Data Sources
To create the school district dataset, EdBuild used the following data sources:
• School district revenues: Revenues from federal, state, and local sources for the 2016-17 school year come from the Census, Annual Survey of School System Finances (F33). The following subtractions were made from total state and local revenues for each school district:
1. Because it can contribute to large fluctuations in district revenues from year to year, we exclude revenue for capital from the calculation of state revenues. 2. Similarly, we exclude money generated from the sale of property from local revenues, because it too can contribute to large fluctuations in revenues. 3. In just under 2,000 districts, revenues received by local school districts include monies that are passed through to charter schools that are not a part of the local school district but are instead operated by charter local education agencies (charter LEAs). This artificially inflates the revenues in these local school districts, because they include money for students educated outside of the district who are not counted in enrollment totals. To address this, we subtract from state and local revenues a proportional share (based on the percent of each districts’ revenues that come from local, state, and federal sources) of the total amount of money sent to outside charter LEAs—an expenditure category included in the F33 survey. 4. In Arkansas, large portions of districts’ revenues that should be considered local are categorized as state revenues. The value of this misattribution for each district is described in the F33 documentation as C24, Census state, NCES local revenue. Before analysis, the value of C24 is subtracted from state revenues and added to local revenues for the state of Arkansas. 5. In Texas, many districts report exorbitantly high per-pupil revenues. This is in part because of the policy and procedures for recapturing and redistributing local revenues raised by property-wealthy districts in the state. In the F33 survey, recapture is reported as expenditure code L12. Because these monies are included in the state revenue for other, receiving districts, we subtract a districts’ L12 expenditures from their local revenues for the state of Texas.
• School district enrollments, racial compositions, counties: School district enrollment characteristics as well as county assignments for the 2016-17 school year come from the US Department of Education, National Center for Education Statistics, Common Core of Data (CCD). Delaware, the District of Columbia, Massachusetts, and Tennessee did not report data on the number of students eligible to receive free- and reduced- price in 2017. For these states, we used the most recent year of data available for the number of students eligible to receive free- and reduced-price lunch-- the 2015-16 school year for Delaware, the District of Columbia, and Tennessee and the 2014-15 school year for Massachusetts.
• School district school-age poverty rates: School district-level data on poverty rates among relevant school-age children in 2017 come from the Census, Small Area Income and Poverty Estimates (SAIPE).
• School district geography: geography for school district borders for the 2017-18 school year come from the US Census Bureau, Education Demographic and Geographic Estimates Program (EDGE), Education Demographic and Geographic Estimates (EDGE).
• School district community indicators: school district-level data on median household income and median property value for the 2016-17 school year come from the US Department of Education, National Center for Education Statistics, Education Demographic and Geographic Estimates (EDGE)
Project Data
EdBuild employed several exclusion criteria in compiling our dataset. Our analysis includes only districts that meet our standard requirements for a geography-based analysis. Therefore any district that does not have geography and is not included in the Composite School District Boundaries File was excluded. EdBuild also excluded any districts from the US territories. Districts that did not have revenues reported in the F33 dataset were also excluded. Additionally, since Act 46 was in process in 2017, all districts in Vermont were excluded from the analysis as their revenues and district borders could not easily be determined.
Methodology
For the purposes of modelling the impact of pooling local revenue, EdBuild pooled local revenue at two levels: for all districts in each county, and for all districts in each state. Districts in each county were determined using the CCD county designations. For county-level pooling, we added the local revenue for all districts in each county to calculate county local revenue. We then added the students in each county to calculate county enrollment. To calculate county-level local revenue per pupil, we divided county local revenue by county enrollment, thus distributing local revenue evenly across all students in any given county. For state-level pooling we added the local revenue of all districts in each state to calculate state local revenue and added all the students in each state to find state enrollment. The division of these evenly distributed revenue across all students and created the state-level local revenue per pupil.
