Searching for the Effect of Immigration on the Labor Market
NBER Working Paper No. 5454
We compare two approaches to analyzing the effects of immigration on the labor market and find that the estimated effect of immigration on U.S. native labor outcomes depends critically on the empirical experiment used. Area analyses contrast the level or change in immigration by area with the level or change in the outcomes of non- immigrant workers. Factor proportions analyses treat immigrants as a source of increased national supply of workers of the relevant skill. Cross-section comparisons of wages and immigration in the 1980 and 1990 Censuses yield unstable results casting doubt on the validity of these calculations. Analyses of changes over time for various education groups within regions give negative estimated immigration effects, which increase in magnitude the wider the area covered. Factor proportions calculations show that immigration was somewhat important in reducing the relative pay of U.S. high school dropouts during the 1980s, while immigration and trade contributed much more modestly to the falling pay of high school equivalent workers. The different effects of immigration on native outcomes in the area and factor proportions methodologies appear to result from the diluting effect of native migration flows across regions and failure to take adequate account of other regional labor market conditions in area comparisons.
Published: American Economic Review, 86, pp.246-251, May 1996.