FinancialDepth

Spending Benchmark

What retirees actually spend, by age and category

Every retirement calculator asks you for an annual spending figure and almost none of them tell you what a plausible figure looks like. This page is the reference: official Bureau of Labor Statistics survey means for ages 55–64, 65–74 and 75 and older, broken down by category, with one adjustment most sources never make.

BLS Consumer Expenditure Survey, 2024Every figure cell-traceableNominal dollars, not deflated

Age of reference person

Household basis

Size and age come from different BLS tables, so choosing a size replaces the age view rather than filtering it.

%

Used only to translate the benchmark into a portfolio size.

Average annual spending

$60,775

Per consumer unit, per year, 2024 dollars.

Monthly equivalent

$5,065

Portfolio needed at 4.0%

$1.52M

The portfolio figure assumes the whole benchmark comes from savings. Any Social Security, pension or part-time income reduces it — see the reverse income calculator.

Where the money goes

The full breakdown, all three age bands

These are means per consumer unit per year, in nominal dollars of 2024, from BLS Table 1300. The share column is that category as a percentage of the band's own total spending, which is where the interesting movement is.

Category55–6465–7475+Share at 75+

Categories shown do not sum to the total: BLS publishes further lines (apparel, education, alcohol, tobacco, reading) that are omitted here for readability.

The category that should not be in your retirement budget

BLS reports a category called personal insurance and pensions. It sounds like an expense a retiree would face. It mostly is not. The bulk of it is retirement plan contributions and Social Security payroll deductions — money going into savings — plus life insurance premiums.

The age pattern gives it away. At ages 55–64, when most people in the band are still working, it runs $12,172 a year. By 75 and older it has collapsed to $1,908. Nothing else in the table falls by 84%.

If you take the headline BLS figure and use it as your retirement spending assumption, you are budgeting to keep contributing to a retirement plan you are now living off. That inflates the portfolio you think you need. Removing it changes the benchmark meaningfully:

Age bandBLS headlineInsurance & pensionsRetiree-adjustedPortfolio saved at 4%
55–64$84,946−$12,172$72,774$304,300
65–74$65,354−$4,579$60,775$114,475
75 and older$55,834−$1,908$53,926$47,700

The last column is the portfolio you do not have to accumulate, at a 4% withdrawal rate, purely from correcting this one line. For someone planning around the 65–74 band that is about $114,000 of over-saving. It is not a rounding error, and it is the single most common way the BLS average gets misused in retirement writing.

One person is not half a couple

A widowed or single retiree cannot just halve a couple's budget. BLS Table 1400 breaks spending down by household size, and the arithmetic does not cooperate: one-person consumer units spend $48,794 a year against $80,830 for two-person units. That is 60% of a two-person budget, not 50%.

Housing is the reason. One person spends $19,052 on housing versus $26,566 for two — 72% of the couple's figure, because a roof, property tax and utilities do not halve when a household does. Food behaves the way you would expect (55.6%) and transportation nearly halves (49.6%), but housing is the largest line in the budget and it barely bends.

This matters for survivor planning specifically. Social Security drops to the larger of the two benefits when one spouse dies, and any pension may drop or stop. Household spending, on this evidence, falls by only about 40%. That gap is the survivor problem in one number, and it is why the portfolio amount pages treat couples and singles differently rather than scaling one to the other.

Three years of the 65–74 band

BLS publishes the same table each year. Below is 2022 to 2024 for the 65–74 band, in each year's own nominal dollars.

Category202220232024Change 2022→2024

These are nominal changes, not real ones. We have not deflated the series to constant dollars. Cumulative U.S. price inflation over 2022–2024 accounts for a large share of every increase in that last column, so treat it as a description of dollars spent, not of consumption rising. Doing the deflation properly requires choosing a specific CPI series, base year and seasonal-adjustment convention, and stating them — not quietly subtracting a number.

One line does move against the tide. Healthcare for the 65–74 band fell in nominal terms between 2023 and 2024, from $7,942 to $7,715. In a period of general price increases, a nominal fall means the real decline was larger still. A single year in survey data is not a trend, but it is a useful corrective to the assumption that retiree healthcare spending only ever rises.

What this benchmark is not

It is not a target. These are population means. Means are pulled upward by high spenders, so more than half of consumer units in each band spend less than the figure shown. Landing below the average is the common case, not a failure.

A consumer unit is not a person, and not reliably a household. BLS defines it by shared financial arrangements. The age bands are the age of the reference person only, not of everyone in the unit, so a 66-year-old with an adult child at home appears in the 65–74 column with that child's spending included.

It is not filtered by retirement status. The 55–64 column in particular is dominated by people still working, which is exactly why its insurance-and-pensions line is so large. Even the 65–74 column contains people who have not retired.

It says nothing about your geography or health. Housing at $22,329 is a national mean across Manhattan and rural Mississippi. Healthcare at $7,715 is a mean across people on Medicare with supplements and people with chronic conditions. Neither describes an individual.

BLS publishes uncertainty; this page does not display it. The source workbooks include standard errors and relative standard errors for every cell. We extract the means only. For any cell you intend to lean on hard, open the source workbook and check the relative standard error before treating the figure as precise.

Use this with

Frequently asked questions

How much does the average retiree spend per year?

For 2024, consumer units with a reference person aged 65–74 spent $65,354 on average and those 75 and older spent $55,834. Removing the personal insurance and pensions line, which is mostly contributions rather than spending, gives $60,775 and $53,926.

Does spending really fall in retirement?

In this data it falls with age, but unevenly. Transportation more than halves from $15,085 to $6,855 between the youngest and oldest bands. Housing barely moves, $27,019 to $21,999. Healthcare rises, $6,711 to $7,918. The shape of the budget changes as much as its size.

Why is the 55–64 figure so much higher?

Mostly because that band is still working. Its insurance-and-pensions line alone is $12,172 against $4,579 for 65–74, and its transportation and food lines carry commuting and larger households. It is a pre-retirement benchmark, not an early-retirement one.

Where exactly do these numbers come from?

BLS Consumer Expenditure Survey published tables: Table 1300 for age of reference person, Table 1400 for size of consumer unit, Table 1502 for composition. The 2024 tables were released in December 2025 and are the latest annual demographic tables available. Each figure on this page is stored with its exact workbook cell coordinate so it can be re-checked against the source.

Sources

  • U.S. Bureau of Labor Statistics, Consumer Expenditure Survey tables — Table 1300 (age of reference person), Table 1400 (size of consumer unit), Table 1502 (composition of consumer unit), calendar years 2022, 2023 and 2024.
  • U.S. Bureau of Labor Statistics, Consumer Expenditures — 2024, released December 19, 2025.
  • Data extracted 2026-08-20. Values are means per consumer unit per year in nominal dollars of the data year. No CPI adjustment has been applied. See our methodology.