What one hour of US median work bought, 1985 vs 2025

Between 1985 and 2025, the US median hourly wage rose from $8.73 to $31.34 — a nominal multiple of 3.6. Some everyday items got more expensive faster than wages and now require more hours of work to buy; others rose more slowly or fell outright. The pattern of which items moved in which direction has a structural shape: the large life-stage purchases — a year of college, a home, a month of rent — explode in wage-time, while everyday consumer goods are stable or significantly cheaper in wage-time. This page documents that pattern for six items with explicit primary sources for each.

The narrower point established by the data is that talking about US “cost of living” as a single number conceals more than it reveals. The 40-year history is not “everything got more expensive”; it is “the structurally important things got dramatically more expensive, while the daily things got cheaper or stayed flat.” Both observations are simultaneously true in the same dataset, and the structural divergence is itself the story.

The 1985 wage benchmark and what it bought

The wage figure used throughout this page is the BLS series for average hourly earnings of production and nonsupervisory employees on private nonfarm payrolls (FRED series AHETPI). This series begins in January 1964 and is the closest available proxy for the wage earned by a typical non-managerial US worker. The annual average for 1985 is $8.73 per hour. The annual average for 2025 is $31.34 per hour. The nominal multiplier is 3.59, meaning a worker in 2025 nominally earns 3.59 times what a worker in 1985 earned for one hour of work.

The six items examined here are chosen to span the typical household budget across three different categories: long-cycle aspirational purchases (a year of college tuition, a home), recurring large monthly outflows (rent, gasoline), and consumer goods bought repeatedly across a lifetime (a Big Mac, a family television). Each is sourced from a primary or institutional source documented in the methodology section. The Big Mac figure is for 1986 rather than 1985 — the Big Mac Index was launched by The Economist in September 1986 — but the price was approximately stable at $1.60 across the mid-1980s per multiple sources, so the substitution does not affect the comparison materially.

The six-item ranking

The table below shows each item’s price in nominal dollars and in hours of US median work required, for 1985 and 2025, ordered by the percent change in wage-time required to buy it.

Notes: Burgundy rows = items requiring more wage-time in 2025 than 1985. Emerald rows = items requiring less wage-time or unchanged. The Big Mac is highlighted as the structural midpoint — its price tracks the median wage almost exactly across the 40-year span. The TV comparison involves a hedonic adjustment documented in the methodology section.

What the structure reveals

1. The dichotomy is not “expensive vs cheap” — it is “long-cycle vs daily”

The dominant narrative about US cost of living over the past 40 years frames the question as one of generalized inflation: things cost more, full stop. That framing is partially correct on aggregate — the CPI rose by approximately 208% between January 1985 and December 2025, meaning a 2025 dollar buys roughly a third of what a 1985 dollar bought. But the aggregate hides the structure. The data shows a sharp dichotomy: items in three categories rose far faster than the headline inflation rate, and items in three other categories rose more slowly or fell in real terms.

The three items that exploded in wage-time share a structural feature: they are large, infrequent, life-stage purchases, and each market has substantial regulatory and supply-side friction. Public college tuition has risen partly because of state disinvestment in higher education (state appropriations per student fell substantially from 1985 to the 2010s) and partly because of expanded administrative and amenity spending; the home price has risen partly because of zoning-constrained supply that has lagged household formation; rent rises in line with home prices in most regional markets. The three items that fell in wage-time share a different structural feature: they are mass-produced, internationally traded or globally productive, and benefit from continuous productivity improvement (gas refining, fast-food labor automation, semiconductor and LCD/LED manufacturing).

2. The Big Mac is the structural fulcrum

The single most surprising data point in the table is the Big Mac. In 1986 — the year the Big Mac Index launched — a Big Mac cost $1.60 and required 11 minutes of US median work. In 2025, a Big Mac costs $5.79 and requires 11 minutes of US median work. The price multiplier (3.62×) and the wage multiplier (3.59×) are statistically indistinguishable. Over a 39-year span, the Big Mac has tracked the median wage so closely that it operates as a kind of de facto wage-deflator at the consumer-goods level.

