The White House · 8 October 2026
The press conference, with the data behind it
On 8 October 2026 the administration announced PERM suspensions for eight companies and a J-1 investigation into nine universities. The case rests on federal records that are difficult to carry in a speech. This site holds those records.
Each moment below quotes the recording and sets the measured evidence beside it. Press a timestamp to hear it said. Where the published record stops short, the page says what is missing and what would settle it — because the gaps are themselves part of the case.
Vice President JD Vance
This is a great day for American workers because we are taking some very concrete actions to cut down on some of the H-1B and J1 visa fraud that we've been tracking over the past several months.
What was announced
Three things followed.
- PERM suspensions for eight named companies.
- A J-1 investigation into nine named universities.
- A visa-fraud strike team.
They will say, maybe they'll put an advertisement in a small town newspaper... and they will use that lack of a response to then go and replace American workers.
The recruitment test
This is the PERM labour-market test: before sponsoring a worker for permanent residence an employer must advertise the role and show that no qualified American applied. The disclosure files record that a test took place. They do not record the advertisement, how many Americans applied, or why any were rejected. The Department of Labor holds that detail on the ETA-9089 and does not publish it.
When somebody comes in on an H-1B, they are completely subordinate to that company. If they lose that job, they have to leave the country.
Why the tie to one employer matters
- $15,300forgone per $100,000 of payroll, employer and employee side together
Status depends on the job, which is a fact of the visa rather than a finding in the data. What the data does show is the price attached: an employer owes no Social Security or Medicare tax on a student working under practical training, and neither does the worker.
The H-1B visa is how the people get into the United States in the first place, and then a PERM program is how those employees stay.
PERM is the step this site can see
- 42,833certified public-employer filings, FY2013–FY2024
PERM is the only federal disclosure file that records a worker's citizenship, which is why this site can say who public employers sponsor and the H-1B files cannot. The series ends with FY2024: the Department of Labor removed every worker field from the FY2025 file, 155 columns down to 137.
Public-employer permanent-residence filings by sector
Certified PERM filings, FY2013–FY2024. This is the only federal disclosure file that records the worker’s citizenship.
Who the public sector sponsors →The ask to restore those fields →
Microsoft laid off last year 6,000 American workers. At the same time, the company benefited from 6,300 H1B visas and almost 3,000 green cards.
The Microsoft figures
Layoff counts are not a federal disclosure, so that half of the comparison comes from company reporting rather than from this site. What is public is everything Microsoft filed, by year, occupation and wage level, and it can be read here directly. One caution when citing approval totals: initial and continuing approvals must never be summed, because a continuing approval is the same worker renewed and adding them counts one person once a year for as long as they stay.
These firms that help corporations apply and find H-1B visa candidates. They're commonly called the witches, the four major outsourcing firms.
The staffing firms
- 1,106employers that meet this site’s staffing-firm test
- 252,806certified filings they account for
This site identifies these firms from the disclosure files rather than from a list. Joining employers to their industry puts Infosys, Tata Consultancy, Cognizant, Wipro, HCL and Capgemini at the top of filing volume — the same firms suspended from PERM later in this press conference.
Certified filings by statutory wage level, nationally
Levels I and II are the two lowest the statute defines. They carry 61.2% of all certified filings.
If you are working as an H-1B visa in the United States of America, you're earning $20,000 less than an American citizen hired in the same position. If you were brought in by one of these foreign outsourcing firms, you're making $48,000 less.
The wage gap
The disclosure files publish the wage an employer OFFERS, not the wage paid, so the offered figures understate any gap rather than overstate it. What this site can show at scale is where the offers sit: the staffing firms named a moment later file overwhelmingly at the lowest statutory wage levels. The figure that would settle the question outright is a Treasury tabulation of W-2 wages against the prevailing wage for the same occupation and area. Treasury holds it. Nobody has published it.
