Funding High-poverty School Districts: Federal Policy Tools and the Limits of Incentives

Public policies that influence the amount and distribution of funding for elementary and secondary schools are important policy levers even if there is no consensus on exactly how much funding is needed and the best way to spend school funds. Funding available in the schools attended by children living in poverty is of particular interest: To help equalize opportunity, higher-poverty schools need to offer more opportunities to learn. In this brief, we discuss the federal role—what it has been and what it could be—in funding schools attended by children living in poverty, with particular attention to a federal effort to incentivize changes to state policies.

Efforts to increase equity in school finance generally focus on the states, but federal policy makers are intrigued by the idea of making the most of their limited budget. Can federal grants be designed to change how much state and local governments spend on schools, leveraging limited federal funds into bigger changes in the level and distribution of school spending? This brief analyzes a policy designed to do just that, the Education Finance Incentive Grant (EFIG) formula of Title I of the Elementary and Secondary Education Act (ESEA).

Title I is a federal “compensatory education” program that makes grants to districts mainly based on child poverty. EFIG is one of four statutory formulas determining how Title I funds are allocated and currently accounts for about a quarter of Title I dollars. We focus on EFIG because of its explicit goal to change the incentives that state policy makers face. The effectiveness of the incentives in EFIG is inherently limited by the small size of the program—there is simply not that much money at stake. For FY 2020, EFIG accounted for 7 percent of federal aid and only half of one percent of total revenue to school districts.

We describe how the formula's incentives are negligible and should not be expected to yield more, or more equitable, state spending on schools. For example, the formula's “Effort Factor” is meant to encourage more spending from state and local sources, but a 1 percent increase in combined state and local funding would have no impact on the Effort Factor in thirty-five states. The largest implicit match rate would be in Mississippi, where each additional dollar of state and local funds would be matched with just one and a half cents more of EFIG funds.

What else might federal policy makers do to increase funding in high-poverty schools? We discuss why designing a formula with stronger incentives would be difficult. More generally, this type of approach faces a fundamental tension between two goals—rewarding “good” state policies as an incentive, and directly supporting districts in need because their states have “bad” policies. In addition, state fiscal capacity is at least as important for spending in schools attended by students living in poverty as which type of school finance system their state adopts. We argue federal policy makers should be more attentive to state fiscal capacity, in addition to indicators of student need, in designing education grant programs.

Federal education grants aim to change the distribution of resources, both directly by distributing funds and, in the case of EFIG, indirectly by changing the level and distribution of state and local funding. To better understand the problem these programs are attempting to address, we first review how and why school spending varies across school districts and states.

An important reason that districts in higher-poverty states spend less is because their states have smaller tax bases. Figure 2b shows the correlation between state fiscal capacity, proxied by per capita income, and school spending for children living in poverty is much stronger than the poverty-spending correlation in the first panel. An increase of about $8,800 in per capita income (again, one standard deviation) is associated with an additional $2,200 in per pupil spending.

The bulk of federal funding for elementary and secondary education is distributed directly as grants to state and local agencies. Title I of ESEA (“education for the disadvantaged”) and the Individuals with Disabilities Education Act (“special education”) together constitute nearly three quarters of the U.S. Department of Education's elementary and secondary spending, accounting for over $30 billion (U.S. Department of Education 2022). Both programs determine district-level allocations using formulas based in large part on school-aged children living in poverty. Federal grants could influence school spending for children experiencing poverty directly (increasing federal revenue to increase total revenue), and/or indirectly by altering incentives for spending on schools from state and local revenue.

The primary goal of federal grants is to support spending on targeted populations or services. But even the seemingly straightforward idea of increasing how much school districts spend by giving them money faces obstacles.

State and/or local governments may reduce their own fiscal effort when they receive federal funds, “crowding out” part or all of the new revenue, such that increases in total spending are smaller than the grant. Evidence on crowd-out is mixed (Gordon 2004; Cascio et al. 2013), but suggests that federal funding increases spending in the targeted districts at least somewhat. Federal policy makers should continue to make grants to districts despite the likelihood that federal funding will be partially offset by reductions in state or local funding. Federal grants can increase spending in high-poverty districts, and only federal programs can address cross-state spending gaps, a major source of inequality in school spending.

