The museum · US
The 1920s room
Selected observations
Signals at the walls
Which parts of modern life appear before the household is fully measured?
Radio: commercial broadcasting began 1920, ~40% of households by 1930. …Radio and automobile, 1920sAFood prep 19.9, house cleaning 9.3, clothing care 11.5, childcare 7.2, …How a 1920s housewife's week broke down (Wilson study)C$0.55/hr, 47.4 hrs/week, $26.02/wkManufacturing production worker wages, 1920Ano reliable record accessible onlineHome value, rooms, plumbing, electricityD
The composite family. Each room is a statistical composite, assembled from separate distributions with separate sources. The family at the median income did not also have the median house, the median car, and the median diet. No single family described here ever existed; each fact tells you, in its provenance drawer, which real population it was measured from.
About this selection · Focused exhibit
The selected route concerns communications and paid and unpaid work. It cannot fill the room's missing household budget and housing evidence.
This describes how the page is arranged; it is not an evidence score.
Objects & everyday life
A closer look
Each label names its measurement and date. Drawings identify subjects, not exact appliance models.
Follow a tour: The work that moved
Inspect the evidence
Browse the records
Every observation includes its measured population and source. Calculations identify their inputs; gaps explain what remains unavailable or unverified.
- Recorded observations
- 16
- Calculated exhibits
- 0
- Documented gaps
- 5
These are counts of collection entries, not people or households. Observations exclude gaps; calculated exhibits are counted separately. A gap is an unanswered or unverified item, not a zero value.
The home
2 observations · 1 gaps45.6%A
Homeownership rate, 1920observed record
% of occupied dwelling units (owner-occupied)
Population measuredAll U.S. occupied housing units (decennial census)
Source recordU.S. Census Bureau · 2000 · Tier A
Full record
Homeownership Rate by State: 1900 to 2000
U.S. Census Bureau, 2000 · source
Confidence: A — Official series
Curator note: From Census Bureau decennial time series (owner.pdf). The low point of the early-century decline. Homeownership had been falling since 1900 (46.7%) as urbanization and renting increased. The 1920s boom would start the recovery.
Source note: Time series of homeownership rates from 1900 to 2000 by state and nationally. 1950 national rate: 55.0%. 1940: 43.6%. 1960: 61.9%. Also see companion table: https://www2.census.gov/programs-surveys/decennial/tables/time-series/census-housing-tables/ownerchar.pdf
Assumption: The composite family
U.S. Census Bureau, 2000 · source
Confidence: A — Official series
Curator note: From Census Bureau decennial time series (owner.pdf). The low point of the early-century decline. Homeownership had been falling since 1900 (46.7%) as urbanization and renting increased. The 1920s boom would start the recovery.
Source note: Time series of homeownership rates from 1900 to 2000 by state and nationally. 1950 national rate: 55.0%. 1940: 43.6%. 1960: 61.9%. Also see companion table: https://www2.census.gov/programs-surveys/decennial/tables/time-series/census-housing-tables/ownerchar.pdf
Assumption: The composite family
no reliable record accessible onlineD
Home value, rooms, plumbing, electricitydocumented gap
n/a
Population measuredAll US housing units and residents (decennial census)
Source recordU.S. Census Bureau · 2024 · Tier D
Full record
Census of Population and Housing — Decennial Publications
U.S. Census Bureau, 2024 · source
Confidence: D — Scholarly estimate
Curator note: No Census of Housing until 1940. Urban electrification was spreading rapidly in the 1920s; by 1930, ~85% of urban/non-farm homes had electricity. Rural electrification remained minimal (~11% of farms by 1935). Indoor plumbing was standard in new urban construction but rare in rural areas.
Source note: Historical Census publications. Census of Housing 1940 was first to ask about amenities (plumbing, electricity, radio, refrigerator). Census 1930 asked about radio ownership — first census technology-diffusion question. Census 1950/1960 asked about TV, telephone, automobile, refrigerator, washing machine.
U.S. Census Bureau, 2024 · source
Confidence: D — Scholarly estimate
Curator note: No Census of Housing until 1940. Urban electrification was spreading rapidly in the 1920s; by 1930, ~85% of urban/non-farm homes had electricity. Rural electrification remained minimal (~11% of farms by 1935). Indoor plumbing was standard in new urban construction but rare in rural areas.
Source note: Historical Census publications. Census of Housing 1940 was first to ask about amenities (plumbing, electricity, radio, refrigerator). Census 1930 asked about radio ownership — first census technology-diffusion question. Census 1950/1960 asked about TV, telephone, automobile, refrigerator, washing machine.
32.1 in 1929, up from 27.4 in 1920A
Rent, 1929 (index, 1982-84=100)observed record
CPI-U rent-of-primary-residence index (1982-84=100), annual average
Population measuredDepends on series (CPI: all urban consumers; food prices: urban consumers)
Source recordU.S. Bureau of Labor Statistics · 2024 · Tier A
Full record
BLS Public Data API v2
U.S. Bureau of Labor Statistics, 2024 · source
Confidence: A — Official series
Curator note: BLS series CUUR0000SEHA. Rent kept rising for five years after other prices crashed, peaking at 34.6 in 1925 — the wartime housing shortage worked its way through leases and construction with a long lag — before easing to 32.1 by 1929. Across the decade rent rose 17% while food fell 21%: shelter took a visibly larger share of the household budget in 1929 than in 1920, without any change in what the family occupied.
Source note: Direct API access to BLS time series. Used for: average food prices (APU series), CPI-U (CUUR0000SA0), wage and hours data. API key registered. Food price series (APU) available from ~1970s onward — no 1950 data. Rate limit: 100 queries/day for registered users.