There are nine universities in particular that we believe merit an investigation. Those universities are Harvard, Yale, Stanford, Brown, the University of Pittsburgh, UC Davis, Caltech, Arizona State, and MIT.
The nine universities
- 14,080postdoctoral appointees across the nine
- 8,545of them hold a temporary visa
An investigation was announced on this date. The federal postdoc survey counts what these nine institutions reported themselves, and the figures below are their own returns.
Each of the nine, by share of postdocs on a temporary visa
Combined they are 60.7%. Every one of the nine is above half.
These universities are employing J1 visas for federally funded grants at a rate of about 61%. The national average is 38%.
The nine are at 60.7%. Nationally it is 58.3%.
- 60.7%the nine named universities, share of postdocs holding a temporary visa
- 58.3%every institution in the survey
- 69,877postdoctoral appointees counted
- 48.1%of those posts are federally supported
The NSF Survey of Graduate Students and Postdoctorates puts the nine named universities at the share given from the podium. The same survey shows this is not confined to them: across every institution it covers, most postdoctoral appointments in the United States are held on a temporary visa. American citizens and permanent residents are a minority of the postdoctoral workforce nationally — in a workforce where roughly half the posts are paid for by federal research money. The nine are where the investigation starts, and the pattern runs through the whole system.
Where every institution falls, not just the nine
175 institutions reporting at least 50 postdocs. The bars right of 70% are the ones no announcement has named.
Labor Secretary Keith Sonderling
I am hereby suspending from the permanent labor certification program some of the largest IT outsourcing firms in the world. Cognizant, Infosys, Tata, Whipro, HCL, and Capgeemini. And also due to multiple active federal investigations, the Department of Labor suspending Microsoft and Adobe from the program.
Eight companies suspended from PERM
- 252,806certified H-1B filings by firms meeting this site’s staffing-firm test
- 12.2%of all certified filings in the country
- 61.2%of all certified filings sit at the two lowest statutory wage levels
PERM is the mechanism described earlier as how an H-1B holder stays permanently. Suspension stops new and pending applications from these employers. It does not affect anyone already holding permanent residence.
Employers by share of their workforce that is H-1B dependent
1,587 employers report a workforce that is entirely H-1B dependent, and 783 more are above nine tenths. The distribution is bimodal: most employers barely use the program, and a cluster runs on almost nothing else.
Attorney General Todd Blanche
This announcement today should send and does send a strong signal to every single employer in the United States that the Trump administration will not tolerate discrimination against American workers.
The enforcement frame
- 32,121permanent-residence filings by universities, FY2013–FY2024
- 7,804by public school districts
No published federal source records graduate applications or admissions by citizenship, so whether American applicants are declining, being declined, or not applying cannot be determined from the data. This site asks for that collection.
Deputy Chief of Staff Stephen Miller
He had at his rallies American workers in good paying middle class jobs with pensions who were laid off and forced to train their foreign replacements as the condition of their severance.
Training your replacement
No federal file records what happens in an office after a filing is certified. The disclosure files stop at the paperwork. Whether a displaced worker trained their replacement is exactly the kind of fact no agency collects, and that gap is itself the finding: the public record ends where the consequence begins.
Labor Inspector General Anthony D’Esposito
Subpoenas have been served, investigations are underway, and nobody will be getting a free pass because their name is carved into an expensive building.
The investigation is live
An investigation was opened, on the record, on this date. Its outcome will be whatever the investigators find. What this page supplies is the federal data the investigators and the public can both read now.
The Chinese Communist Party doesn't need to kick down the doors of America's research laboratories if our own institutions are willing to leave the doors wide open.
The research-security argument
- 72.8%non-resident share of petroleum engineering doctorates
This site measures who is in the laboratories. In several strategic fields the graduate pipeline is overwhelmingly non-resident, and in a handful of programs every doctorate awarded in the window went to a non-resident. These are counts of degrees conferred, reported by the institutions themselves to the federal government.
FTC Chairman Andrew Ferguson
America is not a jobs program for the rest of the world. We didn't build our institutions for the rest of the world. We built them for Americans.
