The latest Census data contains a comforting number. The official U.S. poverty rate fell half a point to 10.2% in 2025, about 34.5 million people, and child poverty hit a record low. The Supplemental Poverty Measure, which accounts for taxes, government benefits, and major household expenses, came in somewhat higher at 13.1%. Median household income reached a record $87,460.12 Read together, those figures describe an economy that is getting better.10
Measure something different and the picture changes.
In December 2025, the Urban Institute found that 24.2% of American adults reported household food insecurity over the previous 12 months, and among working-age adults the rate was 27.7%.4 USDA's separate household survey found 13.7% of U.S. households—18.3 million—were food insecure in 2024, including 5.4% with very low food security.3
Those numbers are not interchangeable. USDA measures households with its long-established food-security module; Urban's survey measures adults reporting household food insecurity. Combining them into one synthetic statistic would be a mistake. But they point the same way:
The official poverty line captures only a portion of the population experiencing material insecurity.
Important
Break the country into income quintiles and the size of that gap becomes hard to ignore.
Poverty is a line. Economic insecurity is a gradient.
The official poverty measure asks a narrow question: does pretax cash income fall below a threshold set by household composition? Food insecurity asks something else entirely—after everything else has been paid, can the household reliably afford adequate food?
The distinction matters because households don't experience their finances as a ratio to the federal poverty line. They experience rent, utilities, insurance, transportation, health care, child care and debt service, in roughly that order of non-negotiability, and then food. The Supplemental Poverty Measure captures more of that reality, which is why it lands nearly three points above the official rate.1 Food insecurity reaches considerably farther still.
USDA found that 39.4% of households below the poverty line were food insecure in 2024.3
Urban's 2025 results show how far the gradient extends above that line for working-age adults:4
| Family income | Reported household food insecurity |
|---|---|
| Below 200% of FPL | 51.0% |
| 200%–400% of FPL | 34.2% |
| 400%+ of FPL | 9.3% |
The middle row carries the story. More than one-third of working-age adults between 200% and 400% of the poverty level reported food insecurity, and the rate rose from 30.6% in 2024 to 34.2% in 2025—the only band where Urban flagged a year-over-year increase.4
These are not households anyone would describe as poor. Many are what we casually call the lower-middle class, and some are plainly the middle class.
Put the problem into quintiles
Income quintiles divide U.S. households into five equal groups. Using the latest Census household-income distribution, the approximate 2025 boundaries look like this:2
| Quintile | Pretax household income | Share of households |
|---|---|---|
| Q1 — Lowest | Below ~$35,800 | 20% |
| Q2 | ~$35,800–$68,600 | 20% |
| Q3 — Middle | ~$68,600–$110,300 | 20% |
| Q4 | ~$110,300–$182,400 | 20% |
| Q5 — Highest | Above ~$182,400 | 20% |
In 2025, 200% of the poverty level was roughly $31,300 for a single adult and $53,300 for a family of three.4 Doubling those gives the top of the band:
| Household | 200% FPL | 400% FPL |
|---|---|---|
| Single adult | ~$31,300 | ~$62,600 |
| Family of three | ~$53,300 | ~$106,600 |
Overlay the two tables.
For a single adult, the 200–400% band runs from about $31,300 to $62,600—the upper edge of Q1 through most of Q2. For a family of three, the same band runs from about $53,300 to $106,600, stretching from the upper part of Q2 through nearly the entire third quintile.
That is where the poverty statistic breaks down as a proxy for economic security.
For a family of three, the income range in which Urban observed 34.2% food insecurity extends almost to the top of America's middle income quintile.
Not the bottom decile, and not only those officially classified as poor. Into Q3.
The distribution looks nothing like the headline rate
Mapped onto the quintiles, exposure looks roughly like this:
| Quintile | Approx. income | Economic exposure |
|---|---|---|
| Q1 | <$35.8K | Acute; poverty and food insecurity heavily concentrated |
| Q2 | $35.8K–$68.6K | High; large overlap with the 200–400% FPL population |
| Q3 | $68.6K–$110.3K | Meaningful, especially for families with children and high fixed costs |
| Q4 | $110.3K–$182.4K | Substantially greater buffer, though geography and household size matter |
| Q5 | >$182.4K | Relatively low national exposure |
The final column is an analytical classification, not a measured food-insecurity rate by quintile. Food-security data are published by poverty ratio rather than Census quintile, and household size prevents a clean one-to-one conversion. But the overlap is large enough to establish the central point...
The population exposed to food insecurity is far larger than the population counted as poor.
How 24% becomes 35%
Starting from Urban's adult rate of 24.2%, reaching 35% requires another 10.8 percentage points of the adult population to cross into food insecurity. Reaching 40% requires 15.8 points.
Those sound like enormous shifts until they are set against the income distribution. Each quintile holds about a fifth of households. A deterioration affecting a share of adults roughly equal to half of one quintile moves the national rate near 35%; a share near four-fifths of a quintile moves it near 40%. Quintiles are counted in households and the food-insecurity rate in adults, so the equivalence is approximate—but the order of magnitude is right.
| Additional population becoming food insecure | Resulting rate |
|---|---|
| +5 percentage points | 29.2% |
| +10 points | 34.2% |
| +15 points | 39.2% |
None of this is a forecast. It is a measure of how little movement is required. The national rate does not need poverty to reach 35% for food insecurity to approach 35%. It only needs enough households sitting on the margin between security and insecurity to lose that margin—and the households immediately above the poverty threshold are already the ones under the most pressure.