The between-state differences in average spending shown in figure 2 are quite large, so closing them would require large increases in federal funding, even setting aside concerns about crowd-out. For example, Mississippi and Massachusetts both spend about what their respective per capita incomes would suggest, but the average student in poverty in Massachusetts lives in a district that spends about $21,000 per pupil, while the average Mississippi student in poverty lives in a district that spends about $12,600 (adjusted for differences in wages). A federal transfer program that substantially shrank this gap would send more than $8,000 per pupil to Mississippi.

The amount that schools need to spend to fulfill their missions depends not only on student needs, but also on the prices of educational inputs, especially wages for staff. Both Title I and IDEA formulas use per pupil spending from state and local funding, known as state per pupil expenditure (SPPE) to proxy for geographic variation in prices. However, SPPE reflects variation in both quantities and prices of educational inputs. Its use in the formula means that more federal funds are directed to districts whose states already are investing more in education. Instead of using SPPE, policy makers could use a measure designed specifically for this purpose, such as the Compensating Wage Index for Teachers (CWIFT) developed by the National Center for Education Statistics.

Because the relative federal contribution to school spending is small, policy makers are interested in using federal funds as leverage to encourage changes in state policies that affect state (and local) contributions to spending. This could take two forms: (1) requiring states or local districts to meet specific policy conditions to receive federal funds, or (2) using matching grants that vary the amount of federal funding as a function of state and/or local spending.

Conditional federal funding has so far involved policies about how education is delivered, rather than how it is funded. For example, under the No Child Left Behind Act of 2002, states had to develop test-based accountability programs, and in the 1960s, districts had to comply with the Civil Rights Act by desegregating to receive funds (Cascio et al. 2010). Federal policy makers could use conditional funding to encourage state and local governments to spend more and/or distribute funding differently. For example, they could condition funding on the adoption or elimination of specific school finance policies, or structure a tournament to promote desired policies. But such approaches would require policy makers to identify which specific state policies are more likely to achieve desired outcomes; unfortunately, there is not a consensus among researchers or advocates about which policies should be incentivized. Further, state school finance systems are complex and vary across states; the best approach may depend on the existing system and could therefore be different in each state. For these reasons, the conditional funding approach is unlikely to improve equity in school finance.

Instead of making grants conditional on other policies or actions, the federal government could match state or local spending at a defined rate, allowing recipient jurisdictions to spend more than they raise from state or local sources. For example, the federal government might give school districts $0.25 for every dollar they spend, or target the match to encourage particular types of spending (e.g., school building funds). Because matching grants provide an incentive for states and school districts to raise more revenue, they are less susceptible to crowd-out than the formula grants described earlier, though matching grants targeted to specific types of spending could be subject to crowd-out (Baicker and Staiger 2005). The simplest version of a matching program would likely be regressive, giving more funding to high-spending states, which tend to be higher income (as shown in figure 2). To mitigate this, federal policy makers could vary the match rate based on state fiscal capacity. For example, the match rate for traditional Medicaid is higher in states with lower per capita income.

The original Title I formula awarded federal funds per “formula count” child (primarily school-aged children living in poverty), with the funding per formula count child varying based on SPPE. Congress has since tweaked the basic grant formula and added three more formulas (all four are described in detail in Gordon and Reber [2023]). In FY 2021, Congress appropriated more than $16 billion, or about $1,900 per child living in poverty, to Title I.

The EFIG formula was designed and passed in 1994 and first funded in 2002; in FY 2021, EFIG accounted for $4.3 billion, about one quarter of total Title I funding. Federal policy makers incorporated matching components into the formula hoping to influence the average level and the distribution of state and local funding across districts within a state. The EFIG formula uses two additional state-specific factors called the Effort Factor and the Equity Factor. These factors create implicit match rates and are intended to create incentives for states to increase average spending (effort) and influence the distribution of funding across districts (equity): Holding the other components of the formula constant, total EFIG funding for a state is proportional to the product of its Effort and Equity Factors. Although the Effort and Equity Factors depend on revenue from local as well as state sources, the incentives are meant to influence state policy since the actions of one district have a very small effect on the Effort or Equity Factors for the whole state.