The thesis, stated from the podium
- 708,393doctoral and master’s theses from 106 American universities
- 29,131postdoctoral appointees who are American citizens or permanent residents
This is the argument this site was built to test with federal data. The measured finding sits on one page and the argument on another, deliberately, because a reader may reject the argument and accept every figure in it.
Vice President JD Vance
Never believe the lie that Americans are not good enough to build a future for Americans.
The close
- 586,177research doctorates awarded in the window this site covers
- 106American universities whose dissertation record is published here
- 7of them publish nothing a reader can search
The last figure is the one worth acting on. No federal rule requires a university to publish a dissertation or even to list the ones it awarded, so the record of what American institutions produced had to be assembled one university at a time and remains a floor. A deposit requirement attached to federal research money would fix it.
The nature of the fraud is fundamentally that when you apply for these visa programs, you make a couple of certifications. One of those certifications is that the person that you're bringing in has a hyper specialized, hyper capable skill set. And that's clearly not the case if you're just replacing one American worker with a foreign indentured servant.
Asked directly: what is the fraud?
- 8,081,883certified H-1B applications in the disclosure files
A reporter asked whether this was illegal conduct or a legal pathway the administration objects to. The answer rests on the specialty-occupation certification. This site can show what occupations and wages were certified. Data about attempts to find American workers are not published by the relevant agencies. The recruitment an employer says it carried out before filing is attested on the form and never released, so no outside party can check it.
What was certified, by industry →The occupation and wage tables →
What the averages hide
Every share quoted from a podium is an average, and an average is the one statistic that cannot show an outlier. The national postdoc figure is 58.3%. The median institution is 60.9% — higher — because the largest universities pull the weighted average down. At the typical American institution the position is worse than the national number suggests.
- 134of 175 institutions have a MAJORITY of postdocs on a temporary visa, covering 56,990 posts
- 98sit above the national average of 58.3%
- 6are at 80% or higher
- 94.6%the highest single institution; the lowest is 5.8%
Institutions by share of postdocs on a temporary visa
Each bar is a count of institutions. The nine named sit at 60.7% combined, inside the main body of this distribution rather than at its edge — which is the point: they are not unusual. The unusual institutions are further right.
Where the distribution runs out
| Institution | Postdocs | Temporary visa | Share |
|---|---|---|---|
| Missouri University of Science and Technology | 56 | 53 | 94.6% |
| University of Texas M. D. Anderson Cancer Center, The | 595 | 513 | 86.2% |
| University of Tennessee, The, Health Science Center | 90 | 77 | 85.6% |
| Augusta University | 133 | 113 | 85.0% |
| University of Missouri, Columbia | 381 | 322 | 84.5% |
| University of Texas Health Science Center at Houston, The | 369 | 301 | 81.6% |
| University of Texas Southwestern Medical Center, The | 602 | 480 | 79.7% |
| Case Western Reserve University | 243 | 193 | 79.4% |
Percentiles across the 175 institutions reporting at least 50 postdocs: 10th 39.3% · 25th 51.0% · median 60.9% · 75th 66.7% · 90th 74.0%.
Every institution ranked →Distributions across the rest of the data →
How to read this page
The quotations are verbatim, taken from the published captions with their timestamps, so any one of them can be heard in context by pressing the time.
Six people spoke and each quotation names its speaker.
Every figure resolves from a federal dataset when the page is rendered. None is typed, and a figure that stops resolving fails the build rather than appearing as a dash. Each one links to the page where it can be checked at source.
Where the record stops, the page says so. Several of these claims cannot be settled from published data — not because they are wrong, but because the agency does not release the field. Those are listed as data requests, and they are the most useful thing this site can offer anyone working on this.
An investigation was announced. That is a fact with a date. Nothing here states that any institution or company broke a law.
Source: Vice President JD Vance Holds Press Conference on Fighting Visa Fraud, The White House, 8 October 2026.