Data Sources
Transparent documentation of all data sources, APIs, and services powering Wage Atlas salary intelligence.
Primary Data Sources
U.S. Bureau of Labor Statistics (BLS)
Occupational Employment & Wage Statistics (OEWS)
Annual surveys covering approximately 800 occupations across all U.S. states and metropolitan areas. Provides mean and median wages, employment levels, and percentile distributions (10th, 25th, 75th, 90th).
Data Years: 1997-2025
Update Frequency: Annual
Coverage: National, State, and MSA levels
Standard Occupational Classification (SOC)
Federal classification system organizing occupations into hierarchical categories. We use SOC 2018 codes for consistent occupation identification across all data years.
Version: SOC 2018
Bridge: SOC 2010 → SOC 2018 mapping
BLS Education & Training Data
Required education levels and training pathways for occupations. Used to identify no-degree-required positions and entry-level opportunities.
Bureau of Economic Analysis (BEA)
Regional Price Parities (RPP)
Cost-of-living adjustments by metropolitan area and state. Used to calculate "real wages" that account for purchasing power differences across locations. Available for Pro subscribers.
Coverage: States and select Metropolitan Statistical Areas (MSAs)
Update Frequency: Annual
Use Case: Cost-of-living adjusted salary comparisons
Additional Government Sources
- Industry Classifications: NAICS (North American Industry Classification System) codes for industry-specific analysis. We use NAICS 2022 with bridges from legacy NAICS 2017 and SIC codes.
- Economic Indicators: Unemployment rates, GDP growth, and labor force statistics for economic context in trends analysis.
Data Processing & Infrastructure
Data Transformation Pipeline
dbt (Data Build Tool)
We use dbt to transform raw BLS data into clean, structured fact and aggregate tables. This includes data quality checks, suppression handling, and bridge tables for classification systems.
Supabase Postgres
Our primary database stores processed OEWS data, aggregates, and user data. Optimized with indexes and year-partitioned tables for fast queries.
Data Quality & Validation
- Suppression Handling: Properly identifies and handles BLS suppression indicators (#, *, **, -) in the data
- Classification Bridges: Maps historical SOC 2010 codes to SOC 2018, and legacy NAICS/SIC codes to NAICS 2022
- Data Validation: Automated checks for data consistency, completeness, and accuracy
Third-Party Services
We use trusted third-party services to support platform functionality:
Google AdSense
Advertising platform that displays relevant ads to help support free access to salary data.
PayPal
Secure payment processing for one-time PDF report purchases.
Data Coverage & Limitations
Geographic Coverage
- National Level: Complete coverage for all 50 states and Washington D.C.
- State Level: All states and territories included in BLS OEWS surveys
- Metropolitan Areas: Coverage varies by MSA size and occupation density. Smaller MSAs may have limited data for specialized occupations.
Data Limitations
Suppression Indicators
BLS suppresses data to protect confidentiality or when sample sizes are insufficient. We preserve these indicators:
- # - Data not available
- * - Data not available or does not meet publication standards
- ** - Data not available or does not meet publication standards (alternative indicator)
- - - Data not available or suppressed
Historical Data Considerations
- Classification Changes: Industry and occupation classifications have evolved over time (SOC 2010 → 2018, NAICS updates, SIC → NAICS). We use bridge tables to maintain consistency.
- Methodology Changes: BLS has updated OEWS methodology over time (e.g., MB3 era changes). We preserve methodology metadata for transparency.
- Geographic Boundaries: Metropolitan area boundaries have changed over time. We use the most current definitions available.
Attribution & Licensing
Government Data Attribution
All government data used by Wage Atlas is in the public domain and provided by federal agencies. We maintain proper attribution to ensure transparency and compliance.
BLS Data: "Data from the U.S. Bureau of Labor Statistics, Occupational Employment and Wage Statistics program."
BEA Data: "Regional Price Parities data from the U.S. Bureau of Economic Analysis."
While the underlying government data is public domain, our analysis, visualizations, calculations, and proprietary insights are protected by copyright. Users may cite Wage Atlas with proper attribution.
Data Usage Rights
- Government data is public domain and may be used freely with proper attribution
- Wage Atlas analysis and visualizations may be cited for research and reporting purposes
- Bulk data scraping or automated data extraction is prohibited per our Terms of Service
Data Freshness & Updates
Update Schedule
BLS OEWS Data
Updated annually, typically released in the spring for the previous year's data. We integrate new releases within 30 days of publication.
BEA RPP Data
Updated annually, typically released in the fall. We integrate new releases within 30 days of publication.
Current Data Year
Our platform currently includes data from 1997 through 2025, with the most recent year representing the latest available BLS OEWS release. Historical data is preserved to enable long-term trend analysis.
Why We Build on Official Sources (and Not Crowdsourced Data)
Most salary websites are built on self-reported figures: users type in what they earn, and the site averages the submissions. That approach produces large sample sizes but a systematically skewed sample — people who submit salaries online are younger, better paid, more urban, and more concentrated in tech-adjacent fields than the workforce as a whole. Nobody verifies the numbers, and there is a well-documented tendency to round up.
The OEWS survey that powers Wage Atlas works the other way around: a statistically designed sample of about 1.1 million business establishments, with wages reported by employers from payroll records, weighted to represent the full economy, and published with documented error bands. It covers occupations crowdsourced sites barely see — there are far more home health aides than data scientists in America, and only one of those groups posts salaries on the internet.
The trade-off is honest: official data is slower (annual releases, three-year survey windows) and excludes equity and bonus compensation. For the fast-moving top of the tech market, crowdsourced data adds color. For everyone else — and for any comparison across occupations, states, or years — the federal survey is the only data with a defensible claim to representing reality. We document its limitations openly on our methodology page rather than pretending it has none.
Citing Wage Atlas
Journalists, researchers, and students are welcome to cite our pages. Please attribute the underlying figures to "U.S. Bureau of Labor Statistics, OEWS (via Wage Atlas)" and link to the specific page used. For custom data questions, contact research@wageatlas.com.
Related Resources
Learn more about how we process and present this data:
Questions About Our Data?
We're committed to transparency about our data sources and processing. If you have questions about our data, sources, or methodology, please reach out:
Data Questions: data@wageatlas.com
Research Inquiries: research@wageatlas.com
Technical Issues: support@wageatlas.com
Note: For questions about the underlying government data itself, please refer to the BLS OEWS program or BEA directly.