Working Paper, Federal Reserve Bank of Chicago 2007-24
Previous study by Card and Lewis (2005) has found (puzzling) that inflows of Mexican immigrants into new metropolitan areas have had no effect on the relative wages of very low-skill (high school dropouts). Rather, Mexican workers do affect relative wages for high school graduates. Whereas Card and Lewis' study uses variations across geographies, this paper considers variations across occupations. Recognizing that Mexican immigrants are highly occupationally clustered (disproportionately work in distinctive very low wage occupations), we use this fact to motivate the empirical approach to analyze the relationship between the composition of Mexican immigrants across occupations/industries and average wages in the occupations/industries. To summarize our finding, we confirm that in spite of the fact that Mexican immigrants are disproportionately in very low skill occupations, (which we define as occupations where the average workers have no high school education), we find no significant impact of Mexican immigrants on wages in those occupations. By contrast, inflows of Mexican immigrants have some small effects on the wages of native workers in low skill occupations (which we define as occupations where the average worker has at least some high school education or is a high school graduate). These results suggest potential spill over effects as natives may be reallocating their labor supply into non-predominant Mexican occupations. An analysis of employment changes of natives into different occupation groupings in response to an inflow of Mexican immigrants, confirms that natives' employment in occupations where the average worker has a high school education increases in response to Mexican inflows in the U.S labor force from previous periods.