Assumption: The composite family
U.S. Bureau of Labor Statistics, 2024 · source
Confidence: A — Official series
Curator note: BLS series CUUR0000SEHA. Rent kept rising for five years after other prices crashed, peaking at 34.6 in 1925 — the wartime housing shortage worked its way through leases and construction with a long lag — before easing to 32.1 by 1929. Across the decade rent rose 17% while food fell 21%: shelter took a visibly larger share of the household budget in 1929 than in 1920, without any change in what the family occupied.
Source note: Direct API access to BLS time series. Used for: average food prices (APU series), CPI-U (CUUR0000SA0), wage and hours data. API key registered. Food price series (APU) available from ~1970s onward — no 1950 data. Rate limit: 100 queries/day for registered users.
Assumption: The composite family
The budget
0 observations · 1 gapsno reliable record accessible onlineD
Median family income, 1920sdocumented gap
n/a
Population measuredAll US families (two or more related persons), CPS money income
Source recordU.S. Census Bureau · 2024 · Tier D
Full record
Historical Income Tables: Families
U.S. Census Bureau, 2024 · source
Confidence: D — Scholarly estimate
Curator note: No national income survey before 1947. Manufacturing wages provide a proxy: $0.55/hr, ~47.4 hrs/week in 1920 ($26.02/week, ~$1,353/yr). By 1929: wages were similar in nominal terms but real wages had risen due to deflation. The NBER and Simon Kuznets produced retrospective national income estimates for the 1920s, but these are aggregate, not family-level.
Source note: Landing page listing Tables F-1 through F-23. Continuously updated. Last revision Sep 2025 (2024 data).
U.S. Census Bureau, 2024 · source
Confidence: D — Scholarly estimate
Curator note: No national income survey before 1947. Manufacturing wages provide a proxy: $0.55/hr, ~47.4 hrs/week in 1920 ($26.02/week, ~$1,353/yr). By 1929: wages were similar in nominal terms but real wages had risen due to deflation. The NBER and Simon Kuznets produced retrospective national income estimates for the 1920s, but these are aggregate, not family-level.
Source note: Landing page listing Tables F-1 through F-23. Continuously updated. Last revision Sep 2025 (2024 data).
The table
2 observations · 2 gapsno reliable record accessible onlineD
The food basket, 1920sdocumented gap
n/a
Population measured2,567 wage-earner families in 33 states, head earning <=$1,200/year, year 1901
Source recordU.S. Department of Commerce and Labor, Bureau of Labor · 1903 · Tier D
Full record
BLS Bulletin No. 49: Cost of Living and Retail Prices of Food (18th Annual Report of the Commissioner of Labor)
U.S. Department of Commerce and Labor, Bureau of Labor, 1903 · source
Confidence: D — Scholarly estimate
Curator note: No consumer expenditure survey for the 1920s. The 1918-19 wartime study and the 1934-36 study bracket the decade. BLS retail food price data exists for the 1920s but not in a household basket format.
Source note: Primary source for 1900s room. Average income $827.19, average expenditure $768.54, food $326.90 (42.54%). Rent 12.95%, clothing 14.04%, fuel and light 5.25%. Average family size 5.31. NOTE: the University of Missouri guide cited average income as $749 — the primary source says $827.19. The $749 figure is WRONG; this is the correct figure from the report itself. File: bls_v08_0049_1903.pdf (268 pages, text-extractable via pdftotext).
U.S. Department of Commerce and Labor, Bureau of Labor, 1903 · source
Confidence: D — Scholarly estimate
Curator note: No consumer expenditure survey for the 1920s. The 1918-19 wartime study and the 1934-36 study bracket the decade. BLS retail food price data exists for the 1920s but not in a household basket format.
Source note: Primary source for 1900s room. Average income $827.19, average expenditure $768.54, food $326.90 (42.54%). Rent 12.95%, clothing 14.04%, fuel and light 5.25%. Average family size 5.31. NOTE: the University of Missouri guide cited average income as $749 — the primary source says $827.19. The $749 figure is WRONG; this is the correct figure from the report itself. File: bls_v08_0049_1903.pdf (268 pages, text-extractable via pdftotext).
665 cigarettes per adult per yearA
Adult per capita cigarette consumption, 1920observed record
cigarettes per adult (≥18 years) per year
Population measuredU.S. adults ≥18 years (NHIS) and total U.S. cigarette consumption (USDA)
Source recordCenters for Disease Control and Prevention (CDC) · 1994 · Tier A
Full record
Surveillance for Selected Tobacco-Use Behaviors -- United States, 1900-1994
Centers for Disease Control and Prevention (CDC), 1994 · source
Confidence: A — Official series
Curator note: From CDC MMWR Surveillance Summary Vol. 43 No. SS-3, Table 1. Cigarette smoking became mainstream in the 1920s. Consumption more than quadrupled from 151 in 1910. Total U.S. consumption was 44.6 billion cigarettes in 1920.
Source note: MMWR Surveillance Summary Vol. 43 No. SS-3. Table 1: per capita cigarette consumption 1900-1994 from USDA Economic Research Service (based on BATF tax data, imports, Census population). Table 2: adult smoking prevalence 1965-1991 from NHIS. Per capita = cigarettes per adult ≥18 per year.
Assumption: The composite family
Centers for Disease Control and Prevention (CDC), 1994 · source
Confidence: A — Official series
Curator note: From CDC MMWR Surveillance Summary Vol. 43 No. SS-3, Table 1. Cigarette smoking became mainstream in the 1920s. Consumption more than quadrupled from 151 in 1910. Total U.S. consumption was 44.6 billion cigarettes in 1920.
Source note: MMWR Surveillance Summary Vol. 43 No. SS-3. Table 1: per capita cigarette consumption 1900-1994 from USDA Economic Research Service (based on BATF tax data, imports, Census population). Table 2: adult smoking prevalence 1965-1991 from NHIS. Per capita = cigarettes per adult ≥18 per year.
