Dispatches From The Nexus Of Boring And Important
School finance is one of those education policy topics located at the extreme ends of the important continuum as well as the boring continuum. On the one hand, school funding is relevant to virtually all major education policy decisions at the state-, district-, and school levels - at least in the background, but usually in the foreground. And the finance research literature is increasingly clear that there is a causal relationship between increased and/or progressive funding and better student outcomes (e.g., Jackson 2018; Baker 2016).
And yet, on the other hand, school finance is probably among the least sexy topics in our public education discourse, in part because the money behind policies is never as exciting as the policies themselves, but also because finance research is complicated and esoteric, and reading the research sometimes feels like reading audited financial statements.
Yesterday, the Shanker Institute, in collaboration with Bruce Baker and Mark Weber from Rutgers, released a new report and public dataset on school finance in the U.S.
It's still not sexy. Just to make sure, we called it the School Finance Indicators Database.
But we did try to make it more accessible and useful to the general public than the typical finance fare. The report presents key findings from the database, specifically state-by-state results on three “core” indicators: fiscal effort, adequacy, and progressivity. We feel that these three indicators provide a pretty good summary of states’ school finance systems. Rather than going through the report’s findings, here are a few things to keep in mind when reading it.
The report expresses strong opinions about how to interpret our measures. Specifically, we assert that school funding should be progressive – that is, districts serving large proportions of disadvantaged students should receive more funding than districts serving fewer disadvantaged students. And we argue, as a principle, that increasing funding, particularly for high-needs districts, is a desirable policy goal in virtually all states, though we acknowledge that some states need to increase spending more than others. These principles are based on our reading of the available evidence, perhaps mixed with our pre-existing values regarding the proper role of government and other issues. But you needn’t agree with these principles to use our report or the data.
We don’t advocate for how money should be spent. You might be in favor of charter schools and performance pay for teachers, or you might advocate for reducing class sizes and higher salaries across the board. The question of how best to spend money on schools is enormously complex, and beyond the scope of this project. We do argue that virtually all policy options that might improve school performance have one thing in common: They all require investment. Yes, how school districts spend money is very important, and there are certainly ways for states and districts to be more efficient with their money. But the money has to be there in the first place, and real, sustained improvement is not and will never be a low-cost endeavor, particularly for districts serving large populations of disadvantaged students.
We don’t assign ratings or grades to states. This is because state school finance systems are so complicated and context-dependent. For example, states like Alabama and Mississippi do poorly on our adequacy measures in part by choice, but also because they serve relatively large proportions of high-needs students (who are more expensive to educate) and lack the economic capacity to raise a lot of revenue. Conversely, states like Alaska and Wyoming do very well on our indicators, but they are in a better position to fund their schools due to revenue from natural resources. Reducing all this context down to a summative rating, or even a small group of ratings, would effectively hide these important considerations. It would also require somewhat arbitrary decisions about where to set the thresholds.
Above all else, we want everyone – researchers, journalists, parents, policymakers, and the general public – to use the data. The three “core” indicators covered in our report are only part of the database, which is freely available to the public (including in Excel format), so you can replicate the analyses presented in the report or do your own analyses. Our primary product is the State Indicators Database, a collection of about 130 variables for each state between 1997 and 2016 (the latest year available). All of the data in the report are from this dataset, and it includes indicators not only for revenue and spending, but also variables pertaining to how money might be spent, such as those measuring teacher salaries, class sizes, early childhood education coverage, and others. More advanced users can also download our District Indicators Database, a much larger and more detailed district-by-district set of variables, many of which are used to construct the state indicators.
In the documentation for the State Indicators Database, we have tried to describe the variables in a manner that is accessible to non-researchers (users can also download a “reduced” version of the State Indicators Database, which includes only the 30 or so variables presented in the report). We have also published a few simple visualization tools for quick looks at the data. In short, the primary purpose of this project is to provide high-quality and accessible data and analyses to inform and improve the public debate on school funding.
We hope the dataset is simple to use, but the measures are not simplistic. In reading our report and using the state dataset, you may notice that virtually none of the indicators are raw, unadjusted measures, such as simple per-pupil spending or revenue. Comparing unadjusted per-pupil spending is like comparing monthly rent between two families, one living in downtown Manhattan and the other in Peoria, Illinois. Without context, this will tell you nothing about the quality of the housing, whether it suits each family’s needs, whether they pay a lot or a little, etc.
The same goes for school finance. For instance, districts serving more disadvantaged students will have to spend more to achieve the same quality of education as districts serving fewer poor students. Similarly, districts located in more expensive labor markets will need to spend more (e.g., on teacher salaries) than districts in less expensive labor markets.
Accordingly, most of our measures are statistically adjusted for factors that affect the value of the education dollar, including poverty (from the Census), labor market costs, population density, and district size. That is, when you use our measures to compare states, you’re comparing outcomes (e.g., per-pupil spending or revenue) between similar districts within those states. You can also compare resources between high-poverty and low-poverty districts within a state.
The database also includes an adequacy measure that, adjusting for many of the same factors, compares actual spending in each state, by district poverty level, with the estimated amount that would be required for students in districts at that poverty level to achieve national average test scores.
Finally, please let us know what you think. This is an ongoing project. The dataset and report will be updated annually. We are committed to making this project as useful as possible to both researchers and non-researchers, and we strongly encourage questions, comments, and suggestions about how to do that. Please get in touch.