At a CEPR discussion meeting, held jointly with the Anglo-German Foundation for the Study of the Industrial Society in Bonn on 29 May 1996, Klaus F Zimmermann discussed the results and policy implications of CEPR’s Research Programme on European Migration. Klaus F Zimmermann is Professor of Economics at Universität München, Director of SELAPO, its Labor and Population Research Group, and Co-Director of CEPR's Human Resources Programme.

In the view of the German government, Germany is not an ‘immigration country’. This is at odds with the fact that immigration has always been significant, in both relative and absolute terms. For the last 50 years, there were explicit immigration policies in Germany, for example the integration of ethnic Germans, the guest-worker programme of the 1960s, the liberal asylum rules of the German Basic Law, and the open labour immigration policies with Eastern Europe in the 1990s. After 1950, the average inflow per year was 700,000 people – not much less than in the (much larger) USA. In 1993, about 900,000 migrants entered the USA, 0.4 % of the resident population. In the years after unification, this rate was 1.4% per year in Germany.

In the past few decades, migration has been largely beneficial in economic terms. In the 1960s, German growth rates would have been substantially lower without the immigrants – by about 4 percentage points on average. In the 1990s and after, Zimmermann argued, it is even more important to strengthen the economic incentives of immigration policies. Given the current problems with unemployment, social security and pension schemes, a rational labour immigration policy must be selective to avoid substantial losses. If the disadvantaged segments of the labour force increase by the size of the current immigrant labour stock, unemployment will rise and natives may accrue income losses of up to 5% of national income. Gains to the natives from selective immigration of the same magnitude would be about 2–4% of national income.

Attracting ‘productive’ immigrants is still beneficial – both in the short and long run. Potential labour market performance is the central criteria of an economic immigration policy. Research suggests that auctions among bidding firms are the most effective way of identifying profitable labour market segments for temporary immigration policies. A permanent immigration policy, however, must deal directly with the immigrants. A point system with a quota similar to the Canadian approach works best only in a world of certainty about the required individual profiles. Otherwise, auctions among potential immigrants are superior if one can exclude individuals from countries with higher rates of return to qualifications than in the receiving country. If returns cannot be identified, a point system may be better at avoiding potential adverse selection of immigrants through the auction, who come to reap social benefits in the host country. A parliamentary committee should thus specify the yearly permanent immigration quota and exclude sending regions with potentially adverse selection problems. Lastly, immigrants would be chosen through an auction which identifies the value of immigration. Immigrants reveal their potential gains from migration if they have to pay the highest rejected bid.