Assumption: The composite family
no reliable record — national Prohibition, 1920-1933: legal sales ended and the series ended with themA
Alcohol consumed per person, 1920sdocumented gap
gallons of pure ethanol per person per year
Population measuredThe whole U.S. population aged 15 and older before 1970, and aged 14 and older from 1970 onward — not drinkers only, and not adults only
Source recordNational Institute on Alcohol Abuse and Alcoholism, Division of Epidemiology and Prevention Research (Alcohol Epidemiologic Data System) · 2025 · Tier A
Full record
Apparent Per Capita Alcohol Consumption: National, State, and Regional Trends, 1977-2023 (Surveillance Report #122)
National Institute on Alcohol Abuse and Alcoholism, Division of Epidemiology and Prevention Research (Alcohol Epidemiologic Data System), 2025 · source
Confidence: A — Official series
Curator note: National Prohibition ran from January 1920 to December 1933, and the source table prints the single word '(Prohibition)' where fourteen years of numbers would be. The series is built from legal sales, so when legal sales ended the measurement ended with them. Drinking did not stop; the state's ability to count it did. The gap is not an archive limitation and is not interpolated: no consumption estimate for these years appears in this table, and any that were entered would be a different kind of evidence.
Source note: Table 1, 'Apparent per capita ethanol consumption, United States, 1850-2023', all-beverages column. Apparent consumption is derived from legal beverage sales and shipment data collected from States, the National Alcohol Beverage Control Association and industry sources, divided by Census population: it measures what the state could count, not what was drunk, which is why national Prohibition appears in the table as the literal word '(Prohibition)' and no numbers for 1920-1933. The report's own note records that data prior to 1977 are from Hyman et al. 1980, so the early half of this official series is republished scholarly work. Pre-1934 figures are published as multi-year ranges, not single years. The archived copy in samples/ was confirmed sha256-identical to the document served at this URL on 2026-09-01 (see docs/verification-log.md).
Assumption: The composite family
National Institute on Alcohol Abuse and Alcoholism, Division of Epidemiology and Prevention Research (Alcohol Epidemiologic Data System), 2025 · source
Confidence: A — Official series
Curator note: National Prohibition ran from January 1920 to December 1933, and the source table prints the single word '(Prohibition)' where fourteen years of numbers would be. The series is built from legal sales, so when legal sales ended the measurement ended with them. Drinking did not stop; the state's ability to count it did. The gap is not an archive limitation and is not interpolated: no consumption estimate for these years appears in this table, and any that were entered would be a different kind of evidence.
Source note: Table 1, 'Apparent per capita ethanol consumption, United States, 1850-2023', all-beverages column. Apparent consumption is derived from legal beverage sales and shipment data collected from States, the National Alcohol Beverage Control Association and industry sources, divided by Census population: it measures what the state could count, not what was drunk, which is why national Prohibition appears in the table as the literal word '(Prohibition)' and no numbers for 1920-1933. The report's own note records that data prior to 1977 are from Hyman et al. 1980, so the early half of this official series is republished scholarly work. Pre-1934 figures are published as multi-year ranges, not single years. The archived copy in samples/ was confirmed sha256-identical to the document served at this URL on 2026-09-01 (see docs/verification-log.md).
Assumption: The composite family
16.5 in 1929, down from 21.0 in 1920A
Food prices, 1929 (index, 1982-84=100)observed record
CPI-U food index (1982-84=100), annual average
Population measuredDepends on series (CPI: all urban consumers; food prices: urban consumers)
Source recordU.S. Bureau of Labor Statistics · 2024 · Tier A
Full record
BLS Public Data API v2
U.S. Bureau of Labor Statistics, 2024 · source
Confidence: A — Official series
Curator note: BLS series CUUR0000SAF1. The postwar collapse was abrupt: food fell from 21.0 in 1920 to 15.9 in 1921, a 24% drop in a single year, and never regained its wartime peak — 16.5 by 1929. For a family, the 1920s were a decade in which eating got cheaper while the rent did not (rent rose 17% over the same span). The often-told story of 1920s prosperity is usually an income story; on the price side the decade was deflationary.
Source note: Direct API access to BLS time series. Used for: average food prices (APU series), CPI-U (CUUR0000SA0), wage and hours data. API key registered. Food price series (APU) available from ~1970s onward — no 1950 data. Rate limit: 100 queries/day for registered users.
Assumption: The composite family
U.S. Bureau of Labor Statistics, 2024 · source
Confidence: A — Official series
Curator note: BLS series CUUR0000SAF1. The postwar collapse was abrupt: food fell from 21.0 in 1920 to 15.9 in 1921, a 24% drop in a single year, and never regained its wartime peak — 16.5 by 1929. For a family, the 1920s were a decade in which eating got cheaper while the rent did not (rent rose 17% over the same span). The often-told story of 1920s prosperity is usually an income story; on the price side the decade was deflationary.
Source note: Direct API access to BLS time series. Used for: average food prices (APU series), CPI-U (CUUR0000SA0), wage and hours data. API key registered. Food price series (APU) available from ~1970s onward — no 1950 data. Rate limit: 100 queries/day for registered users.
Assumption: The composite family
The day
9 observations · 0 gaps56.3 male, 58.5 female (white)A
Life expectancy at birth, 1919-21observed record
years
Population measuredVarious national aggregates (depends on table)
Source recordU.S. Census Bureau · 1975 · Tier A
Full record
Historical Statistics of the United States, Colonial Times to 1970 (Bicentennial Edition)
U.S. Census Bureau, 1975 · source
Confidence: A — Official series
Curator note: From Historical Statistics Vol 1, Series B 116-117 (white population, at birth). Up from 50.2/53.6 in 1909-11 — a gain of ~6 years in a decade. The 1920s would see further gains as public health measures (clean water, sanitation) expanded. By 1929-31: 59.1 male, 62.7 female.
