After I was laid off in 2024, I applied for a lot of jobs.
I applied broadly at UMass Amherst and the Five Colleges, including two positions at Hampshire College. I also applied at CVS, Stop & Shop, Big Y and other local employers.
The response was remarkably thin.
Out of all those applications, I received one callback and interview, for a janitorial position at UMass.
That left me with a question.
What was happening to all the other jobs?
I wasn't particularly interested in guessing. I wanted to see what I could actually find.
Looking at H-1B records
One place I started looking was the federal employment data associated with H-1B workers.
Before an employer can sponsor someone for an H-1B visa, it generally files a Labor Condition Application, or LCA, with the U.S. Department of Labor. Those filings create a public record containing information such as the employer, occupation, location and offered wage.
Several websites make that data searchable, including H1BData, MyVisaJobs and LayoffHedge.
So I started searching employers where I had been applying.
UMass Amherst produced a lot of records.
One database search showed 423 UMass Amherst LCA records covering 211 job titles. Another source reported 52 certified LCAs for UMass in fiscal year 2024 alone.
That was enough to establish something fairly easily:
UMass Amherst was an active participant in the H-1B system.
But that wasn't actually the question I was trying to answer.
Did one of those workers get the job I applied for?
That's where things became much more difficult.
An LCA isn't a hiring record.
It shows that an employer took a step that could support an H-1B petition. It doesn't necessarily establish that the petition was ultimately filed, approved or used. And it certainly doesn't tell me that a particular person hired under the program received a particular job that I had applied for.
I could compare dates.
I could compare job titles.
I could compare salaries and locations.
I could look at the university's job advertisements and compare them with the government data.
But similarity isn't proof.
A job title appearing in an LCA database at roughly the same time that I applied for something with a similar title might be worth noticing. It isn't enough to say they were the same position.
That distinction became increasingly important as I continued looking.
Finding evidence isn't the same as proving the hypothesis
This is one of the things I like about research.
I started with a suspicion that could be tested, at least partly, against public records.
