How the household services valuator works

Household services are the unpaid tasks a person performs to run a household: cooking, cleaning, laundry, home and vehicle upkeep, yard work, household management, shopping, child and adult care, pet care, and related travel. The replacement-cost method values that work at what it would cost to pay someone else to do it.

For a demographic group matching the individual, the tool multiplies average hours per day on each activity by the corresponding occupation's hourly wage, then annualizes. It reports three valuation methods side by side and shows the sample size, geography, and any wage fallback used. It is a planning tool that summarizes public averages; it is not a substitute for an individualized expert opinion, and it does not model household-specific facts, secondary activities, or lost-productivity (opportunity-cost) framings.

Hours come from the American Time Use Survey, accessed through IPUMS. Estimates pool multiple survey years (2019-2024) to stabilize the demographic cells and are weighted with the ATUS final person weight (WT06) so they represent the U.S. adult population rather than the raw sample.

The year 2020 is excluded because ATUS field collection was suspended for part of that year, leaving it unrepresentative of a normal year. The frame is adults age 18 and older.

Citation: American Time Use Survey Data Extract Builder. IPUMS, University of Minnesota. https://doi.org/10.18128/D060.V3.2

Hours are estimated for cells defined by five dimensions: sex, age band (18-24, 25-34, 35-44, 45-54, 55-64, 65-74, 75+), employment status (full time, part time, not employed), marital status (married or partnered vs. single), and the age of the youngest own household child (none, under 6, or 6 to 17).

A cell is flagged as thin when its unweighted sample is below the floor of 50 time diaries. When the exact cell is thin, the tool rolls up to a broader group: first collapsing the youngest-child dimension, then, if still thin, collapsing marital status as well. Every result discloses the sample size and whether a roll-up was applied.

For each activity the cell carries a weighted mean and a weighted standard error, approximated as the square root of [ sum of w_i² (x_i - mean)² ] divided by [ sum of w_i ]². This is a documented approximation; using the ATUS successive-difference replicate weights for design-consistent standard errors is a planned refinement.

Wages come from the BLS Occupational Employment and Wage Statistics program (the current release). The tool offers the hourly mean or the hourly median (50th percentile). Wages are available at three geography levels: national, state, and metropolitan area.

OEWS suppresses some cells for confidentiality or reliability. When the wage for the selected geography and occupation is suppressed, the tool falls back along a documented chain - metro to state to national - and labels every fallback it uses, so the reader can see exactly which geography priced each activity.

Each activity is priced with one or more Standard Occupational Classification (SOC) codes. Where an activity lists several SOCs, the first with an available (non-suppressed) wage for the chosen geography is used. The generalist method prices all hours at a single housekeeping wage.

Daily figures are annualized as daily times 365.25, and weekly figures as daily times 7. The tool reports three methods:

1. Occupation-specific: for each activity, hours times that activity's matched occupation wage; the daily total is the sum across activities.
2. Generalist: total household hours times a single generalist housekeeping wage.
3. Composite: a blended hourly rate equal to the occupation-specific daily total divided by total hours, applied to total hours. By construction this equals the occupation- specific total; it is shown to make the effective blended rate explicit.

• This is a replacement-cost estimate (what it would cost to hire the work out), not an opportunity-cost estimate (the person's own foregone earnings). The two can differ substantially.
• Only primary-activity time is counted. Secondary activities - most importantly childcare that happens while doing something else - are not included in this version, so care time is conservative.
• ATUS time diaries are self-reported for a single day, and hours reflect group averages rather than any one household.
• Standard errors use a weighted approximation rather than design-based replicate weights (see above).
• Wages are area averages for an occupation, not the price of any specific hire.

Every figure derives from public data: ATUS 2019-2024 (via IPUMS) for hours and BLS OEWS (the current release) for wages. The category definitions, demographic cells, occupation crosswalk, and valuation formulas are fixed for a given data vintage, so a result is reproducible from the inputs shown on each valuation. The Word and Excel exhibits restate the inputs, hours, wages, methods, and these citations; the Excel workbook exposes the arithmetic as live formulas.