What Retirees Actually Spend

Updated August 3, 2026. Quick answer: in 2024, the average American household headed by someone aged 65 to 74 spent $64,232 a year. Households headed by someone 75 or older spent $54,915. But the average is the wrong number to plan with, and the reason is the finding of this study: the typical household spends far less than the average one. The median for 65 to 74 is $50,068 — about 22 percent below the mean.

What these numbers are. Computed by Clear Money Guide from the Bureau of Labor Statistics Consumer Expenditure Public-Use Microdata, Interview survey, 2024 (released 2025). Figures are annual means per consumer unit, weighted with FINLWT21, with standard errors from the 44 balanced-repeated-replication weights. Our totals reconcile to BLS’s own published Table 1300 within about 2 percent. The Interview survey is not the whole survey — BLS’s published table also folds in the Diary survey — so our category figures run slightly below theirs, which we state page by page rather than hide.

The table

Medians are annualised from each household’s single observed quarter, which widens the spread relative to a true annual distribution. We publish them because no official table does, and we flag the limitation rather than omit the number.

Why the mean misleads

Across all three cohorts the mean sits about 28 to 32 percent above the median. That gap is not noise — it is the shape of the distribution. A relatively small number of high-spending households pull the average up, and every retirement calculator that asks you to compare yourself to “the average retiree” is quietly comparing you to a household that spends more than most.

If you are budgeting, the median is the more useful anchor. If you are reading a headline, it is almost certainly quoting the mean.

What we checked our work against

Every figure was reconciled against BLS’s own published Table 1300 before publication. Household counts land within 0.3 percent, income within 1 percent, and total spending within about 2 percent. Where we diverge we say so:

What we deliberately do not publish

Food. The Interview survey under-captures it badly, because food at home is collected mainly in the companion Diary survey. Our figure came in at roughly 55 to 62 percent of the published integrated one, so we dropped the category rather than print a number we knew was wrong by 40 percent. Housing, which runs about 8.5 percent below published for the same reason, is published with that shortfall stated.

The rest of the study

• Does spending really fall with age? Our answer
• Health care — the one category that rises
• Housing, and the mortgage that did not get paid off
• The gap between income and spending, and where it goes
• The reference numbers on one page

Method, in full

The method was written down and committed before any estimate was computed, so the results could not be chosen after seeing them. Three of the pre-registered decisions turned out to be wrong — including one that was off by a factor of three — and all three are recorded as amendments rather than quietly edited. That record is part of the study.

Cite this

Clear Money Guide (2026). Retirement Spending by Age from CE PUMD 2024: Medians, Tenure Splits and BRR Standard Errors [Data set]. Zenodo. https://doi.org/10.5281/zenodo.21786091

The deposit contains the estimates, the full computation script, the reconciliation against BLS Table 1300, and the pre-registered method including the three decisions that turned out to be wrong — recorded as amendments rather than edited away. CC-BY-4.0.