The Effect of Immigration on Productivity: Evidence from US States
A one percent increase in employment in a US state, attributable only to immigration, is associated with a 0.4-0.5 percent increase in income per worker in that state.
Immigration during the 1990s and the 2000s significantly increased the presence of foreign-born workers in the United States, but the increase was very unequal across states. In The Effect of Immigration on Productivity: Evidence from US States (NBER Working Paper No. 15507), NBER Research Associate Giovanni Peri analyzes state-by-state data to determine the impact of immigration on a variety of labor market outcomes, including employment, average hours worked, and average skill intensity, and on productivity and income per worker.
Peri reports a number of distinct findings. First, immigrants do not crowd-out employment of (or hours worked by) natives; they add to total employment and reduce the share of highly educated workers, because of their larger share of low-skilled relative to native workers. Second, immigrants increase total factor productivity. These productivity gains may arise because of the more efficient allocation of skills to tasks, as immigrants are allocated to manual-intensive jobs, promoting competition and pushing natives to perform communication-intensive tasks more efficiently. Indeed, a measure of task-specialization of native workers induced by immigrants explains half to two thirds of the positive effect on productivity.
Third, Peri finds that inflows of immigrants decrease capital intensity and the skill-bias of production technologies. The decrease in capital intensity comes from an increase in total factor productivity; the capital-to-labor ratio remains unchanged because investment rises coincident with the inflow of immigrants. The reduction in the skill-intensity of production occurs as immigrants influence the choice of production techniques toward those that more efficiently use less educated workers and are less capital intensive.
Finally, Peri finds that for less educated natives, higher immigration has very little effect on wages, while for highly educated natives, the wage effect of higher immigration is positive. In summary, he finds that a one percent increase in employment in a US state, attributable only to immigration, is associated with a 0.4 to 0.5 percent increase in income per worker in that state.
A central challenge in establishing a causal link between immigration and economic outcomes is the fact that immigrants may be disproportionately attracted to states with strong economic performance. Peri recognizes this problem, and uses information on state characteristics, such as the location of a state relative to the Mexican border, the number of ports of entry, as well as the existence of communities of immigrants there before 1960 to predict immigrant inflows. He then studies how these predicted inflows, rather than actual inflows, are related to labor market outcomes. He argues that the state characteristics that underlie his predictions are not likely to be associated with either labor market outcomes or productivity. He also controls for several other determinants of productivity that may vary with geography such as R and D spending, computer adoption, international competition in the form of exports, and sector composition.
-- Claire Brunel