Methodology: Illinois Veteran Employment & Business Data 2026

Story version: v2.0 (published 2026-04-28). Data retrieved April 2026.

What this story is, and what it is not

This story is a public snapshot of Illinois veterans in 2024 and 2026: where they live, who they are, which industries employ them, what those industries pay, and where the veteran population is headed through 2030. It is built from publicly available federal data sources. It is not a ranking of employers, not a measure of job quality, and not financial, tax, legal, or medical advice.

This story is educational data analysis. If you are making personal decisions about employment, benefits, or finances, consult an accredited Veterans Service Organization or a qualified professional. Authoritative starting points include VA.gov for benefits, the CFPB for consumer guidance, and the ABA Pro Bono Locator for legal help. If you or a veteran you know is in crisis, call the Veterans Crisis Line: dial 988 then press 1, or text 838255.

Acronym decoder

Where this data comes from

U.S. Census Bureau, American Community Survey (ACS)

The ACS is the largest ongoing household survey in the United States. The Census Bureau sends it to about 3.5 million addresses every year and publishes results in two forms: a 1-year estimate (released about a year after data collection, covering geographies of 65,000 people or more) and a 5-year pooled estimate (released every year, covering all geographies). This story uses the 2024 1-year estimate for the state and 22 Illinois counties with populations of 65,000 or more. It uses the 2020-2024 5-year estimate for all 102 counties, for service-era breakdowns, and for race and ethnicity detail. The 5-year pool is more stable in smaller counties but smooths over recent population change. Where both are available, this story shows the 1-year first and notes the difference. Official methodology: census.gov/programs-surveys/acs/methodology.html.

U.S. Bureau of Labor Statistics, Occupational Employment and Wage Statistics (OEWS)

OEWS is a twice-yearly survey of about 1.1 million establishments. It collects how many workers are employed in each occupation and how much they earn. Results are published by state and by 2-digit NAICS sector. This story uses the May 2024 release (published April 2025), the most recent vintage available in April 2026. OEWS does not collect veteran status, race, ethnicity, or gender at the occupation level, so the wages shown are all-worker averages. Official methodology: bls.gov/oes/current/methods_statement.pdf.

U.S. Department of Labor, VETS-4212

VETS-4212 is the annual filing federal contractors with $150,000 or more in federal contracts must submit to the Department of Labor. It reports, by establishment, how many protected veterans the company hired and employed during the reporting cycle. "Protected veteran" is narrower than the Census definition of veteran: it covers veterans who served during a named wartime, on a campaign, who are disabled, or who separated from service within the last 3 years. This story uses 2024 (complete) and 2025 (in-progress; filing deadline September 30, 2026) filings for Illinois establishments. Official methodology: dol.gov/agencies/vets/programs/vets4212.

Industry exhibit: definition, vintage, and group breakdowns

The "where Illinois veterans work" chart is computed from person-level ACS Public Use Microdata Sample (PUMS) records, 2020-2024 five-year file: civilian veterans (served on active duty in the past), currently employed. Two build-time checks lock this definition: the PUMS veteran total must match the published ACS veteran table within 1 percent, and the industry-code mapping must agree with the Census Bureau's official code dictionary.

A correction (2026-06-11): the earlier version of this chart reported about 257,000 employed veterans. That figure only reproduces if people whose service was training-only (Reserve or National Guard, never activated) are counted as veterans, which the ACS definition does not do. The corrected estimate, using the ACS definition, is 205,653 employed veterans (five-year average).

Gender and age breakdowns follow a suppression floor: an industry is shown for a group only when at least 30 survey respondents back the estimate. Slices below the floor are left out and the chart says how many. The county map's gender views suppress counties where the group's veteran estimate is under 50. The wage table and the survival chart carry no demographic breakdowns because their federal sources don't publish them.

U.S. Census Bureau, Business Dynamics Statistics (BDS): firm survival

The five-year survival figures on the Founder path are computed from the public BDS time series. For each industry: firms reaching age five, divided by firms founded five years earlier (age zero), averaged over the 2015 to 2017 founding cohorts. Firm counts are national, all founders, not veterans only. A firm that is acquired stops being counted even if the business continues, so these rates slightly understate business continuation. Two checks run on every build: the age categories must sum to the published total for every industry and year, and no founding cohort may grow as it ages.

Earlier drafts cited secondary summaries claiming roughly 70 to 80 percent five-year survival for health care and 40 to 50 percent for retail and food service. Those figures did not reproduce from the primary BDS firm data and were replaced with the computed rates shown on the page (2026-06-11).

