GuestWorkerVisas.com
State Dept. Table XV(B) · BLS CES · ICE SEVIS · FY2008–FY2024

The Math Nobody Is Doing:
Work Visas vs. New Jobs

Every year, the U.S. government issues hundreds of thousands of work-authorized visas. Every year, the economy creates (or loses) a certain number of net new jobs. Almost nobody puts those two numbers side by side. This report does.

Sources: State Dept. / BLS / ICE SEVIS   Period: FY 2008 – FY 2024   Published: April 2026   GuestWorkerVisas.com
17-year totals · FY2008–FY2024
Core Work Visas Issued
H, L1, O, P, TN, E, R1, Q1, I, CW1 (uncontested categories)
Spousal Work Auth Added
H4 + L2 + E3D (policy-change dependent)
J-1 Exchange Visitors
Many work-authorized; subcategory data varies
Net New Jobs (CES)
Dec-to-Dec net change; includes 2020 COVID crash
Why this matters

New Jobs vs. New Work Authorizations

The question is simple: when the economy creates a net new job, how many newly-authorized foreign workers are also entering the labor market seeking work? The answer, year after year, is: a very large number — often exceeding the total of net new jobs created.

This does not prove displacement. The relationship between visa issuances and employment outcomes is complex. Seasonal agricultural workers (H-2A), artists on short-term engagements (P-3), and corporate transferees (L-1) fill very different roles than domestic job seekers. But the raw math has almost never been presented to the public — and it should be.

Note: This report uses State Department fiscal year (October–September) for visa data and calendar year (January–December) for BLS job data. This creates a roughly three-month offset in the comparison.

Year-by-year comparison
Net New Jobs Created (bars) vs. Total Work Visas Issued (line) — both in thousands, FY/CY 2008–2024
Net New Jobs — positive (bars, amber) Net New Jobs — negative / job losses (bars, red) Total Work Visas Issued (line, blue)

Bars = net nonfarm payroll change (BLS CES, Dec-to-Dec, SA). Amber = job gains; red = net losses. Blue line = total work-authorized visas (Core + Spousal H4/L2/E3D + J-1), State Dept. fiscal year. Both series in thousands. The visa line runs flat and stable compared to the dramatic swings in job creation — the story of the chart. OPT excluded (different source; see table below).

"Even in 2008 and 2009 — when the economy destroyed a combined 8.3 million jobs — over 1.8 million work-authorized visas were still issued."

Complete data table · all figures in thousands
How to read this table

Core Work Visas = H-1B/B1, H-2A, H-2B, H-3, L-1, O-1/2, P-1/2/3, TN, E-1/2/3/3R, CW-1, R-1, Q-1, I (media). These holders are unambiguously work-authorized. Spousal = H-4 + L-2 + E-3D — work authorization for these categories changed significantly in 2015 and 2022; see footnotes. J-1 covers dozens of subcategories with varying work eligibility. OPT (not shown here) adds substantially more — see separate note below.

Year — Core Work-Authorized Visa Groups — — Spousal — J-1 — Summary —
FY / CY H-1B +H-1B1 H-2A Ag H-2B Other All Other Core¹ H-4² L-2³ E-3D J-1⁴ Core Total All Incl. Spousal+J1 Net New Jobs (CES) Visas as % of Jobs⁵

¹ "All Other Core" = L-1, O-1, O-2, P-1, P-2, P-3, TN, E-1, E-2, E-3, E-3R, CW-1, R-1, Q-1, and I (media). Full breakdowns in methodology section.

² H-4 spouses of H-1B holders were NOT work-authorized until USCIS rule effective May 26, 2015. Pre-2015 figures shown for context only; those holders could not work legally.

³ L-2 spousal work authorization was automatic only after a federal court settlement formalized in January 2022. Pre-2022, holders required a separate EAD.

J-1 covers professors, researchers, au pairs, camp counselors, trainees, interns, and government visitors. Work eligibility varies by subcategory. Many categories include unpaid or stipend-only participants. Treat as a partial work-authorization figure.

"Visas as % of Jobs" = All (Core + Spousal + J-1) ÷ Net New Jobs × 100. Shown as N/A in years with net job losses (denominator is negative). Does not account for renewals, multi-year holders, or OPT.

