Where a Household Income Stood: U.S. Percentiles Since 1967

An inflation calculator answers one question: what a 1985 paycheck would be worth in today’s dollars. That answer is useful and incomplete. It describes what the money bought. It says nothing about where that household stood among American households at the time.

This tool adds the second reading. Enter a year and a household income, and it returns both the constant-dollar equivalent and the position of that income in the distribution of its own year — which quintile, how it compared with the median, and how far it sat from the top five percent. The two readings frequently disagree, and the disagreement is the point.

What the tool computes

Two separate operations run on a single figure.

Purchasing power. The amount is restated in 2024 dollars using the Census Bureau’s own deflator — the chained CPI-U from 2000 onward, the research series R-CPI-U-RS before that. Using the Census deflator rather than headline CPI-U keeps the tool’s constant-dollar figures identical to the ones published in the Census historical income tables.

Position. The nominal figure is compared with the income limits the Census Bureau publishes for that exact year: the upper limit of each fifth of households, the lower limit of the top five percent, and the median. The result is expressed first as an interval between two measured anchors, then as an interpolated rank inside that interval.

A third line closes the loop: holding purchasing power exactly constant, where would that same income sit in the 2024 distribution?

One example: $30,000 in 1985

A household earning $30,000 in 1985 sat between the 60th and 80th percentiles, at roughly the 62nd — comfortably above the 1985 median household income of $23,620, at 1.27 times that level.

Restated in 2024 dollars, that income is $78,204. A household with exactly that purchasing power in 2024 falls between the 40th and 50th percentiles, at roughly the 47th — below the 2024 median of $83,730.

Why the rank falls when purchasing power does not

Two mechanisms act at once, and the tool does not separate them for you — but they are separable.

The first is a level effect. Real median household income rose 53% between 1967 and 2024, from $54,880 to $83,730 in constant 2024 dollars. A fixed real income traverses a rising distribution, so its rank declines arithmetically. Income is one side of that ratio; the other is what it buys, tracked in how many hours of work a US household item costs.

The second is a shape effect. The ratio between the top-five-percent threshold and the upper limit of the lowest fifth widened from 6.3 in 1967 to 9.7 in 2024. Part of that widening reflects changes in household composition rather than pay alone — single-person and dual-earner households have both grown as shares of the total, and household income aggregates whoever lives under the roof. The direction is robust across measures; the magnitude attributable to earnings specifically is not settled.

Where the level effect dominates, ranks slip uniformly. Where the shape effect dominates, they slip fastest in the middle — which is where the 1985 example lands.

Resolution and its limits

The Census historical income tables publish six anchors per year, unchanged since 1967: the upper limits of the four lower fifths, the median, and the lower limit of the top five percent. That yields seven zones, consistently, across all 58 years — no resolution ladder, no methodological cliff in the middle of the series.

What it does not yield is any resolution inside the top five percent. An income above the 95th-percentile threshold returns “above the 95th percentile” and nothing further. Distinguishing the top one percent from the top five requires tax-based sources with different definitions of income, and mixing them into a survey-based series would produce a number that looks precise and is not.

Limits worth knowing

• The unit is the household, not the worker. All households are included — single-person, retired, multi-earner. A one-earner household in 1985 and a two-earner household in 2024 can hold the same rank while representing very different amounts of labour.
• Money income before tax. The Census concept covers wages, self-employment income, interest, dividends, rents, Social Security and other cash transfers. It excludes capital gains, employer-provided benefits, and the value of non-cash transfers such as SNAP or Medicaid, and it is measured before federal and state income tax.
• Survey estimates carry sampling error. These figures come from the CPS Annual Social and Economic Supplement, not an administrative census. Small year-to-year moves in a single percentile are not necessarily signal.
• Two methodological breaks. The redesigned income questions first applied to income year 2013, and the updated processing system first applied to income year 2017, both create discontinuities. Where the Census publishes two rows for a year, this tool uses the newer methodology.
• Rank is not a life path. The tool places an income among all households at a point in time, not among households of the same age. Position typically rises with the age of the householder and then falls after retirement.
• Interpolation between anchors. Ranks inside an interval are estimated log-linearly between two published limits. That is a readable convention, not a measurement.

Data

The file behind the tool is downloadable: quintile limits, median and top-five-percent threshold for every year from 1967 to 2024, in current dollars, with the implied Census deflator to 2024 dollars.

Download the CSV (U.S. household income percentiles, 1967–2024)

Primary sources: U.S. Census Bureau, Current Population Survey ASEC, Historical Income Tables — Table H-1 (income limits for each fifth and top 5 percent of all households) and Table H-5 (households by median and mean income), both covering 1967 to 2024.

Frequently asked questions

Should I enter household income or my own salary?

Household income — the combined pre-tax money income of everyone in the dwelling. Entering one person’s salary in a two-earner household understates the rank substantially, because the benchmark distribution counts the household’s total.

Why does the series stop at 2024?

Income for a given calendar year is collected in the March supplement of the following year and published in September after that. The 2024 income figures released in 2025 are therefore the most recent complete distribution available.

Why is mean household income so much higher than the median?

Because the distribution is right-skewed: incomes extend far above the middle and are bounded below. In 2024, median household income was $83,730 while the mean was $121,000 — a gap of 45%.

Can an unchanged income move between quintiles?

Yes, and it happens routinely during inflationary periods. If prices and incomes both rise 5% while one household’s income is unchanged, that household loses purchasing power and slips in rank. The two effects are separate and the tool reports them separately.

Does this say whether an income was “good”?

No. It reports a statistical position within a defined universe at a given date. It carries no information about household size, wealth, local housing costs, health coverage or hours worked — all of which shape material living standards and none of which a rank captures.

Last updated — 21 August 2026