Data Methodology

· Labor Condition Application (LCA) — DOL OFLC FY2024–2025. CERTIFIED records only. An LCA filing is a required step before an H1B petition, but not every LCA leads to a petition or approval. LCA volume indicates employer filing intent, not confirmed hires.

· PERM (Program Electronic Review Management) — DOL OFLC FY2021–2025. Both CERTIFIED and CERTIFIED-EXPIRED records are included. Both statuses represent applications that received DOL labor certification. CERTIFIED-EXPIRED indicates the certification was issued but the I-140 petition was not filed within the 180-day validity window — the underlying DOL approval is identical, so we include both as historical sponsorship activity. PERM is the first step in EB-2/EB-3 employment-based green card sponsorship.

· H1B Employer Data Hub — USCIS FY2022–2024. Approval counts by employer, used to cross-validate LCA volumes.

· Visa Issuance Statistics — U.S. Department of State FY2024. Used for nationality-level data (H1B, E-2, L-1 by country).

· Type 1 normalization only: Standardizing punctuation, trailing commas, and period variants (INC./INC, LLC./LLC, LLP./LLP).

· We do not merge subsidiaries into parent groups. Amazon, Amazon Web Services, and Amazon Fulfillment are tracked separately — their hiring profiles differ meaningfully.

· NAICS hybrid system across 60 sectors. For codes 54xx (Professional Services) and 51xx (IT/Media), we use 4-digit codes. Others use 2-digit.

· NAICS codes are self-reported by employers on LCA filings. Some classifications may feel counterintuitive (e.g., JPMorgan files under HOLDING_COMPANIES, not BANKING). We do not override self-reported codes.

· MEAN_LCA_WAGE_FY2025 — Arithmetic mean of offered wages reported on certified LCA filings (WAGE_RATE_OF_PAY_FROM field). Annual wages only. LCA wages represent employer-reported offered wages and may not reflect total compensation, bonuses, or equity. For a more representative measure, see MEDIAN where available.

· MEDIAN_LCA_WAGE and P75_LCA_WAGE — Shown only for employers with 4+ FY2025 LCA records, to avoid single-filing distortion.

· PREVAILING_WAGE_AVG — DOL-set minimum wage for the occupation and location. Offered wages typically exceed the prevailing wage; the difference varies by occupation, location, and employer. Individual offers vary significantly.

· PERM offered wages — Mean of WAGE_OFFER_FROM on certified PERM applications. Same caveats as LCA wages apply.

· Important: All salary figures reflect government-reported wages for visa-sponsored positions. Actual market compensation may differ.

· Where shown, "PERM certification rate" = CERTIFIED records ÷ (CERTIFIED + CERTIFIED-EXPIRED) records for that employer. This metric reflects the share of DOL-approved applications that the employer subsequently moved forward (filed I-140 within 180 days), not the DOL approval rate itself.

· A low rate may indicate workforce changes, business decisions, or strategic shifts after DOL approval — not DOL rejection.

· We use the following definitions to categorize year-over-year LCA filing changes. These labels describe historical filing patterns only and do not predict future sponsorship activity.

· INCREASING — FY2025 ≥ FY2024 × 1.10

· STABLE — within ±10% year-over-year

· DECREASING — FY2025 ≤ FY2024 × 0.90

· NEW — filed in FY2025 but not FY2024

· STOPPED — filed in FY2024 but not FY2025

· LCA volume ≠ H1B approvals. Not every LCA leads to a petition, and not every petition is approved. LCA data represents employer applications to the Department of Labor, not confirmed hires or USCIS decisions.

· PERM ≠ green card issuance. PERM is the first of three steps (PERM → I-140 → I-485) for EB-2/EB-3 sponsorship. A certified PERM does not confirm green card approval or current employment.

· Entity fragmentation: Large employers (Deloitte, PwC, IBM) file under multiple legal entities. Totals may be understated.

· IT outsourcing firms: Firms like Cognizant, TCS, and Infosys dominate raw counts and may distort sector averages.

· Data cutoff: Data covers filings through FY2025 (ending September 2025). Post-cutoff hiring changes are not reflected.

· NAICS self-reporting: Sector classifications rely on employer-reported NAICS codes, which may not align with common industry perception.