John Hood: School size may not matter, but innovation does
RALEIGH — Because of the tremendous benefits conferred by better education, it would be great if policymakers knew precisely what silver bullets to fire to eliminate obstacles to higher achievement.
But obstacles to achievement aren’t werewolves. They are complex and deeply rooted, not simple and fictional. Unfortunately, far too many political discussions of education reform turn into debates about how best to cast silver bullets.
Some years ago, the magical ammunition of choice was the idea of making schools smaller. It attracted media attention, foundation funding and political momentum. There was some early research support for the cause, which also seemed like common sense: smaller schools would be easier to manage and differentiate, allowing more innovation and individualized instruction.
Like so many other education-reform fads, however, policymakers let a few successful cases and a plausible theory displace critical thinking and patient evaluation. They didn’t look before they leaped. States and districts across the country hurried to break up existing schools and found new, smaller ones. The results proved to be mixed and, in many cases, disappointing.
It turns out that, yes, some students thrive in smaller schools. They feel safer and receive more attention. Some principals and teachers also do their best work in smaller schools. But for other students and educators, the tradeoffs aren’t to their advantage. Smaller schools may lack the scale necessary to justify the high-level courses that challenge the gifted or the extracurricular programs that keep some students excited about school.
If policymakers had studied the issue more carefully, they wouldn’t have greeted the small-school movement with unrealistic expectations. Over the past quarter-century, scholars have published some 100 peer-reviewed studies exploring the relationship between school size and student performance. In a third of the studies, smaller schools were associated with higher achievement. In half of them, there was no statistically significant relationship between school size and outcomes. In the rest of the studies, smaller schools were linked to lower student achievement.
Importantly, these findings do not suggest that there are no benefits to be had from smaller schools. As I said, they may be precisely what the doctor ordered for some students. But policymakers shouldn’t expect a general program of downsizing public schools to result in substantial improvements in the efficiency or effectiveness of those schools. Other factors play a larger role in shaping student achievement.
For small schools that do have a strong track record of success, size may be only part of a larger bundle of characteristics — a shared vision, a dynamic leader, rigorous curriculum, community support — that can’t easily be replicated in other settings by other people.
Indeed, some of the studies finding benefits from small schools are really finding benefits from new, innovative schools — those that naturally begin their existence with small enrollments but then retain their effectiveness even as they grow. A 2013 study in the Journal of Urban Economics, for example, found that when New York City broke up existing high schools into smaller units, there wasn’t necessarily any effect on graduation rates or other outcome measures, whereas newly founded small schools did outperform other high schools in the city.
Policymakers out to improve education should abandon the search for a silver bullet and approach the problem differently. After setting rigorous standards and ensuring that there will be independent assessments of student performance to provide critical information to parents, educators and taxpayers, they should then largely get out of the way and let districts, schools and educators make their own choices about operational matters.
It’s best to think about school reform as a process of discovery and replication, not one of social engineering and bureaucratic compliance.
Unlike small schools, school autonomy does offer consistent benefits. About two-thirds of peer-reviewed studies find a statistically significant link between the degree of school autonomy and student outcomes. Some autonomous schools — public and private — choose to keep total enrollments low. Others don’t. As long as they get results, who cares?
Hood is chairman of the John Locke Foundation.