We next describe the Effort and Equity Factors and use simulations to estimate the formula's implied match rates. That is, we estimate how much more EFIG funding a state could expect for its districts if it increased state and local support for schools overall, or disproportionately for school districts serving children living in poverty.¹ We find the implicit match rates are exceedingly small, and therefore unlikely to influence state policy.

The Effort Factor is intended as an incentive for states to spend more on education. To explain the Effort Factor, we draw on the following (indented) text from Gordon and Reber (2023), with minor edits and updated data.

Mississippi's very low implicit match rate is actually the highest in the nation. Idaho is more representative: Thirty-five states would receive no additional EFIG funding if they increased state and local funding by 1 percent, an implicit match rate of zero. No simple tweak to the EFIG formula could make the match rate significantly higher, because total EFIG funding is so small compared to total spending. Even a tenfold increase in EFIG funding would only increase the match rates in Mississippi to 15 cents on the dollar.

Removing the minimum and maximum for the Effort Factor would ensure that all states face at least some incentive. However, policy makers face a fundamental tradeoff when they attempt to incentivize higher spending with a match (or a conditional grant, for that matter). The goal is to get more resources to students in states that have low spending, but states that do not respond as much to the incentive—and choose to spend less relative to their fiscal capacity—also get less in federal funding; the federal funding accentuates rather than offsets spending disparities. Indeed, the caps on the Effort Factor are a way of moderating this relationship, but they also blunt the incentive effects. Uncapping the Effort Factor would mean more states would face a positive (though still small) match rate, but at the cost of transferring more federal funds away from districts who already suffer from low state spending conditional on income.

The Equity Factor is a measure of how combined state and local spending varies across districts within a state. The goal is to reward states where the distribution of per pupil spending (from state and local sources) is more equal, under the assumption that state policy influences the distribution of state aid and/or how much local districts tax and spend.² To explain the Equity Factor in greater technical detail, we draw on the following (indented) text from Gordon and Reber (2023), with minor edits and updated data.

EFIG faces several serious challenges. First, the simulations above show that the financial incentives states face under EFIG are small. In addition, state policy makers may not understand the EFIG incentives nor know what policy changes would increase their EFIG funding. The formula is complicated and uses a three-year moving average of spending data up to five years out of date, making it difficult to see the relationship between changes in spending and changes in EFIG allocations.

There is no way around the limited ability of the federal government to significantly reduce inequality of school funding, either by providing funding directly to school districts or by incentivizing state governments and local school districts to spend more, given the key role of state governments under American federalism. When spending is low in a high-poverty school district, it is often because the district is in a state where spending is low for all districts. To fully close these gaps, spending in low-spending states would need to increase substantially, often by 50 percent or more. That level of federal funding would prompt serious concerns about crowd-out, and state and local control of schools.

States with below-average fiscal capacity would get more federal funding, all else equal, and states with above-average fiscal capacity would get less (similar to how many states approach the distribution of state aid to school districts based on local fiscal capacity). The constant k and the amount appropriated determine the generosity of the program.⁴ Equation 2 is a template that could be modified based on policy goals. The count of “formula children” used could be based on children living in poverty and other disadvantaged groups (e.g., Title I's Basic Grant), or it could be a weighted count of such children to direct more funding to districts with greater concentrations of poverty (e.g., Title I's Targeted Grant). Similarly, funds could be awarded as a nonlinear function of relative per capita income. If policy makers want to provide more funding where input costs are higher, they could incorporate an adjustment for price differences in the formula as well using the CWIFT.

Finally, equitable funding alone will not create equitable American schools. The federal government is uniquely positioned to invest in research and development (which could cover a broad range of topics from curriculum and instructional practices to the effects of state school finance systems), provide public goods (like open-source high-quality instructional materials), and enforce civil rights laws.

The authors have received compensation from All4Ed for consulting services delivering two reports on Title I. Some findings in the brief, as explicitly attributed, were developed for the report Title I of ESEA: How the Formulas Work, published by All4Ed on 24 January 2023.

We are grateful to Emily Pelles for excellent research assistance and to Phillip Lovell, Anne Hyslop, and Rebeca Shackleford for helpful discussions.