This is consistent with the original analytic intent of The Economist’s Big Mac Index, which used the burger’s globally standardized inputs (beef, bread, labor, real estate) as a proxy for purchasing power parity. The Big Mac’s stability against the US median wage suggests that mass-produced consumer outputs with significant labor content track wages closely over the long run — which is what one would expect if labor is the dominant input. Where the Big Mac departs from this stability (briefly above 13 minutes during 2022-2024 wage and food inflation episodes; briefly below 10 minutes during the 2001-2002 deflation interval), the deviation reverts within two to three years.

3. The TV is not a single product comparison — but the directional claim is robust

The 79% decline in wage-time for a mid-range family TV is the largest single change in the table, and it warrants explicit qualification. The 1985 reference is a 25-inch color CRT television from a mid-tier brand (RCA, Zenith), typically priced at $350-450 per Statistical Abstract retail surveys and contemporary newspaper advertising; $400 is used here as the representative midpoint. The 2025 reference is a mid-range 50-inch 4K LED television from a major retailer, typically priced at $250-350; $300 is used as the representative midpoint. The two products are not the same product. The 2025 product has approximately four times the screen area, sixteen times the pixel count, materially lower power consumption, and a thinner form factor. By any consumer-utility metric, the 2025 product is dramatically superior.

The BLS publishes a hedonic-adjusted CPI for televisions (series CUSR0000SS31021), which attempts to control for these quality changes. By that measure, the real price of a television fell by approximately 90% between 1985 and 2025 — even larger than the 79% wage-time figure used here. The direction and approximate magnitude of the TV collapse are therefore robust to the choice of methodology. The structural point — that consumer electronics have become dramatically cheaper in wage-time, regardless of how one adjusts for quality — holds across multiple comparison frames.

Counter-arguments and limitations

Four objections to the framing of this page deserve direct engagement.

First, the AHE production-and-nonsupervisory series is not the only available wage benchmark, and the choice of series matters. The all-employees AHE series (which includes managers) gives a 2025 figure of approximately $37/hour, raising the wage benchmark and modestly lowering the work-time required for each item. The median weekly earnings of full-time wage and salary workers (BLS series LES1252881600) gives a slightly different trajectory, partly because of distributional shifts. The hourly-AHE-production series is used here because it has the longest continuous monthly history (back to 1964) and because it best represents the typical non-managerial wage. Using the all-employees series would lower the tuition figure to approximately 325 hours in 2025 (from 381 hours) but would not change the cluster placement. The downloadable dataset reports both.

Second, the median wage is not the typical wage for everyone. Using a single median wage as the benchmark obscures the fact that real wage gains have been concentrated at the top of the wage distribution since 1985. The bottom quintile of US wages has gained substantially less than the median in real terms over the same period. The figures presented here therefore understate the wage-time required for low-wage workers to purchase the same items. A college tuition that requires 381 hours of median work might require 600 hours or more of bottom-quintile work. The Hamilton Project at Brookings maintains an interactive tracker that adjusts these comparisons by wage decile; that finer-grained view is consistent with the structural finding here but with steeper gradients at the bottom of the distribution.

Third, the hedonic adjustment problem is significant in three of the six items. The 1985 TV and the 2025 TV are not the same product. The 1985 college education and the 2025 college education differ in technology, amenities, and learning outcomes. The 1985 family home and the 2025 family home differ in size (1985 median new home was 1,610 square feet; 2025 figure is approximately 2,100 square feet), in materials, in energy efficiency, and in included features. Strictly interpreted, the comparisons in this page describe the cost of the dominant market product at each date, not the cost of an identical good across time. The directional findings — TV got dramatically cheaper, college got dramatically more expensive — survive any reasonable hedonic adjustment, but the precise magnitudes do not.

