Average annual consumer spending reached $81,905 in 2025, up 4.3% from $78,527 in 2024, according to Bureau of Labor Statistics results released October 29. Housing accounted for nearly half the dollar increase.
The figure describes last year's spending by a "consumer unit," which can be one person or several people who share major expenses. It is not a per-person amount or a recommended budget.
Where spending increased
These annual averages show four major categories. Dollar changes are calculated from the BLS figures.
| Category | 2024 | 2025 | Increase |
|---|---|---|---|
| Housing | $26,266 | $27,790 | $1,524 |
| Food | $10,169 | $10,633 | $464 |
| Transportation | $13,318 | $13,492 | $174 |
| Health care | $6,197 | $6,470 | $273 |
Total spending rose $3,378. Housing's $1,524 increase supplied about 45% of that change, and housing took 33.9% of the 2025 budget. These are spending amounts, so a change can reflect what people bought and how much they bought as well as prices.
Whose budget does the average describe?
Under the BLS definition, a family living together is a consumer unit. So is a financially independent person living alone or sharing a home with others. Unrelated people can form one unit when they make joint decisions about major expenses. A household address does not necessarily equal one spending unit.
Income makes a large difference. Among five groups ranked by income before taxes, average annual spending was $37,935 in the lowest fifth, $68,983 in the middle fifth and $159,512 in the highest fifth. The middle group's average is not the median spending amount for all Americans.
The averages also include people who bought nothing in a particular category. The survey FAQ explains that an average spread across all consumer units can be much lower than the amount paid by buyers alone. That is especially relevant for purchases that many people skip in a given year. A car buyer should not read an all-households vehicle average as the price of a car.
Why spending and cash payments differ
The survey's accounting rules can make its totals look different from a bank statement. BLS generally records the full acquisition cost of a purchase, including sales tax, during the period when it was acquired, even if it has not been paid off. Later installment payments for an earlier purchase are excluded from expenditures.
For homeowners, the housing category includes mortgage interest, property taxes, insurance and maintenance. Mortgage principal is excluded because BLS treats it as repayment of a loan. A household comparing its entire mortgage payment with the survey's housing figure would therefore be comparing different measures.
The total also includes categories beyond everyday purchases. Personal insurance and pensions includes employee Social Security and retirement contributions. Those definitions explain why the expenditure total cannot be treated as either a shopping bill or the amount of cash a family needs to cover each month's payments.
Find a comparison that fits
BLS publishes spending tables by characteristics including income, age, household composition and housing tenure. Its table guide distinguishes average spending per consumer unit from aggregate spending across the population. It also identifies standard errors, which describe sampling uncertainty, and the share of units reporting a purchase.
A small difference between groups or years may reflect sampling uncertainty. A higher number alone does not establish a statistically significant increase. The spending release marks significant changes separately.
Location adds another limitation. BLS cautions that regional spending comparisons also reflect differences in income, family size, age and preferences. A city's higher average expenditure is not, by itself, a measure of how much more an identical basket costs there.
How the shutdown affected the results
The 2025 federal shutdown interrupted data collection. Collection stopped in October and November and resumed December 5. Interview respondents recall earlier spending, allowing December interviews to recover some information about the missing months.
The diary survey, in which households record purchases as they occur, had no October or November data and reduced September responses. BLS increased weights on available interview reports and doubled the weights on September and December diary data to account for the missing period. The published annual results therefore incorporate adjustments rather than a full set of observations from every month.
