The Paycheck Scoreboard: Where American Wages Stand

Over the course of this series, we have examined American wages from every angle the Bureau of Labor Statistics allows: nominal and real, by industry and occupation, by state and metro area, by hours worked and worker classification. The data spans 62 years, covers every major private-sector industry, and includes more than 800 detailed occupations across all 50 states.

Three facts stand above the rest.

First, the nominal paycheck has never been higher — but real purchasing power only recently recovered from a 46-year trough. The average production worker earns $31.90 per hour today, up from $2.50 in 1964. In constant 1982–84 dollars, however, that translates to just $10.00 — barely 25% more than six decades ago. Real wages peaked in 1973, fell for 22 years, and did not recover their peak until 2019. The nominal record is a mathematical certainty in an inflationary economy. The real record is the one that matters for what a family can actually buy.

Second, where you work matters more than ever. The gap ratios tell the story: the highest-paying industry (Information, $54.00/hr) pays 2.3 times the lowest (Leisure and Hospitality, $23.30/hr). The highest-paid occupation (Family Medicine Physicians, $238,380/yr) earns 7.9 times the lowest (Cooks, Fast Food, $30,160/yr). The highest-paying state (Washington, D.C., $88,000 median) pays 2.3 times the lowest (Mississippi, $39,070). Industry, occupation, and geography are not minor adjustments to a national average — they are the national average.

Third, the American labor market is bifurcated. Knowledge-economy jobs — information technology, finance, professional services, healthcare specialties — pay between $80,000 and $238,000 per year. Service-economy jobs — food preparation, retail, hospitality — pay between $30,000 and $37,000. The two economies share a country but not a paycheck. The premium on education, credentials, and proximity to capital has widened over the past two decades, and there is no sign it is narrowing.

The table below compiles every key metric from the series into a single reference. These are the numbers that define the state of American wages in early 2026.

The horizontal bar chart below recaps the industry pay ladder from Episode 1: average hourly earnings by supersector as of January 2026. The dashed line marks the all-employee average of $37.20. Five sectors sit above it; six sit below. The gap between the top (Information at $54.00) and the bottom (Leisure and Hospitality at $23.30) is $30.70 per hour — a full-time worker in Information earns roughly $64,000 more per year than a full-time worker in Leisure and Hospitality, before taxes, benefits, or overtime.

How much has each sector’s paycheck grown over the past two decades? The table below compares average hourly earnings in January 2007 with January 2026. Every sector posted gains of at least 70%, but the pattern is revealing. Leisure and Hospitality, Information, and Financial Activities all grew by 93% — but from radically different starting points. A 93% gain on $12.10 adds $11.20 to the paycheck; a 93% gain on $28.00 adds $26.00. Percentage growth is symmetric. Dollar growth is not.

The line chart below traces nominal production worker hourly earnings from 1964 to 2026, annotated with the three eras that define the American wage story. The Golden Age (1964–1973) saw wages rise alongside productivity, real purchasing power hitting an all-time high. The Lost Decades (1973–2014) brought relentless nominal growth but stagnant or declining real wages — inflation outpaced pay raises for an entire generation. The Recovery (2014–present) finally reversed the trend: tight labor markets, minimum wage hikes, and pandemic-era competition pushed real wages past their 1973 peak for the first time in 2019.

The curve looks like progress. The annotations tell the real story.

A scoreboard is only as complete as its data. Several important dimensions of compensation fall outside the BLS wage statistics we have examined.

Benefits. Employer-sponsored health insurance, retirement contributions, and paid leave are not included in hourly earnings. For many workers, benefits represent 30% or more of total compensation. Their growth — particularly health insurance costs — has absorbed a significant share of what might otherwise have been wage increases.

Stock-based compensation. For workers in technology, finance, and senior management, equity grants and stock options can dwarf base salary. None of this appears in BLS wage data. A software engineer earning $180,000 in salary and $200,000 in restricted stock units is counted only at the $180,000 level.

Gig and informal income. The CES survey covers payroll employment at established businesses. The growing population of independent contractors, gig workers, freelancers, and informal workers is not directly captured. Their wages — often lower and more volatile — represent a shadow labor market that these statistics do not fully illuminate.

Regional cost of living. Washington, D.C.’s $88,000 median wage and Mississippi’s $39,070 do not tell you which state’s workers live better. Housing costs alone can erase a $20,000 wage advantage. The BLS publishes regional price parities, but they are not embedded in the wage data themselves.

The trends documented in this series point toward a labor market that is splitting further apart. The premium on knowledge work has widened steadily for two decades and shows no sign of reverting. The question for the next decade is whether artificial intelligence and automation will accelerate that divergence or compress it.

One scenario: AI augments knowledge workers — making lawyers, analysts, engineers, and consultants more productive and therefore more valuable — while automating service-sector tasks, reducing demand for lower-paid roles. Under this path, the 7.9-to-1 occupation gap widens further, and the bifurcation deepens.

The alternative: AI commoditizes knowledge work itself, reducing the scarcity premium on information processing, legal research, financial analysis, and routine software development. If the tools become cheap and widely available, the knowledge premium could erode, narrowing the gap from the top rather than lifting the bottom.

Both paths are plausible. Neither is certain. The data from this series provides the baseline from which any future trajectory can be measured.