Society · U.S. poverty report · September 17, 2026
America’s Poverty Rate Hit a Record Low. A Second Measure Tells a More Complicated Story.
New Census Bureau data show the official poverty rate fell to 10.2% in 2025 and child poverty reached a historic low by that measure. But the Supplemental Poverty Measure held at 13.1%, exposing how housing, taxes, medical bills and government benefits can change the picture of economic security.
Two days after the Census Bureau released its annual poverty report, the most important fact is not that one number went down and another did not. It is that the United States uses two different lenses to answer two different questions about hardship — and in 2025 those lenses moved farther apart.
The official measure, the country’s long-running benchmark, put 34.5 million people below the poverty line and set the national rate at 10.2%, down half a percentage point from 2024. The official child poverty rate fell to 13.4%, the lowest level in the historical series. Yet the broader Supplemental Poverty Measure, or SPM, stood at 13.1%, statistically unchanged from a year earlier.
That is not a statistical contradiction. The official measure counts pretax cash income against a national threshold. The SPM adds resources such as noncash benefits and refundable tax credits, then subtracts expenses including taxes, work costs and out-of-pocket medical spending, while also adjusting housing thresholds for geography. The result is a more demanding test in many places and for many households.
01 · The headline
The official poverty rate fell — and the scale of the improvement matters
The Census Bureau’s September 15 release records a real improvement in the official measure: fewer Americans were counted in poverty, and the national rate reached its lowest point in the series.
The official poverty rate fell from 10.7% in the comparable 2024 estimate to 10.2% in 2025. That represented 34.5 million people living below the official threshold. For a family consisting of two adults and two children, the 2025 official poverty threshold was $32,649.
Children registered a notable decline. Their official poverty rate dropped to 13.4%, a historic low. The Census Bureau also reported that the official poverty rate for Hispanic Americans fell to 13.9%, likewise a historic low for that population. Those changes describe money income relative to the official threshold; they do not, by themselves, measure whether rent, child care, health care or other household costs became easier to pay.
The report arrives alongside another strong headline: real median household income rose 2.6% to $87,460 in 2025, the highest level in the Census series dating to 1967. Post-tax median household income rose 3.1% to $76,060. Income growth and the decline in official poverty therefore point in the same direction, even though the gains were not uniform across the income distribution.
02 · The split screen
Why 10.2% and 13.1% can both be right
The two measures are built to illuminate different parts of a household balance sheet. In 2025, the SPM’s deductions and higher housing-sensitive thresholds outweighed some of the resources that the official measure ignores.
The official measure has its roots in the 1960s. Its basic structure is deliberately stable, which makes it valuable for long-term comparisons. But that stability also means it does not directly reflect many features of a modern household budget.
The SPM was introduced decades later to broaden the picture. It treats refundable tax credits, nutrition assistance and some other benefits as resources. It then deducts items families cannot freely spend on other needs, including taxes, work-related expenses and medical out-of-pocket costs.
Because these additions and subtractions can push in opposite directions, the SPM is not automatically higher or lower. In 2025, it was 2.9 percentage points above the official rate. That gap is a signal to look at what is happening inside the calculation rather than to treat one figure as the “real” number and the other as wrong.
The central 2025 story is not one poverty rate. It is the widening distance between a cash-income benchmark and a measure that also prices the unavoidable costs of daily life.
03 · The threshold
The poverty line itself changes when housing is allowed into the frame
A family can have the same income and face a different SPM threshold depending on whether it rents, owns with a mortgage or owns without one — before geographic housing adjustments are applied.
The Bureau of Labor Statistics finalized the 2025 research SPM thresholds in August. For a two-adult, two-child unit, the national threshold before geographic adjustment was $41,701 for renters, $41,323 for owners with mortgages and $34,326 for owners without mortgages. The official threshold for the same family structure was $32,649.
That difference is not a judgment that every family needs the same budget. It reflects the SPM’s design: thresholds are based on expenditures for food, clothing, shelter and utilities and then modified to recognize housing tenure and local housing costs. A renter in a high-cost metro can therefore face a substantially different poverty test from an owner without a mortgage in a lower-cost area.
- $32,649Official poverty threshold for two adults and two children.
- $34,326SPM threshold for an owner without a mortgage, before geographic adjustment.
- $41,323SPM threshold for an owner with a mortgage, before geographic adjustment.
- $41,701SPM threshold for a renter, before geographic adjustment.
That architecture helps explain why the national SPM can remain elevated even when cash incomes rise. The measure is sensitive not just to what comes into a household, but to the standard of basic need embedded in the threshold and to the costs that must be paid before money is available for food, clothing and shelter.
04 · The safety net
Benefits pull millions above the line; necessary expenses push millions back below it
One of the SPM’s most useful features is its ability to show how specific programs and expenses change the count of people classified as poor.