Fourth, six items is a small sample. The full universe of household consumption includes hundreds of items the typical US household buys regularly. The choice of these six is judgment-laden; reasonable analysts could include or exclude items and produce a slightly different distribution. The broad framing — “long-cycle aspirational purchases explode in wage-time, mass-produced consumer outputs stay flat or fall” — is supported by the BLS detailed CPI breakdown for the same period: services-led categories (medical, education, housing, child care) show real price increases of 60-200%, while goods-led categories (electronics, apparel, household furnishings) show real price declines of 30-90%. The six items here are a representative cross-section of that broader divergence, but the n = 6 should not be over-interpreted.

Common misinterpretations

Reading the table as “Americans are 153% worse off in 2025.” The +153% increase in wage-time for college tuition does not mean Americans are 153% worse off overall. It means tuition specifically requires 153% more hours of work to buy. The same person who pays 153% more for tuition pays 79% less in wage-time for a comparable family TV, the same in wage-time for fast food, and 24% less for a gallon of gas. Whether the net effect on living standards is positive or negative depends on each household’s spending mix.

Equating the TV collapse with “everything in 1985 was overpriced.” The TV collapse reflects the introduction of digital manufacturing, global supply chains, and continuous semiconductor productivity gains, not a 1985 pricing anomaly. The 1985 TV was a fair price for its underlying production cost at the time. The 2025 TV is also a fair price for its underlying production cost. Both prices reflect the cost of producing the dominant television product at each date with available production technology.

Treating the “median home” cost as something the median worker could plausibly buy. The 13,255 hours of work figure for a 2025 median home is approximately six and a half years of full-time work, assuming 40 hours per week and 50 weeks per year. In practice, US homes are not bought outright in cash — they are bought with mortgages and down payments. The figure measures the gross dollar cost in wage-equivalents, not the financing or affordability calculation that an actual household would face. A more conventional measure of housing affordability — the price-to-income ratio — has gone from approximately 3.5 in 1985 to approximately 5.5 in 2025 nationally, with much higher ratios in coastal markets.

Reading the Big Mac stability as evidence that “fast food is fine.” The Big Mac price has tracked the median wage almost exactly over 39 years, but that is a statement about the Big Mac and the median wage relative to each other. It is not a statement about the quality or affordability of fast food in absolute terms. The Big Mac is also reported by McDonald’s franchise data and contemporary press coverage to be smaller in 2025 than in 1985 by approximately 10-15% by weight — a hedonic adjustment in the opposite direction from the TV case.

Methodology and sources

Frequently asked questions

Why these six items?

The six items were chosen to span three categories of household spending: long-cycle aspirational purchases (a year of college, a home), recurring large monthly outflows (rent, a gallon of gas), and consumer goods (Big Mac, TV). They were also chosen for the availability of consistent primary sources back to 1985 (or 1986 in the Big Mac case). Reasonable analysts could include additional items — eggs, a gallon of milk, healthcare premiums, an airline ticket — and each would tell its own sub-story. The directional pattern documented here (aspirational items up, consumer goods stable or down) holds across the broader BLS detailed CPI breakdown for the same period.

Why not use real (inflation-adjusted) dollars instead of hours of work?

Both approaches are valid. Real dollars adjust prices by the headline CPI, which measures changes in the average household consumption basket. Hours of work normalizes by the wage. The two approaches give the same direction of change for each item but slightly different magnitudes. The hours-of-work framing is used here because it ties the comparison to the lived experience of the person earning the money — “how long do I have to work to buy this?” — which is more directly interpretable than real-dollar adjustments and avoids the methodological debates about which CPI series to use.

Has the median wage actually risen by 3.6× since 1985?

The nominal AHE production-and-nonsupervisory wage has risen by a factor of 3.59 between the 1985 and 2025 annual averages. After adjusting for CPI inflation (which rose by approximately 3.0× over the same period), the real wage has risen by approximately 20%. That is, the typical US worker in 2025 earns about 20% more in real purchasing power than the typical US worker in 1985 — a smaller increase than commonly assumed but a positive number nonetheless. The pattern in the table reflects the fact that real-wage gains were largely offset by structural increases in the cost of long-cycle aspirational purchases.

Is the Big Mac really the same product in 2025 as in 1985?