Source note: Free on Internet Archive. Contains population, labor, prices, housing, and diffusion series spanning colonial times to 1970. Essential for pre-1940 decades where no dedicated survey exists. Predecessor to the Millennial Edition.
U.S. Census Bureau, 1975 · source
Confidence: A — Official series
Curator note: From Historical Statistics Vol 1, Series B 116-117 (white population, at birth). Up from 50.2/53.6 in 1909-11 — a gain of ~6 years in a decade. The 1920s would see further gains as public health measures (clean water, sanitation) expanded. By 1929-31: 59.1 male, 62.7 female.
Source note: Free on Internet Archive. Contains population, labor, prices, housing, and diffusion series spanning colonial times to 1970. Essential for pre-1940 decades where no dedicated survey exists. Predecessor to the Millennial Edition.
$0.55/hr, 47.4 hrs/week, $26.02/wkA
Manufacturing production worker wages, 1920observed record
USD, nominal (hourly earnings, weekly hours, weekly earnings)
Population measuredProduction workers in manufacturing (national average, payroll data)
Source recordU.S. Census Bureau · 1975 · Tier A
Full record
Historical Statistics: Average Hourly Earnings of Manufacturing Production Workers (Series D 766, D 765-767, D 802-804)
U.S. Census Bureau, 1975 · source
Confidence: A — Official series
Basis: Hourly rate
Curator note: From Historical Statistics Vol 1, Series D 802-804 (all manufacturing). Hourly: $0.55 (55 cents, exact), weekly hours: 47.4, weekly earnings: $26.02. By 1925: $0.54/hr, 44.5 hrs/week, $24.11/wk. By 1929: $0.56/hr, 44.2 hrs/week, $24.76/wk. The 1920s saw flat nominal wages but rising real wages (deflation) and shorter hours.
Source note: Series D 766 (hourly earnings 1890-1914), D 765-767 (manufacturing production workers 1909-1929), D 802-804 (all manufacturing 1919-1948). Pre-FRED CES wage data. Used as the wage anchor for 1900s-1930s rooms (Plan 013). The population skews urban, industrial, male — see the manufacturing-wage-proxy assumption.
Assumption: Values are shown in period money
Assumption: Manufacturing wages as the early-era worker proxy
U.S. Census Bureau, 1975 · source
Confidence: A — Official series
Basis: Hourly rate
Curator note: From Historical Statistics Vol 1, Series D 802-804 (all manufacturing). Hourly: $0.55 (55 cents, exact), weekly hours: 47.4, weekly earnings: $26.02. By 1925: $0.54/hr, 44.5 hrs/week, $24.11/wk. By 1929: $0.56/hr, 44.2 hrs/week, $24.76/wk. The 1920s saw flat nominal wages but rising real wages (deflation) and shorter hours.
Source note: Series D 766 (hourly earnings 1890-1914), D 765-767 (manufacturing production workers 1909-1929), D 802-804 (all manufacturing 1919-1948). Pre-FRED CES wage data. Used as the wage anchor for 1900s-1930s rooms (Plan 013). The population skews urban, industrial, male — see the manufacturing-wage-proxy assumption.
Assumption: Values are shown in period money
Assumption: Manufacturing wages as the early-era worker proxy
25.54 deaths per 100 million vehicle-miles travelledA
Road deaths per 100 million vehicle-miles, 1920observed record
deaths per 100 million vehicle-miles travelled
Population measuredAll persons killed in motor-vehicle traffic crashes on US public roads (the States and the District of Columbia), measured against total vehicle-miles travelled — not drivers only and not vehicle occupants only
Source recordFederal Highway Administration, Office of Highway Policy Information · 2025 · Tier A
Full record
Table FI-200: Motor Vehicle Traffic Fatalities, 1900-2023 (Highway Statistics 2023)
Federal Highway Administration, Office of Highway Policy Information, 2025 · source
Confidence: A — Official series
Curator note: Twelve thousand people a year were being killed by then. The rate was already falling, because travel was growing faster than the toll.
Source note: The whole 1900-2023 span in one FHWA table, so the fatality-rate series needs no splice between estimate and registration eras. Four rate columns are published side by side (per 1,000 miles of road, per 100 million annual VMT, per 100,000 registered motor vehicles, per 100,000 licensed drivers); the museum takes the per-100-million-VMT column, and the extraction script recomputes it from the table's own fatality and VMT columns for every year rather than trusting its position. FHWA footnote (3) records a definitional change: beginning in 1976 the count includes only people who died within 30 days of the crash, so the earlier years are counted on a wider window and the two halves are not the same measurement. The 1900-1995 rows were cross-checked against FHWA's separately-published Highway Statistics Summary to 1995 (April 1997): 95 of 96 years reproduce exactly, the exception being 1995, that publication's terminal provisional year, since revised from 41,770 deaths to 41,817 (see docs/verification-log.md).
Assumption: The composite family
Federal Highway Administration, Office of Highway Policy Information, 2025 · source
Confidence: A — Official series
Curator note: Twelve thousand people a year were being killed by then. The rate was already falling, because travel was growing faster than the toll.
Source note: The whole 1900-2023 span in one FHWA table, so the fatality-rate series needs no splice between estimate and registration eras. Four rate columns are published side by side (per 1,000 miles of road, per 100 million annual VMT, per 100,000 registered motor vehicles, per 100,000 licensed drivers); the museum takes the per-100-million-VMT column, and the extraction script recomputes it from the table's own fatality and VMT columns for every year rather than trusting its position. FHWA footnote (3) records a definitional change: beginning in 1976 the count includes only people who died within 30 days of the crash, so the earlier years are counted on a wider window and the two halves are not the same measurement. The 1900-1995 rows were cross-checked against FHWA's separately-published Highway Statistics Summary to 1995 (April 1997): 95 of 96 years reproduce exactly, the exception being 1995, that publication's terminal provisional year, since revised from 41,770 deaths to 41,817 (see docs/verification-log.md).
