Schools across the country spend billions of dollars each year on the construction and renovation of their facilities—everything from roof repairs to new science labs and from classroom expansions to whole new buildings. How do these expenditures impact students and the taxpaying community around schools? A recent report goes into fine-grained detail to try and answer those questions.
In most states, capital outlays for school construction and repair are primarily funded using bonds issued by each school district. Roughly three-fourths of those outlays are funded locally, with state aid kicking in less than 30 percent of funds for capital projects on average, and less than 5 percent in about half of states. Again, local funds are generally raised from bond sales, which must be approved by voters in local referenda usually held during a primary or general election. If passed, those bonds are eventually repaid with revenues from local property taxes. And while we know that state-level finance reforms have led to a more equal and progressive distribution of spending across districts, the distribution of capital outlays remains unequal. This study, by Barbara Biasi, Julien Lafortune, and David Schönholzer, identifies which investments in school facilities help students (via positive test score changes) and are valued by homeowners (via housing value increases).
To estimate the effect of bond passage, the analysts compile a dataset with information on school bond referenda (including the text of the ballot issue), student test scores, and house prices for twenty-nine states. Their dataset covers approximately 14,000 bond elections in those states and over 10,000 districts, which together enroll 71 percent of all students in the U.S. They use the SEDA database and information from state departments of education to gather district-level test score averages, ultimately developing a dataset with scores from 2003 to 2019 for nearly all states, and as early as 1994 for some states. They use House Price Index data at the Census tract level and map them to 2010 school district boundaries and then calculate the average house price index for each school district and year.
They set out to estimate the causal effect of bond authorization on scores and house prices. They aim to compare “treated” districts, meaning those that succeed in authorizing a bond, and “control” districts that propose a bond in the same year but fail to authorize it. But since school districts that succeed are different from those that don’t in unobservable ways, and since they want to avoid comparisons with districts that become treated at a later time, they exploit variation from close bond elections that had staggered timing. That is, they’re comparing districts that barely authorize a bond in a given year and cohort with districts that also propose a bond in the same cohort, but fail to authorize it, and in fact, never authorize any bonds in the future. They categorize spending into eight categories, including classroom construction and renovation, HVAC, technology/IT improvement or increases, athletic facilities, and more.
Results show that increased school capital spending raises both test scores and housing prices on average. Specifically, in districts that marginally approve a bond proposal (compared to those that marginally reject it), test scores are 0.04 standard deviation (SD) higher on average one to four years after a bond election; they are 0.08 SD higher five to eight years after; and 0.07 SD higher nine to twelve years after. Bond approval also generates increases in housing prices by 7 percent five to eight years after a bond election in the average district.
However, impacts vary widely across types of projects and types of districts. Spending on new basic infrastructure or upgrades to them (such as HVAC systems, plumbing, furnaces, or roofs) or on the removal of pollutants raises test scores but not house prices—possibly because these things are not “visible” to homeowners without school-age kids. Conversely, spending on athletic facilities or construction of new classroom space raises house prices but not test scores. (Smartly, the analysts also look at changes in student population following big, visible capital expenditures, ruling out the influence of gentrification.) Additionally, disadvantaged districts were found to completely drive the differences; they benefit much more from capital spending, in part because they’ve spent less on it in the past and typically request larger bond amounts. In wealthier districts, test scores remain unchanged post-bond election and the effect on house prices is non-significant.
All of this tracks with previous research and indeed with common sense. A leaky roof is a distraction from teaching and learning, and bad indoor air can impact numerous students with asthma and other ailments, but taxpayers without children in the district schools would have little awareness of such issues. The bottom line is that, while some capital improvements must be made, any discretionary improvements can and should be targeted to generate the highest potential achievement boosts for students.
SOURCE: Barbara Biasi, Julien M. Lafortune, and David Schönholzer, “What works and for whom: Effectiveness and efficiency of school capital investments across the U.S.,” National Bureau of Economic Research (January 2024).
