- Immigration
In February and March of this year, a battle royal emerged among economists about measuring the wage gap between foreign professionals here on H-1B visas and US workers. Researchers at the Institute for Progress have found that 53.6 percent of H-1B workers are paid less than the median wage that American workers receive with the same education, experience, occupation, and location.
Yet, looking at averages across the H-1B program may not be the most useful approach. According to the law Congress enacted, H-1B status should be reserved for foreign-born professionals who are being selected because of their skills and talents. Each must receive at least the median wage for Americans with the same education and experience in the same occupation and geography. Whether there is an average wage gap across the H-1B program is not the key issue.
H-1B Program-Wide Averages Aren’t the Policy Key
At its core, interest in H-1B wage gaps reflects two opposing visions of what the H-1B program actually is. The first is a view that H-1B visas are a tool to recruit exceptional global talent that strengthens American competitiveness and innovation. The second view is that these visas allow employers to undercut American workers by substituting cheaper foreign labor for equally qualified domestic workers.
But both of these visions are true for different parts of the H-1B program. Some employers are trying to use H-1Bs to hire the world’s most exceptional talent and others are using it for wage arbitrage.
The existence and scale of an average wage gap is relevant to the Trump administration’s effort to set a “revenue maximizing” H-1B payment obligation—as opposed to one that maximizes, say, innovation—such as that identified in a September 2025 presidential proclamation creating a new $100,000 H-1B entry payment for some H-1B petitions.
But for actually preventing wage arbitrage and reserving H-1Bs for workers with exceptional skills (two other directives in the presidential proclamation), it is key to understand where individual H-1B workers have a wage gap.
Policy Framing
As far as designing policy to implement the H-1B statute Congress enacted (which includes no H-1B visa taxation and which is unlikely to change any time soon), it must be noted that an average wage gap across the entirety of the H-1B program is not most informative. Instead, the question should be: Is there a systematic approach that can ensure each individual H-1B professional a US employer seeks to hire is indeed paid at least the wages a similar American would receive for that work? Answering that question would best position the H-1B visa category as a tool to attract and retain high-skilled immigrants to the US workforce so they contribute to US innovation, productivity, and progress in critical fields—while making fiscal contributions as well as avoiding underpayment compared to similar American professionals.
The Most Salient Information to Reform the H-1B Program
Averages across all US employers using the H-1B program distort and even conceal the actual solutions to avoid wage gaps, as they fail to disaggregate and therefore show the differences among the many different types of employers, sectors, and occupations that use the H-1B category. These include cap-subject employers whose H-1B petition data is reflected in the data we obtained through the Freedom of Information Act (FOIA). These employers include start-ups; non-profits not engaged in research; big multinational tech companies; small and medium sized enterprises in manufacturing, tech, and a multitude of other ventures; big Wall Street firms; rural employers; large multinational outsourcing firms; large professional services organizations; small and medium sized staffing, outsourcing, and consulting firms; and many others.
For FY2026, for example, US Citizenship and Immigration Services received H-1B petition registrations for cap-subject H-1Bs on behalf of approximately 336,000 unique individuals from about 57,600 unique employers—for 85,000 slots. Historically, cap-subject H-1B visas have included a large segment allocated to outsourcing and staffing firms that provide consulting services explicitly based on wage arbitrage, providing services to the US Fortune 500 at a cheaper rate than it would cost those client firms to recruit and hire the right staff for certain services, including temporary project work. Cap-exempt employers are primarily colleges and universities, teaching hospitals, nonprofit research institutions, and government research entities, which disproportionately use the H-1B program to hire nurses, physicians, faculty, and researchers.
Averages across the H-1B program, therefore, reflect very different employer models and missions, with some more or less likely to undercut wages for similarly employed Americans. Thus, such averages likely should not be sufficient for policymakers trying to perfect the H-1B visa category.