Assumption: The composite family
44.5C
Women's weekly unpaid home-production hours, 1920observed record
hours per week, prime-age women (ages 18–64)
Population measuredPrime-age women and men, ages 18–64 (benchmark years 1900–2005); also all-ages per-capita and per-household aggregates
Source recordCambridge University Press (Journal of Economic History) · 2009 · Tier C
Full record
Time Spent in Home Production in the Twentieth-Century United States: New Estimates from Old Data
Cambridge University Press (Journal of Economic History), 2009 · source
Confidence: C — Reconstructed from period surveys
Curator note: Ramey (2009) Table 6A, 'All Prime-Age Women' column. Reconstruction from historical time-diary studies (Purnell Act 1920s, Wilson 1929, USDA 1944) linked to AHTUS/BLS surveys from 1965. Includes food prep, house cleaning, clothing care, childcare, purchasing, and household management. Numbers in italics in source = partially extrapolated. Draft version (NBER w13985) has identical values. Splice caveat: Ramey measures women-specific home production ages 18-64; ATUS (2020s room) measures all-adult household activities age 15+ — a concept change that must caveat when plotted together.
Source note: Valerie A. Ramey, JEconHist 69(1), March 2009, pp. 1–47. Reconstruction from historical time-diary studies (Purnell Act studies 1920s, Wilson 1929, USDA 1944) linked to AHTUS/BLS modern surveys. Draft version: NBER Working Paper w13985 (May 2008, 63pp) — data tables are numerically identical to the published version; differences are prose tightening and table reformatting (5→5A/5B, etc.). Tier C (period-survey reconstruction). Key tables: Table 5A (nonemployed women), Table 6A (all women, prime-age), Table 7 (men, prime-age), Table 8A (all ages), Table 3 (component breakdown). Note: Ramey does not use the 1992-94 survey (missing data), so no 1990s benchmark. Splice point to ATUS: Ramey measures women-specific home production ages 18-64; ATUS measures all-adult household activities age 15+ — a concept splice that must caveat (Plan 004 Measure guard).
Assumption: The composite family
Cambridge University Press (Journal of Economic History), 2009 · source
Confidence: C — Reconstructed from period surveys
Curator note: Ramey (2009) Table 6A, 'All Prime-Age Women' column. Reconstruction from historical time-diary studies (Purnell Act 1920s, Wilson 1929, USDA 1944) linked to AHTUS/BLS surveys from 1965. Includes food prep, house cleaning, clothing care, childcare, purchasing, and household management. Numbers in italics in source = partially extrapolated. Draft version (NBER w13985) has identical values. Splice caveat: Ramey measures women-specific home production ages 18-64; ATUS (2020s room) measures all-adult household activities age 15+ — a concept change that must caveat when plotted together.
Source note: Valerie A. Ramey, JEconHist 69(1), March 2009, pp. 1–47. Reconstruction from historical time-diary studies (Purnell Act studies 1920s, Wilson 1929, USDA 1944) linked to AHTUS/BLS modern surveys. Draft version: NBER Working Paper w13985 (May 2008, 63pp) — data tables are numerically identical to the published version; differences are prose tightening and table reformatting (5→5A/5B, etc.). Tier C (period-survey reconstruction). Key tables: Table 5A (nonemployed women), Table 6A (all women, prime-age), Table 7 (men, prime-age), Table 8A (all ages), Table 3 (component breakdown). Note: Ramey does not use the 1992-94 survey (missing data), so no 1990s benchmark. Splice point to ATUS: Ramey measures women-specific home production ages 18-64; ATUS measures all-adult household activities age 15+ — a concept splice that must caveat (Plan 004 Measure guard).
Assumption: The composite family
3.9C
Men's weekly unpaid home-production hours, 1920observed record
hours per week, prime-age men (ages 18–64)
Population measuredPrime-age women and men, ages 18–64 (benchmark years 1900–2005); also all-ages per-capita and per-household aggregates
Source recordCambridge University Press (Journal of Economic History) · 2009 · Tier C
Full record
Time Spent in Home Production in the Twentieth-Century United States: New Estimates from Old Data
Cambridge University Press (Journal of Economic History), 2009 · source
Confidence: C — Reconstructed from period surveys
Curator note: Ramey (2009) Table 7, 'All Prime-Age Men' column. Early estimates from Purnell Act studies and Lundberg et al. (1934); 1965+ from AHTUS/BLS. Men's hours rose 13 hrs/week across the century (3.9→16.8), partially offsetting women's decline. Numbers in italics in source = partially extrapolated.
Source note: Valerie A. Ramey, JEconHist 69(1), March 2009, pp. 1–47. Reconstruction from historical time-diary studies (Purnell Act studies 1920s, Wilson 1929, USDA 1944) linked to AHTUS/BLS modern surveys. Draft version: NBER Working Paper w13985 (May 2008, 63pp) — data tables are numerically identical to the published version; differences are prose tightening and table reformatting (5→5A/5B, etc.). Tier C (period-survey reconstruction). Key tables: Table 5A (nonemployed women), Table 6A (all women, prime-age), Table 7 (men, prime-age), Table 8A (all ages), Table 3 (component breakdown). Note: Ramey does not use the 1992-94 survey (missing data), so no 1990s benchmark. Splice point to ATUS: Ramey measures women-specific home production ages 18-64; ATUS measures all-adult household activities age 15+ — a concept splice that must caveat (Plan 004 Measure guard).
