A persistent complaint about the H‑1B visa program for skilled foreign workers is that employers use it to replace U.S. workers, leading many to fail to find jobs. Of course, it is true that many U.S. college graduates fail to find jobs that use their education, but the fact is that the prevalence of this phenomenon has remained roughly constant for 30 years, even as the number of H‑1B visas issued has escalated.

The data for this post comes from the Federal Reserve Bank of New York’s analysis of the Census Bureau, Bureau of Labor Statistics, and Department of Labor data.

Figure 1 compares H‑1B visa issuances to underemployment among college graduates for each year from 1990 to 2019. There’s not a statistically significant correlation between H‑1B visas and underemployment among recent college graduates nor among all college graduates. The overall underemployment rate fluctuated within a range of just 2.9 percentage points over 30 years. The underemployment rate was 34 percent in 1990 when the H‑1B premiered and 34 percent in 2019 after three decades of use. The rate has varied more widely among recent college graduates, but there’s still no statistically significant correlation there (and the coefficient is negative anyway).

These facts will not stop opponents of the H‑1B visa from claiming that H‑1B workers have some more subtle effects that this overall metric demonstrates. But the knowledge that the significant amount of underemployment among American college graduates entirely predates the H‑1B program should lead policymakers to realize that underemployment is a persistent problem, and not one that they can easily fix by nixing the H‑1B.