By McDonald’s specification, yes — same recipe, same bun configuration, same sauce. By weight and size, no — multiple sources document that the Big Mac has shrunk by approximately 10-15% in beef patty weight over the period. The Economist’s Big Mac Index treats the burger as if it were identical across time and across countries. For the purposes of this page, the Big Mac is used as a wage-tracking proxy and its size change is acknowledged as a hedonic adjustment in the opposite direction from the TV case. Neither adjustment materially affects the structural finding.

What about healthcare premiums?

Healthcare premiums would belong squarely in the “exploded in wage-time” category. The KFF Employer Health Benefits Survey, the standard source for typical family employer-sponsored health insurance premiums, begins in 1999, when the average family premium was $5,791. By 2025, the same survey reports an average family premium of $26,993 — a 4.66× nominal multiple over 26 years, dramatically faster than the wage. For 1985, however, no equally rigorous source exists; the available 1985 figures from the Health Insurance Association of America and CMS National Health Expenditure data are not directly comparable to KFF methodology. Healthcare premiums are therefore excluded from this table to maintain methodological consistency, but a separate Eco3min page on healthcare costs is forthcoming.

Why no data for clothing, eggs, milk, or airline tickets?

Each of these would add an interesting data point. Clothing has fallen substantially in wage-time (BLS CPI for apparel has been roughly flat in nominal terms since 1995 while wages have risen). Eggs in early 2025 were sharply elevated due to ongoing avian flu, which would distort the long-run comparison. Milk shows roughly the same wage-time today as in 1985. Airline tickets have fallen substantially per mile in real terms since deregulation in 1978, but the 1985 fare structure is not directly comparable to the 2025 fare structure due to changes in routes, baggage fees, and add-on pricing. These items are candidates for a follow-up Eco3min analysis.

What about CEO compensation, the top 1%, or corporate profits?

This page documents the relationship between the median wage and the cost of common goods. It does not address the separate question of how income gains across the wage distribution were divided between the median worker and the top of the distribution. That question is the subject of a substantial parallel literature (Saez-Zucman, Piketty, the Economic Policy Institute) which is consistent with — but separate from — the framing here. The structural findings of this page hold regardless of one’s view on distributional questions.

Does this mean the 1985 economy was “better” or “worse” than 2025?

Neither, and this is not a value judgment the data alone can settle. A 2025 household has access to a TV with sixteen times the resolution at one fifth the wage-time. A 1985 household paid less for a year of college in wage-time. A 2025 household has access to cheaper, more efficient transportation and to communication technologies (smartphone, internet) that did not exist in 1985. A 1985 household paid less for housing in wage-time and could finance a home with less leverage. Whether one finds the 1985 mix or the 2025 mix preferable is a question of personal valuation, not of the data.

Conclusion

Forty years of US economic data, condensed into six items, reveals not a single inflation story but a structural divergence: the items associated with large life-stage transitions (going to college, buying a home, paying monthly rent) require dramatically more hours of US median work to buy in 2025 than in 1985, while the items associated with daily consumption (food, fuel, household electronics) are stable or substantially cheaper in wage-time. A year of public college tuition that required 151 hours of work in 1985 requires 381 hours in 2025. A mid-range family TV that required 46 hours in 1985 requires 10 hours in 2025. The Big Mac requires the same 11 minutes in both years.

The structural framing supported by the data is that “US cost of living” is not a single number that has either risen or fallen — it is two divergent trajectories that have coexisted for 40 years. The aspirational, long-cycle purchases that define middle-class life-stage transitions have become substantially more expensive relative to wages. The daily consumer goods and energy commodities have become cheaper or stayed flat. Both observations are simultaneously true in the same dataset, and the divergence is itself the structural feature worth attention.

This page does not constitute a forecast about which direction either trajectory will move from here, nor a policy recommendation. It is an empirical placement of six common items within their 1985-2025 wage-time history. The dataset and methodology are designed to be updated annually as new wage and price data arrive.

The data and analysis presented on this page are provided for informational and educational purposes only. They do not constitute investment advice or a recommendation to take any specific action. Eco3min is registered with the AMF as a non-prescriptive financial information publisher.