Assumption: The composite family
Cambridge University Press (Journal of Economic History), 2009 · source
Confidence: C — Reconstructed from period surveys
Curator note: Ramey (2009) Table 7, 'All Prime-Age Men' column. Early estimates from Purnell Act studies and Lundberg et al. (1934); 1965+ from AHTUS/BLS. Men's hours rose 13 hrs/week across the century (3.9→16.8), partially offsetting women's decline. Numbers in italics in source = partially extrapolated.
Source note: Valerie A. Ramey, JEconHist 69(1), March 2009, pp. 1–47. Reconstruction from historical time-diary studies (Purnell Act studies 1920s, Wilson 1929, USDA 1944) linked to AHTUS/BLS modern surveys. Draft version: NBER Working Paper w13985 (May 2008, 63pp) — data tables are numerically identical to the published version; differences are prose tightening and table reformatting (5→5A/5B, etc.). Tier C (period-survey reconstruction). Key tables: Table 5A (nonemployed women), Table 6A (all women, prime-age), Table 7 (men, prime-age), Table 8A (all ages), Table 3 (component breakdown). Note: Ramey does not use the 1992-94 survey (missing data), so no 1990s benchmark. Splice point to ATUS: Ramey measures women-specific home production ages 18-64; ATUS measures all-adult household activities age 15+ — a concept splice that must caveat (Plan 004 Measure guard).
Assumption: The composite family
Food prep 19.9, house cleaning 9.3, clothing care 11.5, childcare 7.2, purchasing/mgmt/travel 4.4 hrs/weekC
How a 1920s housewife's week broke down (Wilson study)observed record
hours per week by activity category, nonemployed town housewives
Population measuredPrime-age women and men, ages 18–64 (benchmark years 1900–2005); also all-ages per-capita and per-household aggregates
Source recordCambridge University Press (Journal of Economic History) · 2009 · Tier C
Full record
Time Spent in Home Production in the Twentieth-Century United States: New Estimates from Old Data
Cambridge University Press (Journal of Economic History), 2009 · source
Confidence: C — Reconstructed from period surveys
Curator note: Ramey (2009) Table 3, '1920s Town Housewives, Wilson Study' column. Total home production: 52.4 hrs/week. Wilson's time-diary study (conducted ~1920s, published 193-). For comparison, 1965 housewives (AHTUS) spent less on food prep (16.5) and clothing care (6.9) but more on purchasing/management/travel (10.4) — market substitution shifted time between categories, not eliminated it. Farm housewives (USDA 1944) spent even more on food prep (23.5 hrs) due to preservation and from-scratch cooking.
Source note: Valerie A. Ramey, JEconHist 69(1), March 2009, pp. 1–47. Reconstruction from historical time-diary studies (Purnell Act studies 1920s, Wilson 1929, USDA 1944) linked to AHTUS/BLS modern surveys. Draft version: NBER Working Paper w13985 (May 2008, 63pp) — data tables are numerically identical to the published version; differences are prose tightening and table reformatting (5→5A/5B, etc.). Tier C (period-survey reconstruction). Key tables: Table 5A (nonemployed women), Table 6A (all women, prime-age), Table 7 (men, prime-age), Table 8A (all ages), Table 3 (component breakdown). Note: Ramey does not use the 1992-94 survey (missing data), so no 1990s benchmark. Splice point to ATUS: Ramey measures women-specific home production ages 18-64; ATUS measures all-adult household activities age 15+ — a concept splice that must caveat (Plan 004 Measure guard).
Assumption: The composite family
Cambridge University Press (Journal of Economic History), 2009 · source
Confidence: C — Reconstructed from period surveys
Curator note: Ramey (2009) Table 3, '1920s Town Housewives, Wilson Study' column. Total home production: 52.4 hrs/week. Wilson's time-diary study (conducted ~1920s, published 193-). For comparison, 1965 housewives (AHTUS) spent less on food prep (16.5) and clothing care (6.9) but more on purchasing/management/travel (10.4) — market substitution shifted time between categories, not eliminated it. Farm housewives (USDA 1944) spent even more on food prep (23.5 hrs) due to preservation and from-scratch cooking.
Source note: Valerie A. Ramey, JEconHist 69(1), March 2009, pp. 1–47. Reconstruction from historical time-diary studies (Purnell Act studies 1920s, Wilson 1929, USDA 1944) linked to AHTUS/BLS modern surveys. Draft version: NBER Working Paper w13985 (May 2008, 63pp) — data tables are numerically identical to the published version; differences are prose tightening and table reformatting (5→5A/5B, etc.). Tier C (period-survey reconstruction). Key tables: Table 5A (nonemployed women), Table 6A (all women, prime-age), Table 7 (men, prime-age), Table 8A (all ages), Table 3 (component breakdown). Note: Ramey does not use the 1992-94 survey (missing data), so no 1990s benchmark. Splice point to ATUS: Ramey measures women-specific home production ages 18-64; ATUS measures all-adult household activities age 15+ — a concept splice that must caveat (Plan 004 Measure guard).
Assumption: The composite family
~9.0%A
Married women's labor force participation, 1920observed record
% of married women (husband present) in the civilian labor force
Population measuredVarious national aggregates (depends on table)
Source recordU.S. Census Bureau · 1975 · Tier A
Full record
Historical Statistics of the United States, Colonial Times to 1970 (Bicentennial Edition)
U.S. Census Bureau, 1975 · source
Confidence: A — Official series
Curator note: From Historical Statistics Vol 1, Series D 49-62. Down from ~10.7% in 1910 (the 1910 figure had enumerator-instruction issues; this 1920 figure is more reliable). Total female LFPR was 23.7%. The 'flapper' era saw more women working, but married women with husbands present remained mostly at home.
Source note: Free on Internet Archive. Contains population, labor, prices, housing, and diffusion series spanning colonial times to 1970. Essential for pre-1940 decades where no dedicated survey exists. Predecessor to the Millennial Edition.
