TANF hasn't gotten a raise since 1996
Summary
Congress fixed the federal welfare block grant at $16.4 billion in 1996 and never adjusted it. Inflation has since cut its real value by half, only 21.8 cents of every TANF dollar now reaches a family as cash aid, and the caseload has collapsed from roughly 5 million families to about 1 million.
Follow the dollar
is really two pots of money stacked together: a fixed federal block grant and whatever states kick in to satisfy their required "maintenance of effort." In fiscal year 2024 the combined pot totaled $37.5 billion, per the Administration for Children and Families' FY2024 TANF and MOE spending report. Only $8.18 billion of it — 21.8 cents of every dollar — was basic assistance, the cash payments the program is best known for.
View data as table
| Basic Assistance (cash) | $8.18B | 21.8% |
|---|---|---|
| Child Care | $6.38B | 17.0% |
| Refundable Tax Credits | $4.05B | 10.8% |
| Program Management | $3.68B | 9.8% |
| Pre-K / Head Start | $3.45B | 9.2% |
| Work, Education & Training | $2.89B | 7.7% |
| All other categories | $8.82B | 23.5% |
The rest is legal, but it isn't cash aid. $6.38 billion went to child care, $4.05 billion funded state refundable tax credits, $3.68 billion paid for running the program itself, and $3.45 billion went to pre-kindergarten and Head Start — all allowable uses under the block grant's loose spending rules. Even work, education, and training — the activity in the program's own name — got only $2.89 billion, 7.7 cents of the dollar, about a third of what cash assistance received.
A grant frozen in 1996
None of this happened because the block grant grew and states diversified spending. It happened because the grant stopped growing at all. The state family assistance grant — $16.4 billion a year — was set by the Personal Responsibility and Work Opportunity Reconciliation Act of 1996 and has been paid at that exact nominal amount every year since, per the Congressional Research Service's TANF primer. No adjustment for inflation, population growth, or how many families are actually poor.
View data as table
| Nominal grant, 1996–2026 | $16.4B | unchanged |
|---|---|---|
| CPI-adjusted 1996 value, in 2025 dollars | $33.66B | BLS CPI-U |
Using the Bureau of Labor Statistics' CPI-U index, average consumer prices have more than doubled since 1996 (index 156.85 to 321.94). To buy in 2025 what $16.4 billion bought in 1996, the block grant would have to total roughly $33.66 billion — more than double what Congress actually appropriates. Put another way: today's $16.4 billion buys about what $8 billion did in 1996. That's a 51% real-terms pay cut Congress never voted on; it just let the calendar do it.
The same freeze, fifty different states
The freeze compounds a second one. Each state's share of the $16.4 billion is locked to that state's peak federal welfare spending during fiscal years 1992 through 1995 — a formula untouched even as state populations and poverty rates shifted for three decades. The result is a per-poor-child grant that varies by more than twelve-fold from state to state.
View data as table
| Texas | $358 |
|---|---|
| Mississippi | $595 |
| Georgia | $822 |
| Ohio | $1,047 |
| California | $1,268 |
| New York | $1,426 |
| Vermont | $4,420 |
Texas — the country's second-most-populous state — gets $358 per poor child, the lowest figure in the nation, per the CRS primer. Vermont gets $4,420, more than twelve times as much, because Vermont's federal welfare spending happened to run high in the early 1990s and Texas's ran low. Nothing about either state's current child poverty rate has any bearing on the number.
That state-level gap sits inside a national collapse in reach. In fiscal year 1995, the year before welfare reform, roughly 5 million families received cash assistance under 's predecessor program, and 71% of program dollars went to that assistance, according to the CRS primer. By fiscal year 2024, the caseload — now under , still funded at 1996's dollar figure — was near 1 million families. Child poverty did not fall by four-fifths over that period; the program's willingness and financial ability to reach poor families did.
The takeaway
- The grant is frozen; the mission drifted. $16.4 billion in 1996 became a real-terms $8 billion by letting inflation run for thirty years — a 51% cut nobody had to legislate.
- "Assistance" is now a minority use of the money. 21.8 cents of every dollar is cash aid to a family; the rest funds child care, tax credits, pre-K, administration, and more, all legal under the block grant's broad rules.
- The state formula never updated either. A child's state of residence can be worth twelve times more or less in aid, based on a funding snapshot from before that child was born.
Figures cover federal fiscal year 2024 (Oct. 2023–Sep. 2024) for spending and caseload data, and calendar year 1996 through 2025/2026 for the block-grant and inflation comparison; sources and vintages are noted on each chart.
Sources
- Administration for Children and Families, Office of Family Assistance — and MOE Spending and Transfers by Activity, 2024: United States (updated Nov. 21, 2025) — the national FY2024 spending pie chart underlying the "Follow the dollar" figure. acf.gov
- Congressional Research Service — Temporary Assistance for Needy Families () Block Grant: A Primer (R48413, 2025) — the $16.4 billion nominal block-grant figure and its 1996 statutory origin, the FY1995 vs. FY2024 assistance-share comparison, the ~5 million vs. ~1 million family caseload comparison, and the state-by-state -grant-per-poor-child figures. congress.gov
- U.S. Bureau of Labor Statistics — CPI-U series CUUR0000SA0 (U.S. city average, all items, not seasonally adjusted), retrieved via the public data API — basis for the 1996-to-2025 inflation adjustment on the block grant (BlackLeaf calculation). data.bls.gov
Comments
Always open. Logged-in readers can annotate paragraphs in place.
Temporary Assistance for Needy Families was built to replace welfare with a block grant states could spend flexibly. Thirty years later, "temporary" and "flexible" turned out to be the operative words. The dollar amount has never moved. What states are allowed to call "assistance" has stretched to cover almost anything. And the number of poor families who actually get a check has fallen by roughly four out of every five.