The Congressional Research Service documented the H-1B and employment-based immigration system for Congress. Here is what it confirms, what it glosses over, and what only ground-level data reveals.
The Congressional Research Service is Congress's nonpartisan think tank — staffed by lawyers, economists, and policy analysts who produce research exclusively for Members of Congress and their committees. Their reports are not advocacy. They are meant to be the most objective, authoritative summary of a policy area available to lawmakers.
Report R47164, U.S. Employment-Based Immigration Policy, is CRS's comprehensive treatment of how the United States admits, tracks, and manages foreign workers — covering green card preference categories, the H-1B specialty worker visa, L-1 intracompany transferees, Optional Practical Training (OPT) for foreign students, and the decades-long policy debate around all of it. It was first published in 2022 and most recently updated in November 2024.
It is, in short, what Congress officially knows about this system. That makes the gaps in it just as informative as the content.
"Increasingly, H-1B workers are hired by a small group of very large firms in the scientific and business services, computer, and information technology staffing industries with expertise in navigating the H-1B application process. These large employers are not typical of U.S. employers; they are a selected group."
That sentence is the entire universe this site's H-1B database and scorecard are built around. CRS treats it as a methodological footnote — a "selection problem" that complicates academic research. We treat it as the policy problem itself.
On several points, R47164 is remarkably direct. These are not contested findings — they are documented in the official record that Congress has in hand.
The report documents that the employment-based green card queue stood at 870,000 people as of September 2021, and that workers from India face waits of decades — in some EB categories, over 100 years at current processing rates. Workers stuck in this queue are sponsored by specific employers. If they leave that employer, they lose their place in line. CRS describes this situation factually. What it does not explore is the structural consequence: an employer who holds your green card sponsorship holds enormous leverage over your wages, your ability to complain about working conditions, and your willingness to organize.
The report documents that Optional Practical Training enrollment grew from under 25,000 in CY2007 to over 214,000 in CY2023. OPT has no annual cap, no prevailing wage requirement, and no Labor Condition Application process. Employers who hire OPT students pay no H-1B fees and face none of the attestation requirements that theoretically protect American workers.
OPT enrollment, CY2007–CY2023. Source: ICE SEVIS via CRS R47164. No annual cap. No prevailing wage requirement.
CRS explicitly notes that H-1B usage is increasingly concentrated among a small number of IT staffing and outsourcing firms. The report describes this as a complication for labor economists trying to measure H-1B impacts, because these firms are not "typical" employers. What it does not discuss is why this concentration matters for American workers, wage-setting, or the fundamental purpose of the prevailing wage requirement — which is largely meaningless if the dominant employers in a market are all submitting H-1B applications at similar wage floors.
A nonpartisan briefing document for Congress presents all sides fairly. That is its job. But "presenting all sides" can obscure what ground-level data makes obvious. Here are the places where R47164 is technically accurate and practically incomplete.
CRS notes that H-1B employers must pay the greater of the actual wage or the prevailing wage. What it does not examine is how prevailing wages are calculated — using the Department of Labor's four-tier wage level system, where the majority of H-1B petitions are filed at Level 1 (the bottom 17th percentile of wages for that occupation). A system that defines "prevailing wage" as the wage most workers in a field earn more than is not a wage protection — it is a floor set below the market. Our H-1B database lets you check any employer's filed wages against prevailing wages directly.
The 870,000-person EB queue is framed in R47164 primarily as a humanitarian issue — workers from India and China waiting unreasonably long for green cards they were promised. That framing is real. But the queue also functions as a retention mechanism that benefits employers, not workers. A software engineer five years into a ten-year green card wait does not negotiate for a raise or quit for a competitor. CRS documents the queue's existence and scale. It does not examine this dynamic at all.
CRS treats OPT as a student benefit and an immigration pathway. It does not examine OPT as a labor market instrument — a mechanism that allows employers to hire highly educated workers with no wage floor, no cap, no attestation requirement, and a tax advantage (OPT workers on STEM extension are exempt from FICA taxes). The 214,000 OPT workers in 2023 represent a workforce larger than many entire industries, operating almost entirely outside the labor protections that are supposed to govern the H-1B program they typically feed into.
