by Paul E. Peterson and Matthew M. Chingos
November 7, 2007
For-profit management of public schools is still in its infancy, and many wonder whether it can have a positive effect on student learning. In Philadelphia, that idea has been put to the test. The results, as we report in a paper issued last Friday by the Harvard Program on Education Policy and Governance, would not surprise Adam Smith.
The 18th-century economist explained that those who need to make a profit have strong incentives to do well by their customers. But can Smith's theory actually work when one is talking about educating students in the most challenging of urban schools -- at the very heart of a major metropolis? The answer appears to be yes.
When for-profit management of public schools was first proposed in Philadelphia six years ago, many in that city were extremely skeptical, if not aggressively hostile. So the Philadelphia School Reform Commission, the entity responsible for the innovation, gave only the 30 lowest performing schools to for-profit companies, while another 16 were given to nonprofit organizations, including two of the city's major universities (Temple and the University of Pennsylvania). Others were reorganized by the school district itself.
In effect, a competition was run among the three types of management -- for-profit, nonprofit, and government-run. Four years into the race, here are the results: Students at schools managed by for-profit firms were roughly six months ahead in math than would be expected had the schools remained in the hands of the school district. In reading, students in schools managed by for-profit firms were two months further along than they would have been if the schools had been under district control, though that difference was not large enough to give us statistical certainty. Meanwhile the nonprofits -- and the school district's own reorganized schools -- did no better than expected.
Our findings are based upon information gleaned from nearly 400,000 student test scores made available to us by the School District of Philadelphia. They gave us the test scores of every tested student for the years 2001 through 2006, allowing us to track student performance at for-profit, nonprofit and low-performing district schools both before and after the management changes took place.
That data was subjected to a rigorous, quasi-experimental, "difference in differences" analysis that estimates management impacts at each type of school by making use of information on how much students were learning both before and after the management change while controlling for the students' characteristics.
Though we believe our methodology to be state of the art, our findings will nonetheless be controversial, because they contradict a prior study by the RAND Corp. in February, which found no impact of private management on student performance. The RAND study, however, failed to separate out the schools managed by the for-profit firms from those managed by the nonprofit organizations. In our study, too, management effects are nil when the two are mixed together, as the positive impacts of for-profit firms are canceled out by the negative impacts of nonprofit organizations.
Even Adam Smith would not expect that for-profit management will work anytime, and anywhere. But the Philadelphia results demonstrate that putting schools in private hands could lead to improvements in education. At the very least, the current Philadelphia for-profit schools should be allowed to continue under private management. If results from the first four years continue to hold up, they make a strong case for giving the private sector a larger role in urban education.