Publish the 2025 Income and Poverty Figures Without Mixing Measures

U.S. real median household income was $87,460 in 2025, and the official poverty rate was 10.2%. The Census Bureau released these estimates on September 15, 2026. Pretax median income rose 2.6% from 2024, while official poverty fell 0.5 percentage points. The Supplemental Poverty Measure (SPM) was 13.1%, not statistically different from 2024. These figures describe different measures and populations; they are not interchangeable indicators of household finances. Source: Census Bureau release.

For a public spreadsheet or data page, keep the observation year, income definition, population and unit attached to each estimate. “2025” describes the year measured, not the year the estimates were published. A page labeled only “2026 income” would misidentify these observations.

Choose the statistic your page needs; the tool shows its value, definition and spreadsheet encoding.

The National Figures Describe Five Distinct Measures

Measure	2025 Estimate	Change From 2024	Population / Unit
Real median income, pretax	$87,460	+2.6%	Households; 2025 dollars
Real median income, post-tax	$76,060	+3.1%	Households; 2025 dollars
Official poverty rate	10.2%	−0.5 percentage points	People; percent
People in official poverty	34.5 million	Not shown here	People; count
SPM rate	13.1%	Not statistically different	People; percent

Income figures come from the Census income explainer; poverty figures come from Poverty in the United States: 2025. “Not shown here” means this summary omits the count’s comparison with 2024—not that the number of people in poverty was unchanged.

The two income rows describe households. The poverty rows describe people. Do not relabel the official poverty rate as the percentage of households in poverty, or present median household income as income per person. A shared national geography does not make the populations equivalent.

Median income is the midpoint: half of households have higher income and half have lower income. It is not average income or an individual worker’s salary. The $87,460 pretax estimate was the highest in the series dating to 1967; the post-tax median was not a record high. Keep that record claim attached to the pretax series rather than applying it to both income rows.

These are national estimates. This table supplies no state, county or city figures and no subgroup breakdowns. If a public page adds a geography filter, the national row should remain labeled “United States”; it should not appear to be a local estimate when a reader selects a place for which the dataset has no observation.

Pretax And Post-Tax Income Need Separate Labels

The headline income measure uses pretax money income and excludes the value of in-kind transfers. Post-tax income accounts for federal and state taxes and credits, plus payroll taxes. Both rows are household medians expressed in 2025 dollars, but their income definitions differ. Source: Census release.

Use “Real median household income, pretax” and “Real median household income, post-tax” as distinct measure labels. A short display heading such as “Household income” is acceptable only when the tax basis and price year remain visible next to the value. Otherwise, a reader cannot tell which of the two estimates a chart or exported row represents.

Do not interpret the gap between these two medians as the tax bill of a typical household. The table contains separate summary statistics, not matched before-and-after tax records for a particular household. It also supplies no household-level tax calculation. A calculator that inferred a personal tax rate from these medians would answer a question the data do not support.

“Real” also needs an explicit price-year field. The dollar sign identifies currency, but it does not identify the inflation basis. In a single-year display, “2025 dollars” can sit in the column heading or caption. In a reusable dataset, store that information with each income observation so it survives filtering, copying and export.

Official Poverty And SPM Answer Different Questions

The official poverty measure compares pretax money income with a national threshold adjusted for family composition. The SPM also accounts for certain assistance programs, taxes, work and medical expenses, and geographic variation in housing costs. Its broader treatment of resources and expenses means it is not simply another label for the official rate. Source: Census poverty report.

The SPM supplements rather than replaces the official measure. Publish the 10.2% official rate and the 13.1% SPM rate as separate rows or series. Do not average them into a single “poverty rate,” and do not treat their difference as a year-over-year change. They are two measures for the same observation year, not successive observations of one measure.

The official poverty count, 34.5 million people, belongs with the official poverty measure. This summary does not provide an SPM count. A page should show an unknown or omitted value for that count rather than copying the official count into the SPM row.

Nor does this national summary provide the poverty threshold for a particular family. The official rate is a population estimate, not an income cutoff that can determine whether an individual household is below its applicable threshold. The supplied figures support a national comparison table, not a personal poverty-eligibility calculator.

Changes Need Their Own Units And Significance Notes

The income changes are percent changes. The official poverty change is a percentage-point change. A fall of 0.5 percentage points in a poverty rate is not a fall of 0.5%. The distinction must remain visible in table headings, chart annotations and downloads.

