AI Generated Image (not the people involved in this story)
A new study on high-skill immigration wages concludes that the H-1B and OPT programs have pushed down earnings for American workers in the fields where foreign graduates cluster most heavily. Economist George J. Borjas, in a National Bureau of Economic Research working paper issued Nov. 16, estimates that natives with science and engineering degrees earn about 3.5 percent less because of temporary high-skill programs, and that those with master’s degrees in computing or electrical and computer engineering earn about 16 percent less.
The timing gives the paper weight beyond the academic journals. It lands the same day the Labor Department proposed a sweeping PERM overhaul, while a prevailing wage final rule sits at the White House and covered H-1B petitions remain subject to the $100,000 payment imposed by Proclamation 10973. Defenders of those measures now have a fresh number to cite.
What the Borjas study claims, in plain terms
The paper, “The Labor Market Impact of Temporary High-Skill Immigration Programs” (NBER Working Paper 36014), starts from a count: about 1.2 million foreign workers held jobs through temporary high-skill programs in 2019, including H-1B and Optional Practical Training. Borjas identifies likely temporary workers in the Census Bureau’s American Community Survey and measures how heavily they are concentrated in each combination of education level and college major.
Natives in the most exposed groups, he reports, saw slower earnings growth and a relative decline in full-time, year-round work. The headline figures are the 3.5 percent and 16 percent losses. Against those, the paper estimates that production complementarities raise average wages by about 1.4 percent, which offsets only part of the loss for the most exposed workers. The abstract’s bottom line is blunt: “Absent externalities, the programs mainly redistribute earnings.”
The novelty is a “stayer multiplier.” Many temporary workers stay after their status ends, and surveys cannot identify them. Borjas’s model argues that the estimated effect of current temporary workers already captures those earlier stayers, so the full impact can be recovered without tracking them.
Two 16 percent figures, two different questions
“16 percent” has now been attached to Borjas twice this year, measuring different things. His February paper, “The H-1B Wage Gap, Visa Fees, and Employer Demand” (NBER Working Paper 34793, since published in Labour Economics), estimated that H-1B workers earn 16 percent less than comparable natives and put the revenue-maximizing visa fee at $95,000 to $152,000. That was a claim about what H-1B workers are paid. The new paper is a claim about what happens to Americans’ pay in the fields those workers enter.
The earlier paper drew pointed pushback. In a Feb. 21 analysis, economists at the Economic Innovation Group, a Washington think tank, argued that Borjas compared H-1B wages from 2020 to 2023 with a 2023 native benchmark and matched the two groups by different geographies. Their preferred model produced a gap of 5.1 percent. EIG added a note after speaking with Borjas, who told the group he does adjust for inflation; EIG said it still could not replicate his result. Expect the new paper’s identification strategy to face the same scrutiny, since NBER working papers are not peer reviewed before release.
The opposing evidence is also recent. A peer-reviewed study of hedge funds that used the random H-1B lottery found that funds winning more visas earned higher returns and posted more jobs for domestic workers in complementary fields, though fewer for domestic STEM workers. Jeffrey Miron of the Cato Institute, a libertarian think tank, wrote on Nov. 13 that the research shows skilled immigrants “add value in finance.” The two findings are not necessarily in conflict: one measures gains to firms, the other measures who absorbs the competition.
Why OPT may be the bigger policy target
Most of the political fire has gone to H-1B, but the study folds OPT into the same supply measure, and that matters. H-1B employers must file a labor condition application and pay at least the prevailing wage. OPT and STEM OPT carry no prevailing wage requirement and no annual cap. A finding that the combined inflow depresses pay for computing master’s graduates gives ammunition to proposals that would raise the cost of OPT, such as the OPT fee rule colleges have warned against. Universities, facing a sharp drop in new international students, are likely to argue that post-graduation work is what keeps their programs competitive.
For employers, the more immediate stakes are regulatory. Agencies writing wage rules must justify them with evidence, and a working paper by a former senior policy adviser to the President’s Council of Economic Advisers, a role Borjas held in 2025 according to his earlier NBER paper, is the kind of evidence likely to appear in a regulatory preamble. Opponents will press the counter-research into the comment record. How DOL and DHS weigh the two in final rules, and whether courts later find that weighing reasonable, is the open question for 2027.
This article is for general information and is not legal advice.
Sources
- George J. Borjas, “The Labor Market Impact of Temporary High-Skill Immigration Programs,” NBER Working Paper 36014 (November 2026)
- George J. Borjas, “The H-1B Wage Gap, Visa Fees, and Employer Demand,” NBER Working Paper 34793 (February 2026, revised August 2026)
- Economic Innovation Group, “The Flawed Paper Behind Trump’s $100,000 H-1B Fee” (Feb. 21, 2026, updated Feb. 23)
- Cato Institute, Jeffrey Miron, “Limits on High-Skilled Visas Harm Domestic, Non-STEM Workers” (Nov. 13, 2026)
- Proclamation 10973, Restriction on Entry of Certain Nonimmigrant Workers (Sept. 19, 2025)