The Effect of Foreign Labor on Native Employment : A Job-Specific Approach and Application to North Carolina Farms

Economists have measured the effects of immigration on native employment primarily with exogenous shifts in the foreign labor supply curve. This paper suggests an alternative, occupation-specific approach: directly describe, for one job, the native labor supply curve. The paper applies this method to seasonal farm work in North Carolina, and uses two natural experiments to estimate native labor supply. The first natural experiment uses a legal requirement for farmers to demand native workers as perfect substitutes for foreign workers, describing the level of native labor supply. The second natural experiment uses the spike in U.S. unemployment during the 2007-2008 economic crisis, describing the local slope of native labor supply. The level and slope of native labor supply to this job at both extensive and intensive margins are nearly zero. This identifies two effects of foreign labor supply on native employment: a direct effect (close to zero) and an indirect effect (positive) via consequent increases in sectoral output and its multiplier effects. The paper estimates that one U.S. job across all sectors of the North Carolina economy is created by 1.5–2.3 foreign seasonal farm workers in the short run (Leontieff production), and by 3.0–4.6 foreign seasonal farm workers in the long run (Cobb-Douglas production)