EdBuild pooled local revenue at both the county and the state level and calculated the following:
1. County-level pooled local revenue per pupil: the new local revenue each district in each county would receive from revenue pooling 2. State-level pooled local revenue per pupil: the new local revenue each district in each state would receive from revenue pooling 3. County-level revenue difference: the difference between a district’s current local revenue per pupil and the county-level pooled local revenue per pupil 4. State-level revenue difference: the difference between a district’s current local revenue per pupil and the state-level pooled local revenue per pupil 5. County-level outcome: based on the county-level revenue difference, a district was classified as receiving equal or greater funding or less funding 6. State-level outcome: based on the state-level revenue difference, a district was classified as receiving equal or greater funding or less funding
For each county, each state, and the country, the following was calculated:
1. Percent of students with equal or greater funding based on the county-level outcome 2. Percent of students with equal or greater funding based on the state-level outcome 3. Percent of nonwhite students with equal of greater funding based on the county-level outcome 4. Percent of nonwhite students with equal or greater funding based on the state-level outcome 5. Percent of students eligible for free-and reduced-price lunch based on the county-level outcome 6. Percent of students eligible for free-and reduced-price lunch based on the state-level outcome
To determine the most beneficial pooling method in every state, the percent of students with equal or greater funding, the percent of nonwhite students with equal or greater funding and the percent of students eligible for free- and reduced- price lunch with equal or greater funding were summed for both county level pooling and for the state level pooling. A state was assigned either state or county level pooling based on which sum was larger. For example, in Kansas under a county pooling system 65% of all students, 71% of FRL students and 77% of nonwhite students would get equal or greater funding while under a state level system, 62% of all students, 72% of FRL students and 71% of nonwhite students would get equal or greater funding. Kansas was assigned a county pooling method, because although FRL students fare better from a state-level system, the sum of these three groups is greater at the county-level.
Hawaii was assigned a state-level system as local revenue is already pooled state-wide. The District of Columbia is included in the analysis, treated as a state, and was also assigned a state-level pooling system. Further, because of uncertainties in local revenue allocation stemming from Texas’ current local revenue recapture system, we pooled Texas revenue at the state level.
Neighbor pooling
County and state pooling are two options for expanding the tax bases to increase equity. EdBuild also modeled pooling local revenue among school district neighbors as an additional alternative. Neighbors were identified by shared school district borders—see our methodology of Dismissed for more information on identifying neighbors. For each school district we added the local revenue of the district and the local revenue of all of the school districts with which it shares a border to calculate neighbors local revenue. We then added the students in the district and the students of each of the district’s neighbors to calculate neighbors enrollment. To calculate neighbor-level local revenue per pupil, we divided neighbors local revenue by neighbors enrollment, thus distributing local revenue evenly across all students in any given neighborhood.
From this we found the following for each district:
1. Neighbor-level pooled local revenue per pupil: the new local revenue per pupil each district would receive from neighbor-level revenue pooling 2. Neighbor-level local revenue difference: the difference between a district's current local revenue per pupil and the neighbor-level pooled local revenue per pupil 3. Neighbor-level outcome: based on the neighbor-level local revenue difference, a district was classified as receiving equal or greater funding or less funding 4. Percent of all students within each neighbor pool with equal or greater funding based on the neighbor-level outcome 5. Percent of nonwhite students within each neighbor pool with equal or greater funding based on the neighbor-level outcome 6. Percent of FRL students within each neighbor pool with equal or greater funding based on the neighbor-level outcome
Neighbor pooling was not an option as a state’s most beneficial method as each neighbor pool is mutually exclusive. Since it is assumed that each neighbor of any given district is sharing its revenue with that district, the neighbor cannot also be sharing its revenue with another neighbor.
Analysis
EdBuild calculated the following at the three pooling levels modeled in this report: county-level, state-level, and neighbor-level:
• Average student revenue increase for all students whose local revenues are equal or greater under each pooling system • Average student revenue increase for all nonwhite students whose local revenues are equal or greater under each pooling system • Average student revenue increase for all students eligible for free- and reduced- price lunch whose local revenues are equal or greater under each pooling system
$23 billion
To determine the amount by which the $23 billion funding gap between predominately nonwhite and predominately white school districts was closed, the most beneficial pooling method for each state was applied. The analysis done in $23 billion was duplicated using 2016 pooled local revenues.