Social Security remained the single largest antipoverty program in the Census Bureau’s 2025 SPM analysis, moving 28.8 million people out of poverty. The effect is especially large for older Americans, but the program also reaches disabled workers, survivors and family members across age groups.
The same framework shows the contribution of programs that the official measure does not count as cash income. Census-based tabulations for 2025 show refundable tax credits moved about 6.1 million people out of SPM poverty, while SNAP benefits moved about 3.1 million above the line. These are “all else equal” estimates: each component is removed or added to the calculation to estimate its marginal effect, not to simulate a world in which every other program and household behavior stays unchanged.
Expenses move the other way. Medical out-of-pocket costs pushed about 7.7 million people into SPM poverty in the 2025 calculation. That is why health costs matter even in a year when the uninsured rate itself did not materially change: insured households can still face premiums, deductibles, copays, prescriptions and other costs that reduce spendable resources.
05 · Children
A historic low for child poverty comes with an important measurement warning
The 13.4% official child poverty rate is a milestone in the series. It should be read as a strong result for cash income, not as proof that every form of child hardship is at a historic low.
Children live inside household budgets shaped by rent, food, transportation, child care and medical needs. The official measure captures the cash-income side of that equation. The SPM goes further by counting benefits and subtracting expenses, which makes it especially useful when evaluating policies that reach families through the tax code or through noncash programs.
This distinction became unusually visible during the pandemic era, when temporary tax and transfer policies sharply altered child poverty under the SPM. The 2025 report is from a different policy and economic environment, and its historic-low child figure is specifically an official-measure result. The safest interpretation is narrow but meaningful: the share of children living in families whose cash income fell below the official federal threshold was lower than in any earlier year of the series.
That still leaves a substantial share of children below the line, and many families above it can experience material hardship. Poverty thresholds are binary by design; household budgets are not. A family a few dollars above a statistical threshold may face nearly the same financial constraints as one a few dollars below it.
06 · Income
Household income reached a record, but the distribution still matters
The poverty decline did not happen in isolation. Median household income rose, yet the strongest movement was not shared evenly across every slice of the distribution.
Real median household income climbed to $87,460 in 2025 from $85,210 in 2024, a 2.6% gain after inflation. The post-tax median rose to $76,060 from $73,760. Those are broad measures, and they help explain why the official poverty rate could fall even as the SPM remained flat.
But the Census Bureau found no statistically significant change in income inequality as measured by the Gini index. Income at the 90th percentile rose 1.7%, while income at the 10th percentile did not change significantly. A rising median can therefore coexist with limited movement at the bottom.
Racial and ethnic household trends were also uneven. Median income rose 4.8% for Black households, 3.0% for White households and 2.9% for non-Hispanic White households. The Census Bureau did not find a statistically significant change for Asian or Hispanic households between 2024 and 2025.
Among full-time, year-round workers, women’s median earnings rose 3.2% while men’s did not change significantly. The female-to-male earnings ratio increased to 83.9% from 80.6%. That narrowing is notable, but the report describes one year of earnings data rather than a declaration that the long-running wage gap has been resolved.
07 · Health costs
Insurance coverage was stable, yet medical spending still changed the poverty count
The 2025 health insurance figures did not show a statistically significant change in the national uninsured rate. The SPM nevertheless demonstrates why coverage and affordability are different questions.
About 26.7 million people, or 7.9% of the population, were uninsured for the entire year in 2025. That was not statistically different from 2024. Employment-based insurance remained the most common form of coverage, reaching 53.5% of the population for some or all of the year. Medicare covered 20.1%, Medicaid 17.1% and direct-purchase plans 10.5%.
Public coverage overall was 35.8%, also statistically unchanged, but there were shifts beneath that topline. Medicare coverage increased by 0.6 percentage points while Medicaid coverage decreased by 0.5 points. Among adults ages 19 to 64, Medicaid coverage declined 0.6 percentage points.
The SPM subtracts out-of-pocket medical expenses because those dollars cannot be used for other basic needs. This is the core reason a stable uninsured rate does not neutralize the poverty effect of health costs. A person can be insured and still face enough premiums, deductibles, copays or treatment expenses to cross the supplemental poverty threshold.
08 · Geography
The poverty map changes when local housing costs are counted
National averages conceal how differently the two measures behave from state to state, especially where housing costs diverge from the national norm.
For state comparisons, the Census Bureau recommends using three-year CPS ASEC averages because a single year can be too noisy for reliable estimates in smaller populations. Across 2023 through 2025, the national SPM rate averaged 13.0% while the official rate averaged 10.7%.
