How to Read the Fed's Survey of Consumer Finances (SCF) — Income, Net Worth, Debt and Median vs Mean

As of 3 November 2026, 00:22 KST; most recently published report: Changes in U.S. Family Finances from 2022 to 2025, the Federal Reserve Board's report on its triennial Survey of Consumer Finances (SCF), dated October 9, 2026 on the survey page checked at 00:22 KST. It compares the 2025 survey (4,367 families interviewed) with the 2022 survey (4,602 families), in 2025 dollars. The survey page gives no date for the next survey report; the source gives no time of day for the release. Primary source: https://www.federalreserve.gov/publications/files/scf26.pdf^[1][2]

Comparison basis: values are survey estimates as published, in thousands of 2025 dollars or percent of families; the report adjusts dollars with a current-methods consumer price index. Published changes are whole percentages; this page's calculated changes are rounded half up to one decimal, and point changes are differences between published percentages. Income refers to the calendar year before each survey (2021 and 2024), while assets and debts are measured at the interview. The survey is triennial, so there is no year-ago comparison. The report and survey page are both Federal Reserve Board publications and count as one source.^[1]

In three lines

1. The SCF is a triennial Federal Reserve survey of U.S. families' income, assets, debts and credit experiences, designed to capture very wealthy families as well.
2. Between the 2022 and 2025 surveys, real median income rose 7% while mean income fell 6%; median net worth rose 2% and mean net worth 7%.
3. The share of families late on loan payments rose from 12.2% to 19.6%, and the median payment-to-income ratio for debtors rose to 15.4%.^[1]

"Real median family income, which is measured for the year before the survey, rose 7 percent between the 2022 to 2025 surveys, while real mean family income fell 6 percent." "Between 2022 and 2025, real median net worth (wealth) rose by 2 percent, while real mean net worth rose more, by 7 percent." "Participation in the stock market declined slightly from 58 percent in 2022 to 56 percent in 2025." "Conditional on holding stock, median stock holdings grew 36 percent from $56,900 in 2022 to $77,400 in 2025." "Despite recent income gains, the median home was worth more than 4.5 times median family income in 2025, about unchanged from the previous high in 2022."^[1]

Who and when. "In the 2025 survey, 4,367 families were interviewed, and in the 2022 survey, 4,602 families were interviewed." "In the SCF, a household unit is divided into a primary economic unit (PEU)—the family—and everyone else in the household." "The survey collects information on families' income before taxes for the calendar year preceding the survey year." "For the 2025 SCF, about one-seventh of the interviews were conducted between January and May 2026."^[1]

Median vs mean. The report explains: "Generally, when data include very large values for a relatively small part of the population—as is the case for income, net worth, and many of the other items considered in this report—estimates of the median are often statistically less sensitive to such very large values than are estimates of the mean." The mean is more sensitive than the median to very large values; both can change when the composition of families changes. The report adds: "In fact, excluding families in the top 1 percent of income indicates that, between the 2022 and 2025 surveys, mean income declined only 1 percent." So the 7% rise in median income and the 6% fall in mean income describe different parts of the distribution.^[1]

Aggregate vs median debt ratios (opposite signs). The aggregate debt-to-income ratio rose from 89.4% to 94.9% (+5.5 points), while the median for debtors edged down from 95.1% to 94.8% (-0.3 points; this page's calculations). "The aggregate is defined as the total amount of debt held (or payments) divided by the total assets held (or income) among all families." The median for debtors is the median of each indebted family's own ratio, so the two measures can move in opposite directions. The median leverage ratio for debtors fell to 26.9% while the median payment-to-income ratio rose to 15.4%.^[1]

Executive summary vs body text (late payments and high burdens). The executive summary says: "Families were more likely to be behind on their financial obligations than at any point since the 2010 survey." The body says: "Families were more likely to be behind on such obligations than in any of the past four surveys, or since 2016." For high payment burdens, the summary says: "The fraction of families with debt payment-to-income ratios greater than 40 percent, which are families with particularly high debt payment obligations relative to their incomes, increased from 6.5 to 8.6 percent, a level last seen in the 2013 survey." The body says: "In 2025, 8.6 percent of families had payment-to-income ratios greater than 40 percent, the largest share since 2010." The passages use different comparison ranges ('since 2010' versus 'since 2016'; 'a level last seen in 2013' versus 'the largest share since 2010') and are not mutually exclusive: Table 5, which starts in 2013, shows 14.9% of families late on payments and 8.2% above the 40 percent threshold in 2013, so the 2025 values of 19.6% and 8.6% are above every survey in that table.^[1]

Direct stocks vs any stock holdings. Table 3 shows directly held stocks owned by 19.0% of families in 2025 (21.0% in 2022), with a conditional median of $30,000. The executive summary's 56% stock market participation and $77,400 median holding include stock held indirectly, such as through retirement accounts and pooled funds, so the two sets of figures measure different things.^[1]

Standard errors vs significance. Table 1 prints standard errors below the all-family rows: 1.1 thousand dollars for 2025 median income and 2.4 thousand for mean income; Table 2 prints 6.4 and 31.0 thousand for median and mean net worth. These describe sampling and imputation variability of each level estimate, not a test of the change between surveys. "Space limits prevent the inclusion of standard errors for all estimates." "Standard errors for all estimates in tables 1 and 2 are available on the SCF website." Two months or more late rose from 4.9% to 8.2% (+3.3 points, this page's calculation).^[1]

SCF vs other surveys. The report's appendix compares 2025 medians of $82,200 in the SCF and $85,800 in the Census Bureau's Current Population Survey (as cited in the report), and notes that the SCF family is a primary economic unit within a household, so SCF and household-based figures are not identical by design.^[1]