A long-time concern in the US K-12 schooling system is that local property taxes within each school district are a primary source of funding. As a result, school districts in high-income areas with higher property values will tend to have more money to spend per student. While this was indeed a pattern 40-50 years ago, school finance issues are no longer that simple. Sarah Reber, Beyond Deng, and Gabriela Goodman explain in “School Spending and Equity since 1976″ (Center for Economic Security and Opportunity at Brookings, June 2026). They write:
Schools in high-poverty or predominantly non-white communities often fail to serve their students well. Many point to American schools’ reliance on local property taxes as the culprit, arguing the school finance system ensures that “affluent areas end up with well-funded schools and low-income areas end up with poorly funded schools,” (American University 2020). However, state governments now contribute about as much to school funding as school districts raise locally, and the federal government also contributes. Moreover, state and federal funding disproportionately benefits school districts serving low-income students, and decades of school finance reforms mean that schools’ spending is less tied to its property wealth than in the past. …
We present a comprehensive descriptive analysis of how school spending varies across districts depending on child poverty and racial composition and how that has changed since 1976. To make sense of these patterns, we also decompose differences in average spending into their between-state and within-state components. Average spending varies tremendously across states—some spend nearly three times what others do (Reber and Goodman 2025)—so we ask: When some districts spend more than others, is it because they are in states where all districts spend more or because they spend more than other districts in the same state? …
The analysis shows that between 1976 and the early 1990s the average level of spending increased, but the distribution of spending changed only modestly, becoming somewhat less progressive. After about 1995, spending within states was increasingly targeted to high poverty districts, though between-state differences increasingly favored lower-poverty districts. That is, low-poverty districts were more likely to be in states where average spending was higher, but they had lower spending relative to high-poverty districts in the same state. We also find that low-poverty and predominantly white districts were more protected from spending cuts during the Great Recession, and high-poverty or predominantly non-white districts were the least protected. Finally, differences in average spending between states remain a major driver of spending inequality overall, even though policy conversations often focus on within-state differences. …
The analysis does not speak to whether high-poverty or minority-serving schools have “enough” funding—or whether or under what conditions additional funding would improve student outcomes. Instead, we document what happened: Which districts spent more, and which spent less? How has the distribution of school spending changed? This nearly five-decade national perspective captures how spending patterns evolved as states reformed their finance systems, sometimes multiple times, during the school finance reform era.
Here are a few figures to illustrate their main themes. This figure looks at school districts across the country in 1976-77 and 2020-21, split up by quintiles in terms of poverty rates and share of nonwhite students. Per student spending for each group is shown above the bars. New York City is a national outlier in terms of how much is spent per student. As the authors write: “Before the 1990s, NYC actually spent less than its own-state average but still had spending well above the national average (because its own-state average was high). By the 2010s, spending in NYC was well above average for New York state, which itself was nearly double the national average spending. Throughout the period we study, spending in New York state grew faster than average (from a relatively high starting point), and spending became more progressive (at least for NYC) faster than in the average state. Together these trends moved NYC from a moderate to more extreme high-spending outlier.” Thus, the white rectangles show wha tthe numbers at the top end would be like if New York City was excluded. The 2020-21 data suggest that, at the national level, there is no linear relationship in which school districts with more poverty or a greater share of nonwhite student have the lowest level of per capita spending. Instead, there is a U-shaped relationship, although the lowest quintile is below the top quintile.
Some of the difference here results from large differences in K-12 spending across states: in 2020, for example, “ranging from about $9,000 per pupil in the lowest spending states to almost $30,000 in the highest spending states.” In addition, “There is some geographic correlation between average spending and the child poverty rate. For example, both high-spending states and low-poverty districts are concentrated in the Northeast. … Nearly all states have at least some school districts in each poverty category, though states
with higher average incomes and school spending have fewer high-poverty districts and vice versa.” Differences in per student K-12 spending can be divided up into variation within states and variation between states. Consider this table:
For 1976 and for 2020, the per pupil spending numbers in the first two rows match the number from the bar chart above (with New York City left out). The third row subtracts the second row from the first. The final row shows how the spending gap would change if every school in each state had the same per pupil spending–but the differences across states were the same as they they are now. In the 2020 data, for example, “Average spending for
the highest-poverty districts was $17,208, compared to $15,921 for the lowest-poverty districts, a difference of $1,287. If every district spent at its own-state average, that gap would have been $1,084 in the other direction. Within states, highest-poverty districts spent $2,371 more, on average, compared to their lowest-poverty counterparts.”
Thus, within states, high-poverty school districts have higher per pupil spending. But across states, high-poverty school districts also tend to be in states with lower per pupil spending, so if one looks at national-level data, the within-state efforts to spend more on low-income students are less apparent.
The bottom line is to be wary of common claims that school districts with greater poverty or more nonwhite students have lower per pupil spending. Within states, this claim is generally wrong. Across states, the claim is more complex, because It costs more to hire teachers and to staff a school in higher-income states, so some of the difference across states is about differences in cost-of-living, not actual resources (like numbers of teachers) available to students. The report digs into such adjustments in some detail. It turns out that studies that make claims about fewer resources for higher-poverty schools have a tendency to do idiosyncratic cost-of-living adjustments and also to exclude federal K-12 spending, which is weighted toward students from lower-income households.
