Two federal agencies each publish their own number every year — the State Department counts how many work-authorized visas it issues, and the Bureau of Labor Statistics counts how many net new jobs the economy actually created. Nobody routinely puts them side by side. Doing that for every year since 1987 shows a ratio that held remarkably steady for most of the last four decades — until 2025, when it broke the pattern entirely: the U.S. issued 859,476 worker visas against just 584,000 net new jobs, a ratio 6.6x the pre-COVID normal and higher than at any point during the 2008 financial-crisis recovery.
Worker visas = annual consular issuances of the visa classes that authorize employment and compete directly with domestic labor: H-1B, H-1B1, H-2A, H-2B, H-3, L-1, O-1, O-2, P-1, P-2, P-3, E-1, E-3, E-3R, TN, CW-1, Q-1, and R-1. Source is the State Department's annual "Nonimmigrant Visas Issued by Classification" table for FY1987–FY2024; FY2025 (which State has not yet published as of this report) is reconstructed by summing all twelve of State's monthly "NIV Issuances by Nationality and Visa Class" PDFs for Oct 2024–Sep 2025, cross-checked against each PDF's own printed grand total. Excluded: dependent visas (H-4, L-2, etc.), tourist/business visitor visas (B-1), exchange visitors (J-1), and all immigrant (green card) categories.
Net jobs created = BLS total nonfarm payroll employment, December-to-December net change, seasonally adjusted, calendar year basis — the same headline number reported each January as "the economy added X million jobs this year."
Two honest limitations, stated up front: First, visa data runs on the federal fiscal year (Oct–Sep) while jobs data runs on the calendar year — a one-quarter offset, not a same-window comparison. Second, consular visa issuances are not a clean count of new workforce entrants: a worker who already holds a job in the U.S. and travels home mid-career gets issued a new visa stamp again on return, so this number somewhat overstates new arrivals. In the other direction, it understates total new work authorization, because H-1B extensions and changes of status approved domestically by USCIS — without the worker ever visiting a consulate — don't appear in State's consular-issuance count at all. Neither of these is a reason to dismiss the ratio; they're reasons to treat it as a strong directional signal rather than a precise headcount, which is exactly how it's used below.
For most of the last four decades this ratio moved in a narrow, boring band — a small fraction of a visa per job, because the economy was creating far more jobs than it was issuing worker visas. Red bars mark years the economy shed jobs outright (ratio undefined — there's no "per job" to divide by). 2025 is the tallest bar on the chart that isn't a recession year, and it's taller than 2010, the worst point of the post-2008 jobless recovery.
| Fiscal Year | Worker Visas Issued | Net Jobs Created (BLS) | Visas per Job | Note |
|---|---|---|---|---|
| 2025 | 859,476 | 584,000 | 1.47 | Record ratio |
| 2024 | 856,314 | 2,012,000 | 0.43 | |
| 2023 | 908,520 | 2,594,000 | 0.35 | Highest visa year on record |
| 2022 | 813,641 | 4,555,000 | 0.18 | |
| 2021 | 496,691 | 7,233,000 | 0.07 | Reopening surge |
| 2020 | 491,272 | -9,247,000 | — | COVID crash — jobs shed |
| 2019 | 679,054 | 1,986,000 | 0.34 | |
| 2018 | 639,223 | 2,286,000 | 0.28 | |
| 2017 | 609,114 | 2,115,000 | 0.29 | |
| 2016 | 584,302 | 2,331,000 | 0.25 | |
| 2015 | 521,220 | 2,980,000 | 0.17 | |
| 2014 | 477,080 | 2,729,000 | 0.17 | |
| 2013 | 435,052 | 2,573,000 | 0.17 | |
| 2012 | 390,668 | 2,492,000 | 0.16 | |
| 2011 | 374,776 | 2,257,000 | 0.17 | |
| 2010 | 360,862 | 323,000 | 1.12 | Post-crisis recovery low |
| 2009 | 347,516 | -5,177,000 | — | Financial crisis — jobs shed |
| 2008 | 455,492 | -3,121,000 | — | Financial crisis — jobs shed |
| 2007 | 425,987 | 1,241,000 | 0.34 | |
| 2006 | 437,129 | 1,966,000 | 0.22 | |
| 2005 | 377,710 | 2,474,000 | 0.15 | |
| 2004 | 371,309 | 2,145,000 | 0.17 | |
| 2003 | 335,177 | -83,000 | — | Recession — jobs shed |
| 2002 | 330,961 | -323,000 | — | Recession — jobs shed |
| 2001 | 373,991 | -180,000 | — | Recession — jobs shed |
| 2000 | 327,777 | 1,772,000 | 0.18 | |
| 1999 | 275,910 | 3,079,000 | 0.09 | |
| 1998 | 226,978 | 3,154,000 | 0.07 | |
| 1997 | 198,774 | 3,068,000 | 0.06 | |
| 1996 | 161,912 | 2,517,000 | 0.06 | |
| 1995 | 153,619 | 3,019,000 | 0.05 | |
| 1994 | 131,504 | 3,352,000 | 0.04 | |
| 1993 | 114,902 | 2,274,000 | 0.05 | |
| 1992 | 110,339 | 288,000 | 0.38 | Jobless recovery |
| 1991 | 109,342 | -1,145,000 | — | Recession — jobs shed |
| 1990 | 100,831 | 1,508,000 | 0.07 | |
| 1989 | 89,991 | 2,550,000 | 0.04 | |
| 1988 | 71,949 | 3,287,000 | 0.02 | |
| 1987 | 36,446 | 2,584,000 | 0.01 |
This is not a claim that guest-worker visas are the sole or even the primary reason any individual American can't find a job — plenty of other forces move the jobs number too, including automation, labor-force participation trends, offshoring, and interest rates. Nobody should walk away from a single ratio thinking it explains everything.
What it does show is narrower and, we think, undeniable: guest-worker visa issuance is supposed to respond to labor shortages — that's the legal justification behind every one of these visa categories. If that were true in practice, visa issuance should track job creation, rising and falling with it. Instead the ratio between the two has been essentially flat for a decade and a half — until a year when job creation nearly stalled and visa issuance stayed near an all-time high anyway. Nobody in the public debate about "why can't Americans find jobs" is putting these two government numbers next to each other. This report is that comparison.
This investigation is produced independently — one person, 23 years of data work, living on $1,419/month Social Security. If this work matters to you, a small donation makes a real difference.
♥ Donate via Square More Reports