J-1 Exchange Visitor Categories
The J-1 requires special treatment

The J-1 visa covers a remarkably broad range of participants — from Ivy League professors to summer camp counselors to au pairs. Some categories are fully work-authorized with wages; others are educational exchanges with stipends only. The State Department does not break out subcategories in Table XV(B). ICE SEVIS data provides subcategory detail separately.

The key fact: J-1 issuances have run between 108,000 (2020 COVID year) and 391,000 (2019) annually — making it one of the largest nonimmigrant work-related categories, and one of the least-discussed in policy debates.

Optional Practical Training (OPT) — Not Included Above
OPT is a major omission from the table above — and it is intentional. OPT is tracked by ICE SEVIS, not the State Department, and does not involve a visa issuance (F-1 holders apply for an Employment Authorization Document while already in the U.S.). Including it would require a different data source and methodology. However, its scale is too large to ignore.
Category CY2016 CY2019 CY2021 CY2022 CY2023 CY2024
STEM OPT (2-yr extension) 41,782 72,116 n/a n/a 122,101 est. 135,000+
Non-STEM OPT (12-month) n/a n/a 115,651 171,635 n/a n/a
Total OPT Authorized (stock) n/a n/a n/a n/a n/a 418,781
Total Practical Training (OPT + STEM OPT + CPT) est. 196,000 est. 230,000 n/a n/a 539,382 n/a

Sources: ICE SEVIS By The Numbers reports; CRS In Focus IF12631 (2025); Center for Immigration Studies analysis (2024). "Stock" counts individuals authorized at any point during the calendar year; some are counted in multiple years if their authorization spans years. "Practical Training" totals include Curricular Practical Training (CPT), which covers students working while still enrolled.

Methodology & Caveats

What This Analysis Shows — and Doesn't Show

What it shows: The volume of new work authorizations issued annually relative to net new job creation. The math is straightforward: if 2 million new jobs are created and 1.3 million work visas are issued, then at minimum those visa holders account for a meaningful fraction of available new employment opportunities — regardless of whether they fill new positions or replace departing workers.

What it doesn't show: Displacement. One visa ≠ one job taken from an American. Many visa holders fill roles with no qualified domestic applicants (H-2A agriculture). Many are seasonal or short-term (H-2B, P-3). Many are in transfers within multinational firms (L-1). And many new jobs are created precisely because of foreign talent. This analysis does not make a causal claim — it presents the scale question that must precede any honest policy conversation.

Data limitations: State Department data counts visa issuances, not actual U.S. labor market entries. A holder may enter months later, re-enter on the same visa, or not enter at all. Renewals and extensions of existing status (change of status, USCIS approvals) are not counted here. The USCIS I-129 petition dataset (approvals by employer) would show a different but related picture. Finally, 2019 visa methodology changed (from workload counts to application-level tracking); pre-2019 data is not fully comparable to 2020 onward.

Primary Sources

Visa data: U.S. Department of State, Report of the Visa Office — Table XVI(B)/XV(B): Nonimmigrant Visas Issued by Classification (Including Border Crossing Cards). Fiscal years 2007–2011 (FY11 Annual Report), 2011–2015 (FY15 Annual Report), 2016–2020 (FY20 Annual Report), 2020–2024 (FY24 Table XV(B)). travel.state.gov

Employment data: U.S. Bureau of Labor Statistics, Current Employment Statistics (CES) — All Employees, Total Nonfarm (PAYEMS). Seasonally adjusted, December-to-December annual change. fred.stlouisfed.org/series/PAYEMS

OPT data: ICE SEVIS By The Numbers (annual reports); CRS In Focus IF12631, "Optional Practical Training (OPT) for Foreign Students" (2025); Center for Immigration Studies analysis (2024). ice.gov, congress.gov

Note on 2024 CES data: The BLS February 2026 benchmark revision reduced March 2025 nonfarm employment by 898,000 (−0.6%). The 2024 annual figure used here reflects post-revision estimates and should be treated as preliminary.

Help Support This Work

This research is produced independently — one person, 23 years of data work, living on $1,419/month in Social Security. If these reports have been useful to you, a small donation genuinely makes a difference.

♥  Donate via Square