Food is the household budget's shock absorber
Most households cannot cut their rent by 15% this month. The car payment is fixed, the electric bill arrives regardless, and insurance premiums and prescriptions don't shrink because groceries got more expensive. Food is the one major line item that bends.
A household under pressure substitutes cheaper products, then narrows variety, then drops meat, then skips meals. Adults eat less so children can eat normally. Someone visits a food pantry for the first time. Somewhere along that sequence the household crosses from food secure to food insecure.
USDA's definition captures exactly that dynamic: insecurity occurs when a household struggles to acquire adequate food for lack of money or other resources, and its most severe category—very low food security—involves disrupted eating patterns and reduced intake.3
Important
That makes food insecurity an early warning indicator for broader financial distress. A family does not have to fall below the poverty line before the grocery budget starts absorbing everything else.
Now introduce OBBBA
The 2025 reconciliation law did not create this vulnerability; the data shows it existed beforehand. What the law does is remove resources disproportionately from the portion of the population already closest to the boundary.
CBO estimates the law reduces federal SNAP spending by roughly $211 billion through 2035, through lower participation, smaller average benefits and shifting part of the program's cost to states, alongside large reductions in Medicaid spending and enrollment.57 Its distributional analysis is more instructive than the topline: resources fall for households toward the bottom of the income distribution, driven mainly by cuts to in-kind benefits such as SNAP and Medicaid, while resources generally rise for households in the middle and toward the top, driven mainly by tax changes.58
Early implementation is running ahead of the projections. The Center on Budget and Policy Priorities reports that SNAP participation and spending are falling faster than CBO assumed, including among people the law did not directly target.6
That shifts the relevant question. It is not simply how many people lose SNAP. It is how many households lose enough total resources to cross from economically stressed but food secure into food insecure. Those are different numbers, and because Q2 and much of Q3 already overlap the income range showing elevated food insecurity, the second number can reach much farther up the distribution than conventional debates about welfare programs assume.
The cliff isn't at the poverty line
Take a family of three earning $80,000. Nationally, that household sits in the third quintile and well above poverty. It is also at roughly 300% of the poverty level—squarely inside the band where Urban found 34.2% food insecurity, and where nearly one-third of working-age adults living with children reported it.4
After federal and payroll taxes and a modest state tax, that family takes home something like $5,700 a month. A plausible set of fixed costs:
| Monthly expense | Amount |
|---|---|
| Rent or mortgage | $2,200 |
| Child care | $1,200 |
| Transportation | $700 |
| Health insurance and medical | $600 |
| Utilities and communications | $500 |
| Total | $5,200 |
That leaves roughly $500 a month for groceries, clothing, car repairs, school costs, debt payments and anything unexpected. The grocery budget is whatever survives the rest. This household is simultaneously above the poverty line, above SNAP eligibility, statistically middle income and financially fragile.
There is no contradiction in that; the measures are describing different things.
At risk is not the same as food insecure
None of this supports the claim that half of America is food insecure. The defensible conclusion is narrower and still alarming:
Roughly one-quarter of adults already report household food insecurity, while the economically exposed population extends considerably farther—through much of the second income quintile and, for many families, into the third.
"At risk" and "food insecure" are different states, but risk is where policy and economic shocks do their work, before households cross the measured threshold. America has three populations, not two: the secure, the food insecure, and a large group between them that is secure but without much margin.9
That third group is the one conventional statistics hide.
The measurement problem
This becomes more consequential over the next several years. Urban positions its survey partly as a way to fill gaps left by the federal government's decision to stop collecting food-security data in selected federal surveys.4
Falling program enrollment cannot, by itself, demonstrate improving conditions. SNAP participation can decline because incomes rose and need fell—genuine progress. It can also decline because eligibility tightened and administrative friction increased while the underlying need stayed put. The observable count falls in both cases.
A reduction in the number of people receiving assistance is not evidence of a reduction in the number of people needing assistance.
Only independent measures—food insecurity, food-bank demand, skipped meals, rent delinquency, utility arrears—can tell the two stories apart. Cutting back on measurement at the same moment the eligibility rules change makes that distinction harder to draw, exactly when it matters most.
The real dividing line
The poverty statistics are not wrong. They answer a question too narrow for the argument they are being asked to settle. A 10.2% poverty rate says how many Americans fall below the official threshold. It does not say how many households lack resilience, how many could absorb a $300 monthly shock, or how many can reliably afford food.
The clearest evidence sits in the middle of the distribution. Below 200% of poverty, food insecurity among working-age adults is already 51%. Between 200% and 400%, it is 34.2% and rising—and for a family of three, that band runs nearly to $107,000, almost the top of the third quintile.
The question is no longer how many Americans are poor. It is how far up the income distribution fragility has already climbed, and how many households can lose their remaining margin before the official numbers register it. By the time the poverty rate signals that something has gone badly wrong, the grocery budget of an $80,000 family will have been saying it for years.