Throughout R47164, the report references "labor market needs" and "tight labor markets" as the rationale for employment-based immigration. It does not examine the evidence for those claims in the specific occupations — particularly software development and IT — where H-1B usage is most concentrated. Our work visas vs. new jobs report shows that in nearly every year from 2008 to 2024, total work-authorized visa issuances ran at or above the total number of net new jobs created in the entire U.S. economy. That is not a picture of a labor market so tight it requires unlimited foreign worker pipelines.
CRS works from aggregate statistics and published research. This site works from the transactional records — the actual LCA filings, the actual contract awards, the actual employer names and dollar amounts. Here is what that ground-level view adds.
In June 2026, we pulled all USASpending.gov awards matching the keyword "staffing" across all federal agencies for FY2026. The result was $4.1 billion in contracts — and the contractor list looks strikingly familiar to anyone who has spent time in the H-1B data. Tribal and Native-owned IT staffing firms that dominate H-1B filings at the State Department are the same firms winning government staffing contracts there. The H-1B ecosystem and the federal contracting ecosystem are not separate worlds.
In our H-1B filings database, covering FY2026 Q2, the wage gap between offered wages and prevailing wages is measurable for tens of thousands of individual petitions. You can look up any employer and see the spread. CRS notes the wage attestation requirement exists. Our data lets you see how it performs in practice — employer by employer, job title by job title.
This is not a criticism of CRS. Their job is nonpartisan synthesis for policymakers. Our job is transparent data for the public. They are different tools.
| Topic | CRS R47164 | GuestWorkerVisas.com |
|---|---|---|
| H-1B wage attestation | ◑ Describes requirement | Searchable by employer — actual offered vs. prevailing wages for every LCA filing |
| Staffing firm concentration | ◑ Acknowledges it | Quantified — top employers, petition volumes, wage patterns |
| OPT growth | ◑ Documents the numbers | Charted with sourced data; FICA exemption and wage floor absence noted |
| EB queue as labor leverage | ✗ Not examined | Documented: workers cannot leave sponsoring employer without losing queue position |
| Visas vs. job creation | ✗ Not in scope | Year-by-year comparison, FY2008–FY2024, all work-authorized categories |
| Federal contracting overlap | ✗ Not in scope | $4.1B FY2026 contracts — same employers, same ecosystem, searchable |
| Foreign-born vs. native-born employment | ✗ Not in scope | 19 years of BLS Table A-7 data, including the COVID baseline problem |
| Individual employer lookup | ✗ Not possible | Search any employer's full LCA history, wages, and job titles |
Because CRS is nonpartisan and Congress-facing, citing R47164 in your own writing carries weight that citing an advocacy organization does not. Several of its findings are directly useful when making the case that this system warrants serious scrutiny:
"Increasingly, H-1B workers are hired by a small group of very large firms in the scientific and business services, computer, and information technology staffing industries." (CRS R47164, updated Nov. 2024) — This is Congress's own think tank confirming that the program is dominated by intermediaries, not the "best and brightest" employer-by-employer narrative that is usually offered in its defense.
The queue of approved-but-waiting EB immigrants totaled 870,000 as of September 2021, with Indian nationals in some categories facing waits of over 100 years. (CRS R47164) — Useful context for discussing why workers cannot simply "vote with their feet" if an employer underpays or mistreats them.
F-1 students working under OPT grew from fewer than 25,000 in CY2007 to over 214,000 in CY2023. (CRS R47164) — The growth of a no-cap, no-wage-floor pipeline that feeds directly into the H-1B system is documented in the official record.
For all three of these, you can then link directly to this site's tools to show what those documented trends look like at the employer level — making the connection between the policy abstraction and the actual filings.
This research is produced independently — one person, 23 years of data work, living on $1,419/month in Social Security. If this analysis has been useful to you, a small donation genuinely makes a difference.
♥ Donate via Square