Methodology.

Data sources

The full list of sources behind every employer profile. Source-specific fields use a consistent prefix (osha_*, whd_*, cms_*) so you can identify provenance from the field name alone.

• OSHA enforcement — inspections, violations, fatalities, and penalties from the Occupational Safety and Health Administration. Includes violation severity (willful, repeat, serious, other-than-serious) and SVEP flags.
• OSHA ITA — Form 300A self-reported DART and TRIR rates by establishment.
• WHD — Wage and Hour Division compliance actions. Case counts, back wages owed, employees affected.
• MSHA — mine safety violations, inspections, and assessed penalties.
• EPA ECHO — Clean Air, Clean Water, RCRA, and SDWA inspections, formal actions, and penalties.
• NLRB — unfair-labor-practice charges and union-representation case history.
• FMCSA — DOT-registered motor carriers with operation type and safety basics.
• OFLC — H-1B, H-2A, and H-2B labor condition applications and prevailing-wage ratios.
• BLS SOII — Bureau of Labor Statistics industry-wide TRIR/DART averages for peer-percentile ranking.
• SAM.gov — federal debarment, suspension, and ineligibility flags.
• CMS — Centers for Medicare & Medicaid five-star ratings, deficiencies, civil money penalties, and staffing hours per resident (nursing-home operators only).
• USAspending — federal prime-contract awards with a 25-year lookback.
• CPSC — Consumer Product Safety Commission recalls tied to the responsible manufacturer.
• NHTSA — vehicle recalls and defect complaints tied to OEMs and upfitters.
• UVA Corporate Prosecution Registry — academic source maintained by UVA Law, cataloging federal corporate prosecutions, plea agreements, DPAs, and NPAs.

OFCCP federal contractor compliance data is planned for a future update — ingestion is under development.

Entity matching

Employer profiles aggregate enforcement records across multiple federal agencies using deterministic matching on normalized employer name, state, and ZIP code. Records from different agencies are linked to the same profile when these three fields match.

Parent-company matching uses a curated seed table of 138 patterns (e.g. WALMART mapping to Walmart Inc.) plus ITA Form 300A company-name data to group subsidiary locations under their parent entity.

Matching errors may occur — particularly for employers with common names operating in the same state and ZIP code. If you believe records have been incorrectly attributed, please email corrections@fastdol.com.

Scoring formula

The raw score sums weighted components, each capped to prevent any single factor from dominating:

Time decay

The raw score is multiplied by a time-decay factor based on the most recent enforcement activity (OSHA inspection or WHD case). Recent activity counts at full weight; older activity is progressively discounted:

The final score is clamped to the 0 – 100 range: min(100, max(0, raw_score × decay)).

Risk tiers

The numeric score is mapped to a risk tier. Willful and repeat violations trigger automatic tier overrides regardless of the numeric score:

Employers with any recorded violations in the last 3 years cannot be LOW — recent enforcement activity is itself a risk signal even when the numeric score is modest.

Trend signal

The trend compares enforcement activity in the last 12 months against the 36-month average. It requires at least 3 violations in the 3-year window to produce a non-STABLE signal:

• WORSENING — last 12 months > 150% of the 3-year annual average
• IMPROVING — last 12 months < 50% of the 3-year annual average
• STABLE — in between, or insufficient data

Confidence tier

Reflects how much federal enforcement data exists for this employer:

• HIGH — 3+ OSHA inspections on record
• MEDIUM — 1 – 2 OSHA inspections or 1+ WHD cases
• LOW — no OSHA inspections and no WHD cases (employer exists in the system but has minimal enforcement history)

Peer percentile

Each employer’s violation count is ranked against peers in the same 4-digit NAICS industry and state. The percentile (0 – 100) indicates what percentage of peers have fewer violations. A peer group of at least 10 employers is required for the percentile to be displayed.

Industry benchmarks

DART (Days Away, Restricted, or Transferred) and TRIR (Total Recordable Incident Rate) benchmarks come from the Bureau of Labor Statistics Survey of Occupational Injuries and Illnesses (SOII), Table 1. Benchmarks are matched at the most specific NAICS level available (6-digit → 5 → 4 → 3 → 2 digit). Self-reported employer rates come from OSHA ITA Form 300A filings.

Limitations

• Risk scores reflect federal enforcement activity only. State-level OSHA programs (in the 22 states that operate their own programs) are included in the DOL data but may have different reporting lag.
• Employers with no inspections receive a score of 0. A score of 0 does not mean an employer is safe — it means there is no federal enforcement data on file.
• OSHA citations may be contested, vacated, or reduced after issuance. The data reflects the most recent status available from the DOL API, but there may be a lag.
• Self-reported injury rates (DART, TRIR) are submitted by the employer and may be subject to underreporting.
• The scoring formula does not currently incorporate MSHA, EPA, FMCSA, NLRB, or OFLC enforcement data in the numeric score. These sources are displayed on employer profiles and contribute to qualitative flags.
• Scores are not normalized for employer size. Larger employers will naturally accumulate more inspections and violations. Employee count data (from OSHA ITA Form 300A) is available for approximately 55,000 employers and is displayed when present, but is not used in the score calculation.
• Scores are not normalized for industry. A construction company and a software company with identical violation counts will receive the same score. The peer violation percentile (shown separately on each profile) provides industry-relative context.

References & sources

Every number on a FastDOL employer profile comes from one of the primary sources below. External links go to the official government API or the canonical data-distribution page.

• OSHA Enforcement Data — Establishment Search (Occupational Safety and Health Administration)
• OSHA Injury Tracking Application (ITA) — Form 300A
• U.S. Department of Labor Enforcement Data Catalog (WHD, OFCCP, OLMS)
• MSHA Open Data (Mine Safety and Health Administration)
• EPA ECHO — Environmental Compliance History
• NLRB Case-Activity Data (National Labor Relations Board)
• FMCSA Safety Measurement System Data Downloads
• OFLC Performance Data (Office of Foreign Labor Certification)
• BLS Survey of Occupational Injuries and Illnesses (SOII) — Table 1
• SAM.gov Exclusions Extracts (System for Award Management)
• CMS Provider Data Catalog (nursing-home compare & penalties)
• USAspending.gov Custom Award Data
• CPSC Recalls (Consumer Product Safety Commission)
• NHTSA Datasets & APIs (vehicle recalls & complaints)
• UVA Corporate Prosecution Registry (academic — federal corporate plea agreements, DPAs, NPAs)

All underlying federal data is in the public domain (see usa.gov/government-works). FastDOL’s aggregation, matching, and scoring work is proprietary to FastDOL.