Brookings has another excellent and useful study out this week. This one examines how much superintendents, on average, contribute to student learning. The authors, Matt Chingos, Russ Whitehurst, and Katharine Lindqiust, analyze student level data in Florida and North Carolina between 2001 and 2010 to see how much variance in achievement can be explained by changes in superintendents. The answer is not very much — only .3%. Other aspects of the school system, including the student, teacher, school, and district matter much more in explaining the variance in student achievement.
The authors are careful to explain that their research does not suggest that there are no dud or superstar superintendents. It’s just that, on average, superintendents don’t make much of a difference. They liken this to the effect of money managers who on average add no value, although it is possible that some of them are great and some awful. Of course, much or all of that difference between great and awful could be random chance. So when you pick a superintendent (or a money manager) you should rationally expect that they don’t make much of a difference. It’s a shame that they still cost so much.
This report helps illustrate how Brookings is really the model of what think tanks should be. It is solid empirical work on a policy relevant question that is written in a way that is accessible to policymakers and other non-experts. Other think tanks would do well to consider how they could emulate Brookings rather than produce more agenda-driven hatchet job research. And more foundations should think about how they could fund this type of quality, policy-relevant work and stop paying for talking points masquerading as research.
Jay P. Greene's Blog 
Of course this doesn’t prove that superintendents can’t matter, only that they don’t currently matter. We have designed the system in a certain way, a way that tends toward the enrichment of central bureaucrats who have large and ever-expanding empires but contribute little value on the ground. As long as the system remains designed that way, the results will be largely the same. But that very fact will create pressure for change, and I expect that a generation from now superintendents will matter because the old system will have been successfully challenged in at least some big ways.
[…] at Jay Greene’s blog, Matt Ladner succinctly clarifies what the Brookings report says: The authors are careful to explain that their research does not suggest that there are no dud or […]