Assumption: The composite family
U.S. Census Bureau, 1975 · source
Confidence: A — Official series
Curator note: From Historical Statistics Vol 1, Series D 49-62. Down from ~10.7% in 1910 (the 1910 figure had enumerator-instruction issues; this 1920 figure is more reliable). Total female LFPR was 23.7%. The 'flapper' era saw more women working, but married women with husbands present remained mostly at home.
Source note: Free on Internet Archive. Contains population, labor, prices, housing, and diffusion series spanning colonial times to 1970. Essential for pre-1940 decades where no dedicated survey exists. Predecessor to the Millennial Edition.
Assumption: The composite family
Men: 24.6, Women: 21.2A
Median age at first marriage, 1920observed record
years (median age at first marriage)
Population measuredU.S. civilian noninstitutionalized population (CPS/decennial census)
Source recordU.S. Census Bureau · 2023 · Tier A
Full record
Census Historical Table MS-2: Median Age at First Marriage
U.S. Census Bureau, 2023 · source
Confidence: A — Official series
Curator note: From Census Table MS-2. Decennial census. Both men and women marrying slightly younger than in 1910. The 1920s prosperity made earlier family formation possible.
Source note: Table MS-2: Median age at first marriage by sex, 1890 to present. Decennial census data 1890-2000; CPS data for intervening years and 2010+. File: ms2-median-age-first-marriage.xls. The median age at first marriage declined from ~26 (men) / ~22 (women) in 1890 to a mid-century low of ~22.5 / 20.3 in 1956, then rose steadily to ~30+ / ~28+ by 2023.
U.S. Census Bureau, 2023 · source
Confidence: A — Official series
Curator note: From Census Table MS-2. Decennial census. Both men and women marrying slightly younger than in 1910. The 1920s prosperity made earlier family formation possible.
Source note: Table MS-2: Median age at first marriage by sex, 1890 to present. Decennial census data 1890-2000; CPS data for intervening years and 2010+. File: ms2-median-age-first-marriage.xls. The median age at first marriage declined from ~26 (men) / ~22 (women) in 1890 to a mid-century low of ~22.5 / 20.3 in 1956, then rose steadily to ~30+ / ~28+ by 2023.
$406 billionD
Nonfarm household net worth, 1929observed record
aggregate net worth of nonfarm household sector, current dollars
Population measuredU.S. nonfarm household sector (aggregate, Goldsmith reconstruction)
Source recordNational Bureau of Economic Research / Princeton University Press · 1963 · Tier D
Full record
Studies in the National Balance Sheet of the United States, Vol. I
National Bureau of Economic Research / Princeton University Press, 1963 · source
Confidence: D — Scholarly estimate
Curator note: From Goldsmith, Lipsey & Mendelson (1963), Table 1, Sector I.1 (Nonfarm households). Total assets: $448B. Tangible assets: $158B. Financial assets: $291B. Debt: $42B. Net worth: $406B. The 1920s boom inflated financial assets to 65% of total — stocks and securities surged. After the 1929 crash, net worth would fall to $341B by 1939 (−16%). Aggregate — NOT median.
Source note: Table 1: Main Balance Sheet Components by Sector, 1900-1958. Benchmark years: 1900, 1912, 1922, 1929, 1939, 1945, 1958. Sector I.1 = Nonfarm households. Provides total assets, tangible assets, financial assets, debt, and net worth in current dollars. Goldsmith's balance sheet construction uses Census wealth data, Federal Reserve flow of funds, and tax return compilations. Aggregate household sector, NOT median. Tier D (scholarly reconstruction). Downloaded as NBER chapter c3926.
National Bureau of Economic Research / Princeton University Press, 1963 · source
Confidence: D — Scholarly estimate
Curator note: From Goldsmith, Lipsey & Mendelson (1963), Table 1, Sector I.1 (Nonfarm households). Total assets: $448B. Tangible assets: $158B. Financial assets: $291B. Debt: $42B. Net worth: $406B. The 1920s boom inflated financial assets to 65% of total — stocks and securities surged. After the 1929 crash, net worth would fall to $341B by 1939 (−16%). Aggregate — NOT median.
Source note: Table 1: Main Balance Sheet Components by Sector, 1900-1958. Benchmark years: 1900, 1912, 1922, 1929, 1939, 1945, 1958. Sector I.1 = Nonfarm households. Provides total assets, tangible assets, financial assets, debt, and net worth in current dollars. Goldsmith's balance sheet construction uses Census wealth data, Federal Reserve flow of funds, and tax return compilations. Aggregate household sector, NOT median. Tier D (scholarly reconstruction). Downloaded as NBER chapter c3926.
What had arrived
2 observations · 0 gaps~33 per 1,000 population (1925)A
Telephone diffusion, 1920sobserved record
telephones per 1,000 population
Population measuredVarious national aggregates (depends on table)
Source recordU.S. Census Bureau · 1975 · Tier A
Full record
Historical Statistics of the United States, Colonial Times to 1970 (Bicentennial Edition)
U.S. Census Bureau, 1975 · source
Confidence: A — Official series
Curator note: From Historical Statistics Vol 2, Series R 1-2. In 1925: ~33 per 1,000 population (~14.4 million total phones). Household percentage not directly reported until 1944. The 1920s saw rapid telephone expansion — the Bell System nearly tripled its subscriber base.
Source note: Free on Internet Archive. Contains population, labor, prices, housing, and diffusion series spanning colonial times to 1970. Essential for pre-1940 decades where no dedicated survey exists. Predecessor to the Millennial Edition.
U.S. Census Bureau, 1975 · source
Confidence: A — Official series
Curator note: From Historical Statistics Vol 2, Series R 1-2. In 1925: ~33 per 1,000 population (~14.4 million total phones). Household percentage not directly reported until 1944. The 1920s saw rapid telephone expansion — the Bell System nearly tripled its subscriber base.
