When a compensation benchmark says "the median wage for software engineers in Austin is $142,000," that's a different and often more useful statement than a mean wage figure. The median tells you what a representative worker earns; the mean is distorted by outliers at both ends. For anyone setting pay ranges, defending a compensation decision, or reading public wage data, the median is typically the number that matches how workers and managers intuitively talk about "typical" pay. But it also has limitations, and knowing when to use the median versus the mean is core to any serious compensation analysis.
How the Median Wage Is Calculated Rank every worker's wage from low to high. The median is the middle value. For an odd-numbered group, it's the exact middle; for an even-numbered group, it's the average of the two middle values. For a team of seven earning $45,000, $50,000, $55,000, $60,000, $70,000, $80,000, and $200,000, the median is $60,000. The mean is about $80,000, distorted by the one high outlier.
The Bureau of Labor Statistics calculates median hourly wages using the Occupational Employment and Wage Statistics program, which surveys approximately 200,000 establishments annually. Median annual wages come from either multiplying median hourly by 2,080 or summing annual earnings for salaried workers and taking the middle value. Both metrics are published quarterly and broken out by occupation, industry, metropolitan area, and state.
When the Median Is the Right Benchmark The median is the default choice for most compensation and benchmarking work. Pay range setting: midpoint target is usually set at or near the market median. Equity analysis: comparing median wages across demographic groups surfaces systemic gaps the mean would hide. External positioning: "we pay at the 75th percentile" is a statement about the distribution, and the median (50th percentile) is its anchor. And job leveling: market medians by level are the raw input for most leveling exercises.
Mean wage has narrower but legitimate uses, especially for aggregate cost forecasting where you need the actual average per-worker dollar amount.
Why Is Median Wage Lower Than Mean Wage ? Almost always because the wage distribution is right-skewed. A relatively small number of high earners (executives, highly specialized senior roles) pull the arithmetic mean up, but the median (the middle of the ranking) stays closer to what most workers actually earn. The gap between the two is a useful indicator of wage inequality: a large gap signals compression at the low end and stretch at the top.
Does the Median Wage Include Bonus and Equity? It depends on the source. BLS OEWS medians are straight cash wages and don't include bonus, equity, or non-cash benefits. Private market surveys (Radford, Mercer, Payscale) vary; some report base-only medians, others total cash compensation, others total direct compensation including equity. When benchmarking, you need to know which definition your source uses and match it to what you're comparing internally.
Where Median Wage Data Comes From BLS OEWS is the public-sector gold standard for U.S. wage medians, published quarterly with five-year trend data. Census Bureau current population survey data is an alternative source, especially for demographic cuts. For specific industries, Radford (tech), Mercer (cross-industry), Aon (HR), and Croner (nonprofit/education) publish subscription surveys with tighter job leveling.
For any benchmark you quote, document the source, the effective date, and the job-leveling methodology. A number from 2023 BLS data is not comparable to a number from 2026 Radford data without adjustment for both time and leveling differences.
Using the Median Wage in 2026 Compensation Decisions Four practices make median wage data useful rather than misleading. Always check the effective date of your benchmark and age it forward using wage growth indexes if it's older than six months. Match job level and geography precisely, because median wages vary dramatically across both. Compare medians to your internal wage distribution at the same levels, not to your averages. And pair median with interquartile range (25th to 75th percentile) so you're looking at the shape of the distribution, not just its center.
For related compensation concepts, see mean wage , pay equity , minimum wage , and living wage . BLS Occupational Employment and Wage Statistics is at bls.gov/oes , and the Current Population Survey wage data is at bls.gov/cps/earnings.htm .
.svg)
.svg)