Instead of looking solely at H-1B program-wide totals, researchers at the Institute for Progress have reviewed and analyzed FOIA data made public by Bloomberg (on H-1B initial cap-subject petitions including I-129 forms filed to DHS) and compared it to the Census Bureau's ACS data (American Community Survey) and other administrative data. Experience benchmarking reveals that the current DOL (Department of Labor) regulations require employers to identify prevailing wages in the US labor market based on their identified minimum job requirements and not based on the specific credentials of the foreign-born professional they want to hire. H-1B dependent firms, those with a significant percentage of their US workforce in H-1B status, are more likely to use the H-1B program to hire experienced, older workers recruited directly from abroad. This rule allows them to comply with the DOL requirements while still providing services more cheaply than if hiring experienced, older Americans.
Experience benchmarking results vary greatly by occupation and by firm type. For example, 58 percent of H-1B professionals working as operations research analysts get paid more than the median wage that similar Americans receive with the same education and experience in the same occupation and location, while that's the case for only about 23 percent of H-1B professionals employed as business intelligence analysts. H-1B dependent firms have a median difference of paying about $9,700 less annually for their H-1B staff compared to native workers with the same qualifications while all other firms have a median difference of paying approximately $2,600 more in annual salary to each H-1B worker compared to Americans with the same qualifications.
Economists Weigh In
Economists can now wade into the debate by merging data from the American Community Survey (ACS) from the Census Bureau with the FOIA data and other administrative data, allowing sophisticated comparisons to similar Americans that avoid reliance on occupational median wage figures. In early February, George Borjas, a leading immigration economist who served on President Trump’s Council of Economic Advisers in 2025, published a working paper concluding there was a 16 percent H-1B wage gap, based on exactly this type of complex analysis. Borjas concludes, “The average H-1B worker earns about 16 percent less than an American worker with the same education, age, gender, occupation, and who works in the same locality” and that “the revenue-maximizing fee is between $118,000 and $264,000” per H-1B petition.
Later in February, however, Adam Ozimek, the chief economist at the Economic Innovation Group (EIG), and Jiaxin He responded to the Borjas study with a short paper saying that while the analysis might have “influenced the Trump administration’s $100,000 H-1B fee policy” announced in September 2025, the Borjas “findings result from substantial data errors” and that the H-1B wage gap is instead quite small, about -5 percent. The EIG analysis points out that while “this still represents a non-zero difference in pay, the overall effect obscures significant underlying patterns worth examining,” like the fact that the data shows younger H-1B lottery winners, such as those obtaining H-1B status after F-1 student status, outearn their American peers.
In mid-March, the Centre for Research and Analysis of Migration (CReAM) published a discussion paper by Michael Clemens, another heavily cited immigration economist, detailing how Borjas's paper reflects four methodological choices that inflate the wage gap, finding on average the wage differential in the H-1B program is likely "statistically indistinguishable from zero." The Clemens paper documents that the wage gap identified by Borjas arises from four methodological choices: erroneously assigning education levels to more than a third of H-1B workers in FY2023–24; the unexplained pooling of four years of H-1B data against a single year of US native data rather than straightforward within-year comparisons; a nonstandard definition of local labor markets using Census PUMAs (Public Use Microdata Areas), units so narrow they are inconsistent with DOL’s “normal commuting distance” definition of area of employment that controls prevailing wages for the H-1B category; and failing to consider the gap in employer tenure between average native workers and new H-1B employees. Clemens finds that “correcting these choices, sequentially and collectively, accounts for the entire discrepancy between the 16 percent estimate and the 1.2 percent estimate obtained here.”
In response, Borjas posted a revised paper suggesting the average H-1B wage gap remains quite large, so perhaps the debate among economists will continue.
Asking the Right Questions to Improve Outcomes
Despite evidence there may not be much of an H-1B wage gap on average, the debate is an answer to the wrong question. The emphasis should be on constructing policy solutions to boost innovation and productivity among US businesses under the current law. By the end of March, we expect DOL will publish a proposal for a new H-1B prevailing wage methodology. If adopted, experience-benchmarked wage requirements would be tied to the credentials of sponsored H-1B professionals, allowing the executive branch to ensure that no foreign professional obtains H-1B status if the employer intends to pay less than what Americans with the same education and experience are paid.
It’s time to reframe the questions so policymakers can improve the H-1B program, the nation’s flagship visa category for employing high-skilled immigrant talent.
Amy M. Nice is a distinguished immigration fellow at Cornell Law and distinguished immigration counsel at the Institute for Progress.
