Wages at the Floor: Bunching in the H-1B Program

• In Event: Rethinking Guestworker Migration Policy: Empirical Analyses of the H-1B Visa Origins and Labor Market Impacts

Abstract

How strongly do H-1B sponsoring employers adjust wages to satisfy prevailing wage requirements? The answer is central to evaluating proposals that would raise H-1B wage floors and shift visa allocation toward higher-paying jobs, yet existing policy simulations typically hold employer wage-setting fixed. We measure firms' responses directly using the distribution of wages reported on H-1B filings under the current regime. 

Current H-1B rules create a sharp wage floor. For each Labor Condition Application (LCA), the Department of Labor requires employers to certify a wage at or above a prevailing wage determined by occupation, location, and statutory wage level. Filings with offered wages below this threshold are denied. The rule produces a spike of wage offers at the floor, and we measure its excess mass above a smooth counterfactual density.

We merge USCIS I-129 records with Department of Labor LCA disclosures for 283,000 approved cap-lottery new-employment H-1B petitions filed by over 50,000 firms in fiscal years 2021 through 2024. Using a bunching design, we estimate excess mass at the prevailing wage floor overall, by statutory wage level, and by H-1B-dependent employer status.

In preliminary analysis, wage offers concentrate at the prevailing wage floor. About one-third of approved petitions report wages within 1 percent of the applicable prevailing wage, and nearly one-half fall within 5 percent. Concentration is strongest at the lower statutory wage levels, consistent with the floor binding more firmly when it sits lower in the cell's occupational wage distribution. Petitions filed by H-1B-dependent employers (about a third of the sample) concentrate at the floor substantially more than petitions filed by non-dependent employers, with a pooled spike-density ratio roughly three times larger at dependent firms.

Our contribution is to document how strongly prevailing wage rules shape wages reported in the H-1B program. The reduced-form bunching estimates provide an empirical input for evaluating proposed reforms, allowing policy simulations to be disciplined by observed bunching rather than fixed-behavior assumptions. More broadly, the paper applies bunching methods to an administratively set wage floor in a high-skill labor market, with direct relevance for immigration and labor-market policy.

Author

• Jesús Villero, University of Pennsylvania
  Presenting Author