BLS Data Validation Study

Exhaustive source-file comparison of all BLS labor market data

Date:

February 14, 2026

Scope:

559,010 data points · 2 BLS datasets

Result:

100.0% verified

559,010

Data Points Verified

389

Metro Areas Checked

0
Discrepancies Found

Data Sources Validated

Dataset	BLS Source File	Data Points	Match Rate	Status
OEWS Regional Wages & Employment	MSA_M2024_dl.xlsx	553,192	100.0%	✓
National Employment Projections	occupation_1_.xlsx	5,818	100.0%	✓

Validation Methodology

1. OEWS Regional Data

BLS OEWS Excel MSA_M2024_dl.xlsx

→

Row-level Match Area × SOC code

→

553,192 Verified 0 discrepancies

OEWS — Exhaustive Source Comparison

The complete BLS Occupational Employment and Wage Statistics file MSA_M2024_dl.xlsx was downloaded from bls.gov/oes/tables.htm (May 2024 release). Every row with AREA_TYPE = '2' (metropolitan) and O_GROUP = 'detailed' (141,164 rows) was compared against the corresponding entry in our deployed oews_msa_minimal.json, keyed by area code + SOC code. Four fields were compared per row: TOT_EMP, A_MEDIAN, LOC_QUOTIENT, and JOBS_1000. An additional 9,012 summary-level occupation groups (e.g., "11-0000 Management Occupations") are present in the JSON for completeness but are never displayed in individual reports.

Result: 553,192 of 553,192 Values Verified

Four fields × 138,298 occupation-metro pairs = 553,192 individual values compared. Every value in the backend matches the BLS source file exactly. Zero discrepancies found. An additional 11,464 null-pair values (both source and system null due to BLS suppression) were confirmed as expected.

2. National Employment Projections

BLS EP Table 1.2 occupation_1_.xlsx

→

Row-level Match 832 SOC codes

→

5,818 Verified 17 top-coded*

National Projections — Exhaustive Source Comparison

The 832 occupations in bls_projections_2024_2034.json were compared row-by-row against BLS Employment Projections Table 1.2 (occupation_1_.xlsx), downloaded from bls.gov/emp. Seven fields were compared per occupation: employment_2024, employment_2034, change_numeric, change_percent, annual_openings, median_wage, and education. Of 5,818 values compared, 5,801 were exact matches. The 17 differences are all high-wage medical specialties (surgeons, anesthesiologists, etc.) where BLS reports "≥ $239,200" — the Excel cell is blank, and our system correctly stores the BLS top-code value of $239,200. This is standard BLS wage top-code handling.

Additionally, mathematical consistency was verified: employment_2024 + change_numeric ≈ employment_2034 held for all 832 occupations with zero errors, and all 832 outlook labels matched the published growth-rate thresholds.

Cross-Dataset Plausibility

National median wages from the EP projections were compared against the regional OEWS distribution for 763 occupations with data in 10+ metros. For all 763, the national wage fell within the regional p5–p95 range, confirming cross-dataset consistency.

Sample Verification Detail

Both datasets were validated exhaustively by automated comparison. The following occupations were additionally verified by manual lookup on bls.gov/ooh on February 13, 2026:

SOC	Occupation	BLS Median Wage	Our Value	Growth	Result
31-9092	Medical Assistants	$44,200	$44,200	+12.5%	✓
29-1141	Registered Nurses	$93,600	$93,600	+4.9%	✓
43-4051	Customer Service Reps	$42,830	$42,830	−5.5%	✓
29-2061	LPN/LVN	$62,340	$62,340	+2.6%	✓
15-1252	Software Developers	$133,080	$133,080	+15.8%	✓

Regional Verification

Metro Area	SOC	Field	BLS Source	Our Value	Result
Raleigh-Cary, NC	43-4051	Median Wage	$41,020	$41,020	✓
Raleigh-Cary, NC	43-4051	Employment	14,770	14,770	✓
Charlotte, NC-SC	31-9092	Median Wage	$45,330	$45,330	✓
Charlotte, NC-SC	31-9092	Employment	6,890	6,890	✓
Raleigh-Cary, NC	31-9092	Median Wage	$42,900	$42,900	✓

How to Verify

Any data point in a Semantic Insight report can be independently verified:

Regional data (wages, employment by metro): Download the OEWS dataset from bls.gov/oes/tables.htm → "Metropolitan and nonmetropolitan area" → select the May 2024 release. Open the Excel file and filter by your report's area code (shown in report metadata) and SOC code.

National projections (growth outlook, openings): Download Table 1.2 from bls.gov/emp (the "Occupation" XLSX file). Filter by SOC code. The employment, growth rate, openings, and wage columns will match the figures in your report. Alternatively, visit bls.gov/ooh and search by occupation title.

Notes & Known Limitations

Wage top-coding. BLS does not publish exact median wages for occupations earning above $239,200 per year. The source Excel leaves these cells blank. Our system stores the BLS standard top-code value of $239,200 for these 17 occupations (all medical specialists: surgeons, anesthesiologists, cardiologists, etc.). This is the standard handling recommended by BLS and used by all systems that consume this data.

Two data sources, two survey methodologies. National projections (Employment Projections program) and regional data (OEWS survey) are produced by different BLS programs. National median wages may differ slightly from the weighted average of all metro medians. For Software Developers (15-1252), the national median of $133,080 is slightly above the OEWS metro 90th percentile of $131,050 — this reflects the impact of high-paying tech metros and non-metro employment, and is consistent with BLS methodology.

Suppressed data. BLS suppresses values in some metro × occupation cells to protect employer confidentiality. These appear as null values in both the source Excel and our system. The audit found 11,464 null-pair values (both source and system null), which is expected.

Growth percentage rounding. The EP table publishes change_percent independently from the rounded employment figures. For 161 of 832 occupations, recalculating the percentage from the rounded employment figures produces a slightly different value (median difference: 0.4 percentage points). Our system uses the BLS-published percentage, which was verified to exactly match the source Excel for all 832 occupations.

Data Currency & Refresh Schedule

Dataset	Current Version	BLS Publication Cycle	Next Expected Release
OEWS Regional Wages	May 2024	Annual (spring)	~April 2026 (May 2025 data)
Employment Projections	2024–2034	Every 2 years	~2027 (2026–2036 projections)

When BLS publishes updated data, we re-run this validation process against the new source files before deploying updated data to reports. The automated validator (bls_data_validator.py) is included in the CI pipeline to catch any discrepancies during the update process.