The SPM rate was significantly higher than the official rate in 26 states and the District of Columbia. It was lower in five states, with no statistically significant difference in the remaining 19. That pattern reflects the mix of local housing costs, taxes, benefits, household characteristics and other expenses incorporated into the supplemental measure.
The official measure showed a different geographic concentration: all but one of the 11 states with three-year official poverty rates significantly above the national rate were in the South; the exception was New Mexico. That does not mean high-cost coastal states are uniformly better off. Rather, it illustrates how a national cash threshold and a geographically adjusted resources-and-costs measure can tell different stories about the same places.
09 · A statistical wrinkle
Part of the flat SPM result came from how the threshold was updated
A Census Bureau working paper released with the report adds an important piece of context: 2025 resources improved enough that the SPM would have declined under a threshold updated only for price change.
The SPM threshold is not simply last year’s line adjusted by inflation. It is tied to a moving standard of expenditures on basic needs. In 2025, that updated standard rose faster than a pure price adjustment would have. Census researcher Adrienne DiTommaso showed that if the 2024 SPM threshold had been carried forward using only price changes, the overall SPM rate would have fallen instead of remaining statistically flat.
That does not invalidate the published 13.1% rate. It explains why a year of rising median income and falling official poverty could still produce a stable supplemental measure. The SPM is intentionally designed to reflect changes in the standard of basic need, not just changes in consumer prices.
There is also a separate historical revision to keep in mind. In July, the Bureau of Labor Statistics corrected errors in the SPM threshold series for 2019 through 2024. The Census Bureau’s August review found that corrected overall rates were 0.1 to 0.3 percentage points lower than previously published for 2019 through 2023, while 2024 moved from 12.9% to 13.0%. Census concluded that the trend was not meaningfully altered.
Readers comparing this year’s figures with older articles should therefore check which vintage of the SPM series is being used. Census also notes that updated population controls can make some 2024 estimates in the newest reports differ from values published a year earlier.
10 · The takeaway
The best reading is neither triumphalist nor alarmist: the measures are telling us different things
The official rate’s decline is real and historically significant. The SPM’s lack of improvement is also real. Together they describe a country where cash incomes strengthened while the broader balance of resources, thresholds and unavoidable expenses remained tighter.
For workers and families, the 2025 report offers evidence of progress: median income rose, post-tax income rose, official poverty fell, and official child poverty reached a historic low. Those are not small changes. They indicate that earnings and cash resources improved enough to move a meaningful number of households above the longstanding federal benchmark.
At the same time, the SPM shows why a rising income number does not settle the question of economic security. Housing costs vary widely. Medical expenses can erase disposable resources. Taxes and work costs matter. Programs delivered through food assistance or refundable tax credits can lift families above a threshold even though they never appear in the official cash-income measure.
The 2025 data also caution against treating “the poverty rate” as a single universal fact. The official measure provides a long, consistent historical series. The SPM is designed to show how taxes, benefits, housing and necessary expenses reshape the household budget. Researchers, journalists and readers need both to see what changed and why.
Finally, the numbers are estimates from a large household survey, not a census of every family. The Census Bureau tests comparisons for statistical significance and warns users about sampling and nonsampling error. The SPM also depends on methodological choices about thresholds and resource accounting. Those caveats do not make the measures unhelpful; they define the boundaries of what the data can support.
For 2026 and beyond, the value of the new release is as a baseline. Future reports will show whether the cash-income improvement continued, whether medical and housing costs changed the SPM balance, and whether today’s gap between 10.2% and 13.1% narrows or widens. The next headline will be easier to understand if the country remembers what those two numbers actually measure.
Sources
Primary federal data and independent reporting used for this article. All links were accessed for the September 17, 2026 edition.
- U.S. Census Bureau, Poverty in the United States: 2025, September 15, 2026.
- U.S. Census Bureau, Income, Poverty and Health Insurance Coverage in the United States: 2025, September 15, 2026.
- U.S. Census Bureau, Income in the United States: 2025, September 15, 2026.
- U.S. Census Bureau, Health Insurance Coverage in the United States: 2025, September 15, 2026.
- U.S. Bureau of Labor Statistics, 2025 Research Supplemental Poverty Measure Thresholds, finalized August 24, 2026.
- U.S. Census Bureau, How Updating Annual Poverty Thresholds Affects Poverty Rates: 2024–2025, September 15, 2026.
- U.S. Census Bureau, Revising Supplemental Poverty Measure Estimates from 2019 to 2024, August 19, 2026.
- U.S. Census Bureau, States With Highest Poverty Rates Clustered in the South, September 15, 2026.
- Reuters, U.S. poverty rate dips to lowest on record, Census Bureau says, September 15, 2026.
- Associated Press, American incomes finally top pre-pandemic levels, fighting against inflation all the way, September 15, 2026.
Comments
Post a Comment