Study: Immigrant Waves Cut Job Openings

New research contributes to on-going debate on immigration’s labor market impact

Getty Images
• May 11, 2018 12:00 pm


Bursts of immigration can cut the number of job opportunities available to previous residents, a recently released paper argues.

Paper authors Jason Anastasopoulos, George Borjas, Gavin Cook, and Michael Lachanski take advantage of a commonly used index of job openings to track the way in which several bursts of immigration from Cuba affected employment in the city of Miami. They find across three waves of immigration that job openings declined for several years following each wave, indicating that the arriving migrants are filling jobs previously available to native-born individuals.

"Our evidence consistently indicates that immigration-induced supply shocks are typically followed by a short-run period of slackness in the local labor market, as measured by the number of advertised job openings. The labor market, however, tends to recover after a few years," they conclude.

They further find that the jobs being advertised and subsequently filled tend towards the lower end of the skills spectrum. This is a byproduct of the low-skilled or skilled-discounted status of the immigrants who arrived from Cuba in the analyzed timed periods.

These findings may not seem surprising to the average reader, but in fact contribute to an ongoing academic controversy over whether or not low-skilled immigration has any impact on wages or employment. Indeed, the choice to analyze Miami is no accident, as the immigrant-filled metropolis has been at the center of the immigration debate for almost 30 years.

The story starts with the so-called Mariel Boatlift. Between May and October of 1980, 125,000 Cuban immigrants fled the Castro regime, departing Cuba’s Mariel harbor to the United States. There, they settled in the port city of Miami.

The boatlift, which increased Miami's labor force by approximately eight percent, also made for what economists call a "natural experiment." The sudden increase in the immigrant population (a "supply shock") allowed researchers to study the way immigration affects labor market conditions.

In 1990, Princeton economist David Card published a study that did just that. Conventional wisdom would argue that an increase in the supply of low-skilled labor should lower its cost, i.e. the wages of low-skilled workers. But this, Card suggested, was not the case. Rather, "the Mariel influx appears to have had virtually no effect on the wages or unemployment rates of less-skilled workers, even among Cubans who had immigrated earlier."

Card's work has been hugely influential in the debate over immigration's labor market impact, cited by more than 1,500 of Card's peers. Proponents of more immigration routine point to the study as evidence that fears of immigrants depressing wages and taking jobs are overblown.

The Mariel boatlift study has also attracted criticism, primarily from George Borjas, America's most-credentialed immigration skeptic and a co-author on the new study. Borjas concluded in his own analysis that while general wage levels held steady, the boatlift reduced the wages of high-school dropouts specifically by between 10 and 30 percent. (Borjas's approach has since received criticism.)

This back-and-forth, the new paper's authors argue, misses the point, insofar as the wage impact of the boatlift depends on which subpopulation is analyzed.

"By the time this debate runs its course, the presumed impact of Mariel will depend on the reader's [emphasis original] choice of which of the studies best represents the true wage trend in Miami, and there will be a menu of choices for that reader to pick from. This menu will allow some to argue that Mariel had a substantial impact on wages while simultaneously allowing others to argue that Mariel had no impact," they write.

Instead, Anastasopoulos et. al. turn to a different measure: the number of job vacancies in Miami pre- and post-Mariel, as compared against other cities. They capture this statistic using the Conference Board’s Help-Wanted Index (HWI), a compilation of the number of help-wanted ads in 51 metropolitan areas' major newspapers between 1951 and 2010. According to Anastasopoulos et. al., the HWI has never before been used to analyze immigration's impact on labor market, in spite of the index's wide use in economic research generally.

The HWI does not allow the paper's authors to measure wage levels, but it has been established as robustly related to employment rates—and, they argue, is an even better proxy for the low-education employment rate. That means measuring the fluctuations of the HWI in relation to labor supply shocks gives a good sense of how influxes of immigrant labor affect the overall supply of jobs, especially of low-skilled jobs.

The paper looks at fluctuations in the HWI for three different immigrant waves: the Mariel boatlift, as well as the wave of Cuban refugees fleeing to the United States following the establishment of the Castro regime, and another wave of refugees in 1995. The changes in the HWI in Miami are compared to a "synthetic" control group based on the other 50 cities in the HWI, in order to give a sense of how much Miami's HWI varied from the norm when a wave of immigrants arrived.

The results are striking. The HWI fell about 25 percent compared to the control in the two years following Mariel; by 1985, it had fallen 40 percent. For the 1960s drop, the HWI had fallen 27 percent compared to the control as of 1963. The mid-90s supply shock, finally, cut the HWI by 15 to 20 percent compared to the control.

The HWI eventually returned to pre-wave levels in the 1960s and 1980s, although it did not for the 1995 wave. Anastasopoulos et. al. attribute this either to "a steady (and increasing) flow of Cuban immigrants after the 1994-1995 shock" or to the transition from paper to digital help-wanted ads.

The authors generalize their findings across the other cities captured by the HWI (although the available immigration data for many of those cities is mixed at best). Their findings hold in those cities, too, as they find a negative, cross-city correlation between the HWI and the number of immigrants entering a local labor market.

What implications does all of this have for immigration policy? The answer remains unclear. That immigrants are filling jobs previously available to natives does not mean, necessarily, that immigration depresses wages or "steals jobs," so to speak.

However, Anastasopoulos et. al.’s work does at least contribute to an ongoing, and important, debate. If their analysis is correct, immigration shocks do have a measurable impact on employment, calling into question the idea that immigration can increase the supply of jobs or will have no impact on native workers' access to the labor market.

Published under: Immigration