Source note: Free on Internet Archive. Contains population, labor, prices, housing, and diffusion series spanning colonial times to 1970. Essential for pre-1940 decades where no dedicated survey exists. Predecessor to the Millennial Edition.
Radio: commercial broadcasting began 1920, ~40% of households by 1930. Automobile: ~26M registrations by 1930.A
Radio and automobile, 1920sobserved record
n/a
Population measuredVarious national aggregates (depends on table)
Source recordU.S. Census Bureau · 1975 · Tier A
Full record
Historical Statistics of the United States, Colonial Times to 1970 (Bicentennial Edition)
U.S. Census Bureau, 1975 · source
Confidence: A — Official series
Curator note: From Historical Statistics Vol 2. Commercial radio began with KDKA (Pittsburgh) in November 1920. By 1930, ~40% of households had a radio — one of the fastest technology diffusions in history. Automobile registrations: 1910: ~458K, 1920: ~8.1M, 1930: ~26.7M. The Model T (1908-1927) made car ownership accessible to the middle class.
Source note: Free on Internet Archive. Contains population, labor, prices, housing, and diffusion series spanning colonial times to 1970. Essential for pre-1940 decades where no dedicated survey exists. Predecessor to the Millennial Edition.
U.S. Census Bureau, 1975 · source
Confidence: A — Official series
Curator note: From Historical Statistics Vol 2. Commercial radio began with KDKA (Pittsburgh) in November 1920. By 1930, ~40% of households had a radio — one of the fastest technology diffusions in history. Automobile registrations: 1910: ~458K, 1920: ~8.1M, 1930: ~26.7M. The Model T (1908-1927) made car ownership accessible to the middle class.
Source note: Free on Internet Archive. Contains population, labor, prices, housing, and diffusion series spanning colonial times to 1970. Essential for pre-1940 decades where no dedicated survey exists. Predecessor to the Millennial Edition.
A day's work buys
1 observations · 1 gapsno reliable record accessible onlineD
Share-of-family-income comparison, 1920sdocumented gap
n/a
Population measuredVarious national aggregates (depends on table)
Source recordU.S. Census Bureau · 1975 · Tier D
Full record
Historical Statistics of the United States, Colonial Times to 1970 (Bicentennial Edition)
U.S. Census Bureau, 1975 · source
Confidence: D — Scholarly estimate
Curator note: No national median family income figure exists for this decade, so a share-of-income comparison cannot be computed. The hourly manufacturing-wage anchor does support hours-to-afford calculations for the priced records that follow.
Source note: Free on Internet Archive. Contains population, labor, prices, housing, and diffusion series spanning colonial times to 1970. Essential for pre-1940 decades where no dedicated survey exists. Predecessor to the Millennial Edition.
U.S. Census Bureau, 1975 · source
Confidence: D — Scholarly estimate
Curator note: No national median family income figure exists for this decade, so a share-of-income comparison cannot be computed. The hourly manufacturing-wage anchor does support hours-to-afford calculations for the priced records that follow.
Source note: Free on Internet Archive. Contains population, labor, prices, housing, and diffusion series spanning colonial times to 1970. Essential for pre-1940 decades where no dedicated survey exists. Predecessor to the Millennial Edition.
~$949 (average wholesale, computed from factory sales)A
Average new car price, 1920observed record
USD, nominal (wholesale value per passenger car)
Population measuredVarious national aggregates (depends on table)
Source recordU.S. Census Bureau · 1975 · Tier A
Full record
Historical Statistics of the United States, Colonial Times to 1970 (Bicentennial Edition)
U.S. Census Bureau, 1975 · source
Confidence: A — Official series
Basis: One-time price
Affordability anchors: Production workers in manufacturing (national average, payroll data)
Denominators measure: hours axis: average hourly earnings
Curator note: Computed from Historical Statistics, Series Q 148-162. 1920: 1,905,500 passenger cars sold at $1,809 million wholesale value = ~$949 average per car. The Model T (1908-1927) drove prices down — by 1920 a new Model T cost ~$395 (retail), but the average wholesale across all cars was higher due to luxury models. Wholesale value; retail/transaction price would be higher. Population: all US factory sales of passenger cars.
Source note: Free on Internet Archive. Contains population, labor, prices, housing, and diffusion series spanning colonial times to 1970. Essential for pre-1940 decades where no dedicated survey exists. Predecessor to the Millennial Edition.
Assumption: Values are shown in period money
U.S. Census Bureau, 1975 · source
Confidence: A — Official series
Basis: One-time price
Affordability anchors: Production workers in manufacturing (national average, payroll data)
Denominators measure: hours axis: average hourly earnings
Curator note: Computed from Historical Statistics, Series Q 148-162. 1920: 1,905,500 passenger cars sold at $1,809 million wholesale value = ~$949 average per car. The Model T (1908-1927) drove prices down — by 1920 a new Model T cost ~$395 (retail), but the average wholesale across all cars was higher due to luxury models. Wholesale value; retail/transaction price would be higher. Population: all US factory sales of passenger cars.
Source note: Free on Internet Archive. Contains population, labor, prices, housing, and diffusion series spanning colonial times to 1970. Essential for pre-1940 decades where no dedicated survey exists. Predecessor to the Millennial Edition.
Assumption: Values are shown in period money
Computed affordability
≈ 1,725 hours of work
Confidence & flags
A — official statistical series
B — official microdata, computed by this project
C — reconstructed from period surveys
D — scholarly estimate
Gap — no reliable record; shown as the gap it is
Reading the atlas
Provenance is in the scan-line: every exhibit's row names its value, confidence tier, the population actually measured, and the source record. Chart points and stage glyphs open the same record.
Rising series draw in petrol, falling in ember. An omitted illustration means no exhibit was selected, not that the object was historically absent.