RALEIGH – The issue of the optimal size of school districts remains a salient one in several North Carolina communities. Some disaffected parents and taxpayers in Mecklenburg and Wake counties would still like to see those huge school systems broken up into more manageable and accountable parts. On the other hand, there is still a spirited debate going on in Orange County about whether to merge the county system and the Chapel Hill-Carrboro Schools, while in Davidson County three different school districts remain – the county, Lexington, and Thomasville – as does strong sentiment not to merge the high-poverty city systems into a unified Davidson system.
The financial argument for school-district mergers has a long pedigree. That doesn’t make it valid. Eliminating “redundant” administrative functions is supposed to result in lower per-pupil costs, which can either be returned to taxpayers as tax cuts or invested in better-quality education. Unfortunately, there isn’t a lot of empirical support for this proposition. Theoretically, one can see the potential for savings, but also in theory merger reduces healthy competition for residents and students while also pushing the governance of the system further away from local neighborhoods and communities.
On another theoretical question – do merged systems produce better or worse outcomes for students? – the good news is that some useful research has recently been published. The bad news for advocates of school mergers is that the research appears persuasively to show that the larger the school district, the less effective it is at graduating students from high school.
The study was written by Dr. Jay Greene and Marcus Winters, both researchers affiliated with the Manhattan Institute for Policy Research. Greene and Winter assembled a data set that included graduation rates (correctly computed), per-pupil spending, the average size of school districts in each state (measured in square miles), and other relevant variables. They then used a fixed-effects regression equation to evaluate the effect of district size on graduation rates holding other potential causes constant.
The study offered several findings that should be of interest to North Carolina policymakers. First of all, there is a wide variation among states in average size of school districts. When you hear stories from Northeastern transplants in North Carolina cities about how many school districts they had choice of in their former abodes, believe it. The average North Carolina school district encompasses 406 square miles. The average district in New Jersey, with a comparable population, is only 12 square miles. Among nearby, Southeastern states, only Florida at 800 square miles has a higher average size. The national median is 260.
Greene and Winters found that state with larger districts had lower graduation rates. The result was statistically significant but not overwhelming in magnitude. If I’m reading their findings correctly, it appears that if North Carolina reduced its district size to the national median – by deconsolidating large urban systems – it might improve our graduation rate by a little less than two points. A radical downsizing to the level of New Jersey might drive the graduation rate up another couple of points.
The authors read their results more broadly to make the case for choice and competition in education. I share their inclination but not necessarily the conclusion, given other potential explanations for the effect (such as less strife over student reassignment or better administrative and financial control). Still, the case for North Carolina’s big, unwieldy urban school systems just got weaker – and it was already pretty shaky.
Hood is president of the John Locke Foundation.