Store the official poverty change as -0.5, with percentage_points in a separate change-unit field. Store the pretax income change as 2.6, with percent as its change unit. Do not place both numbers in a column labeled only “Change (%)”; that heading would misstate the poverty comparison.

For the SPM, preserve the reported wording: “not statistically different from 2024.” That does not mean the estimates were identical, and it does not justify entering a numeric change of zero. This summary does not provide the SPM’s numeric change, so leave that field blank and retain the significance statement in its own field.

Census tests reported comparisons at the 90% confidence level unless otherwise noted. Source: Census release. Statistical significance and change magnitude belong in separate fields: one describes the reported comparison’s statistical status, while the other records the numerical change when available.

Apply the same discipline to the official poverty count. Its comparison is omitted here. A blank change value with an explanatory note preserves that limitation; a zero would turn missing information into a claim of no change.

A Reusable Spreadsheet Keeps Values And Metadata Together

Use one row per observation year, geography, population and measure. That structure lets the same dataset feed a table, a measure selector or a chart without requiring readers to infer definitions from column position.

Field	What To Store
observation_year	2025
release_date	2026-09-15
geography	United States
population	Households or People
measure	Explicit measure label
value	Numeric value only
unit	USD, percent or people
price_year	2025 for real income
change_value	Reported numeric change
change_unit	percent or percentage_points
significance_status	Reported significance wording
comparison_note	Comparison, omissions, rounding
source_url	Original Census source

For the official poverty rate, store 10.2 with unit percent. Do not mix 10.2 and 0.102 in the same value column. Both can represent the same rate under different conventions, but a dataset needs one convention and an explicit unit. Check the spreadsheet’s percent formatting before publishing: the stored value and the displayed percentage must agree.

For the official poverty count, store 34500000 with unit people and a note that the source figure is rounded to 34.5 million. The expanded integer is a convenient numeric encoding, not an exact headcount. Keep the rounding note visible in row details or the table caption so exported precision does not imply greater source precision.

For the income rows, store 87460 and 76060 without currency symbols or thousands separators. Apply dollar formatting in the display layer, while retaining USD and 2025 as the unit and price year. That separates calculations from presentation without stripping away the meaning of the numbers.

Leave price_year blank for poverty rates and counts because it is not applicable to those values. Distinguish that blank from an omitted comparison: both may look empty in a spreadsheet, but they have different meanings. A public page can render an em dash while row notes explain whether the field is inapplicable or unavailable in this summary.

Keep source URLs at the row level if the dataset may later combine measures from different reports. A single page-level citation can become ambiguous after filtering or export. The income and poverty observations should remain traceable even when a reader downloads only one measure.

Historical Comparisons Must Use The Same Release Vintage

Use the same release vintage for historical comparisons. The income report expresses historical estimates in 2025 dollars and warns that its 2024 estimates may differ from the prior release because of updated population controls. Its comparable 2024 pretax median is $85,210. Source: Income in the United States: 2025.

If you add a 2024 income row, use that comparable estimate and its stated price basis rather than silently joining a figure copied from an older release. Otherwise, a two-year chart may combine different dollar bases or revised estimates while appearing to show a clean comparison.

Keep observation_year separate from release_date throughout the dataset. The former determines the position on a time-series axis; the latter identifies the publication vintage. Sorting the observations by release year and labeling the axis “Income year” would place these estimates in the wrong year.

Retain the published 2.6% pretax change alongside the reported dollar estimates. If a spreadsheet recalculates a change from displayed values, distinguish that calculation from the source-reported change; rounding can affect whether the displayed calculation reproduces the published figure exactly. Do not overwrite the source field without labeling the replacement.

The Public Page Must Preserve Meaning After Filtering

Publish the reviewed aggregate table, not household-level records containing sensitive information. Keep source links, definitions and the release date visible alongside the data. A useful caption is: “U.S. household income and poverty, calendar year 2025; Census Bureau estimates released September 15, 2026.” The guide to data table captions and summaries explains how to describe units and missing values.

After applying a measure filter, check that the population and unit remain visible. A card showing only “10.2” loses the percent unit; one showing only “$87,460” loses the pretax definition and price year. Row details can carry longer definitions, but those essential labels should not require a separate lookup.

Use separate chart scales or separate charts for income dollars and poverty percentages. A shared unlabeled axis would obscure both measures. Keep official poverty and SPM series explicitly named, and preserve the SPM significance note wherever the year-over-year comparison appears.

The published page should therefore retain five distinct observations, not compress them into a single household-finances score. Readers can compare the national figures while still seeing which population, definition, unit and comparison each number represents.