In case you’re starved for weekend conversation starters: New paper on California immigration and wages shows zero historical correlation in the same education-experience group.
Rethinking the Area Approach: Immigrants and the Labor Market in California, 1960-2005.
I show that a CES production-function-based approach with skill differentiation and integrated national labor markets has predictions for the employment effect of immigrants at the local level. The model predicts that if I look at the employment (rather than wage) response by skill to immigration in a state, I can estimate the substitutability-complementarity between natives and immigrants. This allows me to infer, other things constant, how immigrants stimulate or depress the demand for native labor. I also use a novel instrument based on demographic characteristics of total Central American migrants or of the Mexican Population to predict immigration by skill level within California. Looking at immigration to California between 1960 and 2005 my estimates support the assumption of a nationally integrated labor market by skill and they support the hypothesis that natives and immigrants in the same education-experience group are not perfectly substitutable. This, in turn, explains the counter-intuitive fact that there is a zero correlation between immigration and wage and employment outcomes of natives. [Emphasis mine]
Source: Peri G. Rethinking the Area Approach: Immigrants and the Labor Market in California, 1960-2005. NBER Working Paper 16217. 2010. Available at http://www.nber.org/papers/w16217.