75% of H-1B hires are outside the Top 25 firms, and the wage gap for those firms is -18.5%. Firms hiring just one H-1B worker show a gap of -22.2%. Underpayment is systemic – not limited to a few bad actors.
| Firm | Avg. H-1B Salary | Wage Gap vs. Americans |
|---|---|---|
| Tech Mahindra | $86,600 | -34.8% |
| HCL | $86,800 | -32.2% |
| Wipro | $78,900 | -26.8% |
| Capgemini | $97,600 | -21.1% |
| Infosys | $82,700 | -13.7% |
| Tata Consultancy | $83,600 | -11.5% |
| All Other Firms (75% of hires) | $97,900 | -18.5% |
| Occupation | % of H-1Bs | Avg. Native Pay | Wage Gap |
|---|---|---|---|
| Software Developers | 38.3% | $146,900 | -29.8% |
| Computer Programmers | 2.9% | $109,500 | -30.0% |
| Financial/Investment Analysts | 1.8% | $130,700 | -40.0% |
| Mechanical Engineers | 2.2% | $104,000 | -22.7% |
| Electrical/Electronics Engineers | 3.0% | $113,100 | -14.0% |
H-1B workers change jobs at just 9.4% annually, compared to 20–25% for comparable American workers. This restricted mobility gives employers market power to suppress wages. Research on similar contractual restrictions shows they reduce wages by 4–14%.
Employers classify 80%+ of H-1B workers below the median prevailing wage level. But actual skill analysis reveals nearly 75% are above the median of the native skill distribution. Amazon classified 61% of its H-1B workers at the lowest wage level; Microsoft classified 53% there. The system designed to protect American workers is being systematically gamed to justify lower pay.
The H-1B program is not filling a talent gap – it is providing a captive, underpaid workforce that depresses wages for American workers in high-skill occupations. The program transfers wealth from American workers to employers: firms pocket the savings while comparable Americans earn less or are displaced entirely. Software developers alone face a 30% wage disadvantage when competing against H-1B hires. Employers would pay $200,000+ per visa and still hire – proving the program is driven by cost savings, not genuine need.
You can read the study here.