U.S. Census Bureau, Annual Business Survey (ABS) and ACS class-of-worker: state comparison

The "Should you stay or go?" rows: veteran-owned share comes from the ABS 2022 company summary (businesses with employees; the share is veteran-owned firms divided by all employer firms, including firms whose ownership cannot be classified, such as publicly held companies). Self-employment comes from ACS table C24070, 2020-2024 five-year estimate: workers self-employed in their own incorporated or unincorporated business, divided by all workers. The unincorporated figure includes unpaid family workers, a very small group.

SBA certification and Mentor-Protege counts

SDVOSB counts come from the SBA's own certification search at search.certifications.sba.gov (the system of record since SBA took over veteran certification in January 2023). Anyone can reproduce a state's number there: filter to the SDVOSB certification plus the state. Counts were retrieved June 11, 2026; the database is live, so numbers drift as firms certify and lapse. Mentor-Protege counts come from the SBA's public list of active Mentor-Protege agreements (file effective May 5, 2026), counted by the protege firm's state. These are raw counts, not rates: larger states naturally have more firms. Note that VOSB and SDVOSB certifications overlap, so the two certification types are never added together on this page.

U.S. Department of Veterans Affairs, VetPop2023

VetPop2023 is the VA's official projection of how many living veterans will reside in each state from 2023 through 2050. It accounts for separations from active duty, mortality (by age cohort), and interstate migration. This story uses the Illinois series through 2030. The VA updates the model periodically as actuarial tables and separation rates change. Official methodology: va.gov/vetdata/veteran_population.asp.

Our methodology

Veteran definitions

Three definitions of "veteran" appear in this story. They are not interchangeable.

• Census ACS veteran (broadest): any civilian adult who served on active duty in the U.S. Armed Forces. Includes Reserve and National Guard members with active-duty service. Used in Sections 1, 2, 5, and 7.
• VA VetPop veteran: all living veterans, including those in VA facilities and nursing homes (not just private households). Used in Section 7.
• DOL VEVRAA protected veteran (narrowest): a veteran who meets one of four criteria (disabled, recently separated within 3 years, served during a war, or served on a campaign or expedition for which a badge was authorized). Used in Section 6 and the federal-contractor companion pages only.

Every chart and caption names the definition it uses.

Small-cell suppression

For counties where the Census ACS veteran count is fewer than 50 people, this story marks the county as "suppressed" in the downloadable data and does not publish a point estimate. This follows Census Bureau data-use rules and protects against identifying individual respondents. For VETS-4212 filings, counts are published as filed; VETS-4212 is an administrative filing by the employer and does not carry the same identification risk as survey microdata.

Geocoding

VETS-4212 filings include a county field. For the roughly one percent of filings where the county was blank, we backfilled via the U.S. Census Geocoder one-line-address endpoint (benchmark Public_AR_Current, vintage Current_Current). Final county coverage on placeable rows is 98.8 percent.

Entity resolution

Multi-establishment filings submitted by the same parent company are retained as filed, one record per establishment. We did not collapse multi-establishment filings into parent rollups. County-level totals reflect establishment-level filings, which is how the Department of Labor publishes them.

Cross-checks

Before publishing, the pipeline runs these checks and refuses to proceed if any fails:

• Every ACS veteran count reconciles against the source API response.
• All 102 Illinois counties appear in the county map.
• Every veteran-terminology caption carries the correct qualifier.
• All nine C21001 race and ethnicity tables are present for the state.
• No Illinois county with fewer than 50 veterans shows a published count.
• No duplicate ReportID appears in the VETS-4212 cache.
• All employer names round-trip cleanly through UTF-8.

ACS-to-NAICS industry crosswalk

Section 3 uses ACS industry codes from the PUMS (Public Use Microdata Sample). Section 4 uses NAICS 2-digit sectors from the BLS OEWS release. These two coding systems overlap but do not match exactly, which is why the story shows them as separate views rather than forcing a join.

Illinois population and veteran change in context

Illinois has lost residents most years since 2019. The Census Bureau's Population Estimates Program (PEP) shows net outmigration every year from 2020 through 2024. Some of those leaving, and some of those passing away each year, are veterans. That is why a 5-year pooled ACS estimate for 2020-2024 reports a veteran count that differs from the 2024 1-year estimate: the pool still includes veterans who were present in 2020 but are not present in 2024.

