Peri’s at it again. Now, he and co-author Ottaviano look at immigration’s impact in Germany in the 90’s (pdf). Germany is an interesting case because much of the immigrants were more highly educated than in the U.S. case. Their results:
[W]e find that the substantial immigration of the 1990’s had no adverse effects on native wages and employment levels. It had instead adverse employment and wage effects on previous waves of immigrants. This stems from the fact that, after controlling for education and experience levels, native and migrant workers appear to be imperfect substitutes whereas new and old immigrants exhibit perfect substitutability. Our analysis suggests that if the German labour market were as `flexible’ as the UK labour market, it would be more efficient in dealing with the effects of immigration.
The last sentence is an implication of the following story: if a labor market was perfectly flexible, wages there would adjust immediately following a supply shock (e.g. a large flow of immigration) and displaced workers would immediately find employment (or they would take lower wages). In other words, there would be no employment effects from labor supply shocks, only wage effects.
Peri and crew run the following thought experiment: eliminate employment effects by changing wages (down for substitute groups like other immigrants and up for compliment groups like natives). To eliminate employment effects (and thus simulating a flexible labor market) wages would have to be about 260 million euros lower in a particular year. In that same year, unemployment benefits for the workers displaced by immigration cost the German state 310 million euros. In other words, the German labor market inefficiency (cushy unemployment insurance) cost the Germans about 50 million euros.
This is a small effect because immigration has a small effect on the labor market. New immigrants really only displace old immigrants, not natives, and even so only one old immigrant becomes unemployed for every 10 new immigrants that become employed.
It should be pointed out, though, 50 million euros is about 15% of these expenditures in the unemployment insurance program. This does seem like significant waste relative the size of the expenditures.