The Pay Report
Employers’ posted pay offers are moving higher, but the gains are not evenly distributed across the wage ladder. Indeed Hiring Lab’s September 24 U.S. labor-market snapshot estimates that wages advertised in job postings rose 2.5% over the year ending August 2026. Its comparison by occupational pay level shows growth accelerating for traditionally higher-paying occupations, from a 2.0% annual rate in January to 2.6% in August, while growth for low- and middle-wage occupations moved sideways, with middle-wage roles lagging. The tracker describes advertised compensation, not wages that workers ultimately receive.
Indeed characterizes overall posted-wage growth as stable over the past six months or so, but says the gains “aren’t being distributed equally.” The occupation comparison is a three-month moving average of year-over-year growth in the Indeed Wage Tracker.
Advertised pay is not a wage index
The distinction matters for anyone treating a published salary range as a measure of pay. Employers may revise posting language, shift the mix of jobs they advertise or raise offers for hard-to-fill specialist roles without delivering the same gains to existing staff. Indeed’s method controls for job titles when measuring posted-wage growth, but it is still based on positions advertised on its platform.
Official earnings data ask a different question. The Bureau of Labor Statistics reported September 11 that average hourly earnings rose 0.3% from July to August while consumer prices rose 0.4%, so real average hourly earnings fell 0.1% over the month and were down 0.3% from a year earlier. Real average weekly earnings rose 0.2% in August because the average workweek lengthened. An advertised-pay tracker cannot tell a worker whether their own paycheck is keeping up with prices.
More postings, not yet more hires
The snapshot also reported firmer demand for workers. Indeed’s Job Postings Index rose 1.5% over the month to 103.5 on September 18, its highest level since late March and about 3% above the February 2020 baseline. The companion index for new postings—those seven days old or fewer—was 94.0, about 6% below that baseline and still drifting slightly lower. Indeed says the improvement “has not reached the JOLTS data”: the official hires rate was 3.2% in July, with quits at 1.9% and layoffs at 1%. For compensation professionals, the practical issue is which roles have genuinely become more competitive and which advertisements have simply stayed visible longer.
What it means for salary benchmarking
The occupational split raises questions about entry-level pay in particular. A national average can hide the very roles small employers recruit most often. Salary benchmarking should start with the exact role, seniority, location and work arrangement, then identify whether a benchmark is advertised pay, measured employee earnings or employer-reported labor cost. Blending those measures can create false precision even when each source is reputable.
The next tests are whether stronger advertising produces more completed hires and whether slower pay growth in middle-wage roles persists. The August JOLTS report is due September 29 and the next Real Earnings release on October 14; neither will reproduce an occupation-level posted-pay tracker. Until comparable data arrive, the evidence points to an uneven advertised-pay market rather than a universal raise.
Sources
- Indeed Hiring Lab, “US Labor Market Snapshot: September 2026,” September 24, 2026. https://hiringlab.indeed.com/2026/09/24/us-labor-market-snapshot-september-2026/
- U.S. Bureau of Labor Statistics, “Real Earnings — August 2026,” September 11, 2026. https://www.bls.gov/news.release/realer.nr0.htm
- U.S. Bureau of Labor Statistics, “Job Openings and Labor Turnover — July 2026,” September 1, 2026. https://www.bls.gov/news.release/jolts.nr0.htm
Featured photo: A job interview in progress. Representative image of hiring conversations; it does not depict any employer or posting in the Indeed data. Photo: Amtec Photos (www.amtec.us.com), via Wikimedia Commons (CC BY-SA 2.0). Cropped from the original.
.jpg){kind=link}