The VA's VetPop2023 model projects this decline forward. Illinois is projected to lose about 76,364 veterans between 2023 and 2030, driven mostly by mortality of Vietnam-era and older cohorts. This is the context behind the snapshot-vs-trend framing throughout the story: a single-year number can look stable while the underlying population is shrinking.

Success-evidence research: full citations

The Founder path's "What tends to predict success" section summarizes published research. Full sources, in page order. Evidence type is named per finding on the page itself; none of these findings is causal proof.

Five primary findings:

1. Industry survival rates: U.S. Census Bureau Business Formation Statistics and Business Dynamics Statistics (census.gov); SBA Office of Advocacy, "Startup Rates and Firm Age" (advocacy.sba.gov, 2016).
2. SBA-backed financing and survival: SBA lending data; quasi-experimental study in the Journal of Planning Education and Research (2021, journals.sagepub.com, DOI 10.1177/0739456X211028291); Walden University dissertation comparing SBA-loan recipients with matched non-recipients (scholarworks.waldenu.edu).
3. Founder experience status (first-time roughly 18 percent, prior failure roughly 20 percent, prior success roughly 30 percent over five years): Equidam meta-analysis using Bureau of Labor Statistics and EU statistical data (equidam.com).
4. Service versus product business architecture: U.S. Chamber of Commerce founder research summary (uschamber.com/co).
5. Mentoring outcomes for small firms: Entrenova small and medium enterprise mentoring meta-analysis (ojs.srce.hr/entrenova); SBA Mentor-Protege program documentation and public directory (sba.gov).

Nine additional findings (the "More research" section):

6. Founding-team composition: Florence Honore, Wisconsin School of Business (2025), using Census Longitudinal Employer-Household Dynamics data (business.wisc.edu).
7. Revenue concentration risk: Naval Postgraduate School thesis on government-contract management success factors (calhoun.nps.edu).
8. Employment model and time use: SBA Office of Advocacy, "Working for Yourself: How the Self-Employed Use Time Differently" (advocacy.sba.gov, 2025); Harvard Business School working paper 88-046 on employer commitment (hbs.edu).
9. Industry time-burden patterns: Bureau of Labor Statistics Occupational Employment and Wage Statistics sector data (bls.gov/oes). Inferred from operational structure; direct time-allocation studies by sector are rare.
10. Geographic ecosystem support: Kauffman Foundation, 2021 Early-Stage Entrepreneurship State Report (indicators.kauffman.org).
11. Prior government-contracting experience: FedBizAccess practitioner guide (fedbizaccess.com); Christoph LLC on past performance versus experience (christophllc.com); SAM.gov.
12. CMMC compliance: Washington Technology, "CMMC compliance gap now competitive risk" (washingtontechnology.com, January 2026).
13. Project-management certifications: practitioner guidance (Entrepreneur, Purdue Lean Six Sigma program documentation). Direct success-impact research is limited.
14. Structured entrepreneurship training: Small Business Institute Journal program-evaluation study (sbij.scholasticahq.com). The confidence effect for women participants is a Cohen's d of 0.75, a large standardized effect size; this is the figure behind the page's "notably stronger benefit for women" note. Long-term success-rate impact is not established.

Timing-window citations: every timing callout on the path pages carries its sources inline, and each claim has a verification date and review deadline tracked in the build. State tax claims are the most change-prone; verify against each state's revenue department before acting.

Reproducibility

The full pipeline is a Jupyter notebook that reads from a local cache of API responses. The cache can be regenerated from the live APIs by setting A1_REFRESH=1 before executing. Source code and exact dependency pins are available on request.

Reading-level scores

• Summary: Flesch-Kincaid grade 7.2.
• Body copy: Flesch-Kincaid grade 8.0.
• This methodology page: Flesch-Kincaid grade 8.5.
• Wage caveat: Flesch-Kincaid grade 7.2.

Charting library attribution

Interactive charts powered by Plotly.js (MIT licensed). See LICENSE for full attribution.

This product uses the U.S. Department of Labor API but is not endorsed or certified by the U.S. Department of Labor.

This product uses the Census Bureau Data API but is not endorsed or certified by the Census Bureau.

BLS.gov cannot vouch for the data or analyses derived from these data after the data have been retrieved from BLS.gov. Data retrieved April 23, 2026.

Veteran population projection from the VA VetPop2023 model. Department of Veterans Affairs, National Center for Veterans Analysis and Statistics.

© 2026 Rakvia (Henry Montoya). Story content licensed under CC BY 4.0.