Abbott Froze H-1B Filings at Texas Universities. They Kept Filing.

What The Order Actually Says

Abbott's directive, announced January 27, 2026, bars Texas state agencies and public institutions of higher education from filing new H-1B petitions — or renewing/initiating them without written permission from the Texas Workforce Commission — through the end of the next legislative session, May 31, 2027. It does not touch existing H-1B employees already on staff, and it explicitly does not apply to private-sector employers. Institutions were also ordered to report, by March 2026, how many H-1B workers they currently sponsor, in what job classifications, from what countries, with what visa expiration dates, and documentation that qualified Texans were given a "reasonable opportunity" to apply for each position first. Abbott framed it around a September 2025 Trump administration proclamation and allegations that some employers had used H-1B hires to displace American workers at lower wages.

The order does not create a new law — it's an executive directive to entities within the governor's own chain of authority, enforced (in theory) through budgetary and administrative leverage rather than a court order or statute.

Can A Governor Actually Do This?

H-1B is a federal program — the petition itself is adjudicated by USCIS, and the labor condition application underneath it is certified by the U.S. Department of Labor. No governor can override that process or make it federally illegal for a Texas institution to sponsor someone. What Abbott is actually claiming authority over is narrower: whether Texas's own public institutions, as instruments of state government, can be directed not to initiate that federal process in the first place. That's a real distinction, and it's the one the order leans on — citing the state's control over budgets and administrative practices, 2025's Senate Bill 37 (which expanded state oversight of public university governance), and the governor's appointment power over agency leadership.

But that authority is genuinely contested, not settled. Texas public universities are run by governing boards with statutory independence over hiring and employment decisions — the governor doesn't directly control a university provost's hiring pipeline the way he controls a state agency head he appointed. Legal analysis published shortly after the order flagged real ambiguity: compliance depends on indirect levers (funding, institutional governance) rather than a direct legal command, the order's reach into "semi-autonomous or affiliated entities" is unclear, and employers now face two, potentially conflicting sets of obligations — federal immigration law says they can sponsor; a state executive order says (for now) they shouldn't. Nobody has definitively answered whether a university board of regents is actually bound by it, or what happens if one decides it isn't.

Data Note: DOL's Own Files Triple-Count Filings, and We Almost Missed It

The first version of this report, published earlier today, put the total at 350 new-employment filings since the freeze, 319 certified, across 28 institutions. Those numbers were wrong, and the reason is worth explaining rather than quietly correcting: DOL's FY2026 quarterly LCA disclosure files are cumulative, not incremental. The Q3 file (received dates through June 30, 2026) doesn't just contain April–June cases — it republishes nearly every Q1 and Q2 case alongside them, refreshed with each case's current status. Our local combined table stored Q1, Q2, and Q3 as three separate loads; summing across all three, as the original version of this report did, counted the same DOL case number two or three times for any month that had already shipped in an earlier quarterly file.

We confirmed the fix two ways. First, internally: of 731,003 raw rows across all three quarters, only 437,497 distinct CASE_NUMBER values exist, and the Q3 file alone accounts for 437,496 of them — a near-total superset of Q1 and Q2, exactly as DOL's cumulative-file design would predict. Second, externally, against DOL's own published aggregate: the Office of Foreign Labor Certification's "Selected Statistics" report for FY2026 Q3 lists 437,496 total applications processed nationally through June 30, 2026 — matching our deduplicated count almost exactly. Every figure in this report now uses only the Q3 snapshot (the most current status for each case), which is both the correct and the simpler approach. We're leaving this note up rather than pretending the first version didn't happen, because if we can make this mistake with data we mirror locally, it's worth asking how many other analyses — ours or anyone else's — are quietly built on the same unflagged assumption about how DOL packages its releases.

What The Filing Data Shows

Method: identified employer-name variants across our FY2026 H-1B disclosure data (Q3 snapshot, 437,496 deduplicated filings, received dates October 2025 through June 2026) matching the University of Texas, Texas A&M, Texas Tech, University of Houston, University of North Texas, and Texas State University systems, plus the state's independent regional public universities — 30 of which have at least one new-employment filing in this window. Private institutions in the same data (Rice, SMU, Baylor, TCU) were excluded; they aren't covered by the order and aren't in this count. Each LCA record carries a NEW_EMPLOYMENT flag from DOL itself, distinguishing a genuinely new hire from a continuing employee's renewal, transfer, or amendment — exactly the line the order draws, so that's the figure charted below rather than raw filing volume.

New-employment H-1B labor condition applications received by DOL, by month, for Texas public university and university-system employers, deduplicated by DOL case number using each case's current (Q3) status. Nov–Dec 2025 reflect normal pre-cap-season filing buildup ahead of April 1 (when new cap-subject H-1B petitions open) — a pattern visible across the entire dataset, not specific to these institutions. Source: DOL LCA Disclosure Data FY2026 Q3 snapshot.

The Dip Was Real. It Also Didn't Last.

February 2026's drop isn't noise. Across the entire H-1B dataset — every employer, not just Texas public universities — new-employment filings actually rose from January to February 2026 (5,517 to 6,156, +11.6%), the normal run-up to April's cap season. Texas public universities moved the opposite direction over that same month, falling 86% (58 to 8). That divergence from the broader market is the clearest evidence in this data that institutions genuinely reacted to the order in real time.

But March reversed it, and by April, new-employment filings at Texas public universities (113) were nearly double where they'd been in January, the last full pre-freeze month (58) — even as the order remained (and remains, as of this writing) formally in effect through May 2027. May (88) and June (78) stayed elevated well above the sub-10/month pace these institutions ran in the immediate post-freeze trough. Nothing about DOL's own processing changed: of the 323 new-employment LCAs these institutions have filed since the freeze took effect, 295 have already been certified and zero denied — the federal side of the pipeline is functioning exactly as it always has. Whatever stopped in February, it wasn't the federal government's willingness to process these applications.

By Institution, Since The Freeze

New-employment H-1B filings received by DOL from Texas public universities and university-system agencies, February 1 – June 30, 2026 (top 15 by volume; 15 smaller institutions not shown, ranging from 1–4 filings each).

Sample Certified Positions, Filed After The Freeze

A random sample of individual, DOL-certified new-employment filings received March 2026 or later — specific job openings Texas public universities sponsored for H-1B workers after the freeze took effect, and DOL approved.

Investigative Assessment

💬 Comments