the social costs imposed on the family of the managers and skilled works which follow the investment. By contrast if the firm uses immigrant labour it reverses the social costs on the hosting community and, quite often, on its poorest members. This externality induces an asymmetric difference between social and private costs which implies that the overflow of capital and direct investment is inefficiently lower than the flow of immigrant labour.

D. Mario Nuti said...

The asymmetry you mention is convincing. I am not sure about your bottom line, that "the overflow [I suppose you meant the flow] of capital and direct investment is inefficiently lower than the flow of immigrant labour". Inefficiency presumes that social capital should be provided entirely free of charge by the host country. Otherwise you could argue instead that labour immigration is "inefficiently higher" than it should be.

chilosi said...

I quite agree by and large with your treatment of the emigration issue, but one point needs some qualification:

"Immigration pressure can be reduced by a less unequal worldwide distribution of income and wealth"

Emigration, in particular illegal emigration can be very expensive in relation to the living standards of the countries of emigration. If those standards increase, the resources available for financing immigration increase too. As a consequence migration pressures may increase rather than decrease. In the end the relation between relative living standard and emigration can have the usual shape of a inverted U curve. Of course your interest was mainly to immigration processes inside the EU where the relevant branch of the inverted U is the descending one, but for the world as a whole, and in particular concerning the emigration from Africa to Europe, we may be rather still in the ascending branch.
A final point: there is also the usual Hecksher-Ohlin argument of trade being a substitute for factor movements, in particular migration. It is irrelevant for the EU, but it may be quite relevant for the world at large.

D. Mario Nuti said...

An increase in living standards in a poor country may encourage rather than discourage emigration by making travel costs more affordable - up to a point: granted.

Hecksher-Ohlin processes are long run, and their impact can actually be reduced or offset by short term factors.