This post is from Matthew Steinberg (University of Pennsylvania).
A recent trend in school reform efforts is the transfer of decision-making authority in many large, urban districts to individual schools, giving principals greater control over how they meet performance targets. In California, two of the state’s largest districts, San Francisco and Los Angeles, have shifted decision-making authority to a small but growing number of schools; similarly, Oakland has allowed some schools to opt out of many district-wide mandates. Beyond California’s borders, Boston, Chicago, Houston, New York City, Seattle, and St. Paul have implemented school-based autonomy programs in the last ten years.
These shifts in the locus of control are based on the common-sense notion that decision-making authority should be in the hands of those who actually deliver the education services to students. In other words, principals are in the best position to know how to support their students and teachers in order to maximize student learning. Specifics vary from district to district, but responsibilities transferred to schools generally fall into one or more of the following areas: budget; curriculum, instruction, and assessment; calendar and schedule; and teacher professional development.
Despite the growing emphasis on reform strategies that decentralize decision-making authority, we know little about how greater local autonomy impacts school performance. In Chicago, the administrator charged with running the program acknowledged its inherent gamble, noting that the district was “asking schools to take new risks with us.” To minimize the role of chance in these initiatives and nudge them into the realm of evidence-based reforms, policy makers need to see actual results. In particular, they need to see the impact of greater school autonomy on math and reading performance.
My study of the first two years of a school-based autonomy program in the Chicago Public Schools (CPS), which was recently published in the journal Education Finance and Policy, is an early step to providing such evidence. In the Chicago venture, participating principals chose from among ten autonomy options, which they could implement as they saw fit. I looked at 450 CPS elementary schools, of which 73 were granted greater autonomy. Leveraging the fact that school selection into the autonomy initiative was determined almost entirely by prior school achievement, I employed regression discontinuity methods to identify the effect of receiving greater school-based autonomy.
I found that the share of elementary students meeting state-determined reading proficiency standards increased after two years of autonomy. There were no significant improvements in the share of elementary students meeting math proficiency standards. As for the schools’ average achievement scores, there was no improvement in either math or reading. Among the autonomy options, principals were more likely to exercise autonomy over the school budget and curricular/instructional strategies than over professional development and the school’s calendar/schedule. Principals most often chose budgetary control, indicating that they saw it as the most critical component in the provision of autonomy over school operations.
Based on these findings, it is possible to sketch some modest policy recommendations. First, policy makers should not expect immediate gains. Instead, they need to allow time for school leaders to learn how best to implement their new autonomy. The gains in reading proficiency at CPS didn’t appear until the end of the second year. And although my empirical strategy limited my ability to look out past two years, the positive effect on reading proficiency rates after two years suggests that as principals are given time to learn how to use their resources and teachers become accustomed to organizational changes, school performance can improve over time.
Second, it’s important to be aware of the potential adverse impact on the lowest- and highest-achieving students if resources are focused on improving proficiency rates rather than average school performance. This is because, for the purposes of sanctions under NCLB legislation, schools are evaluated on the share of their students proficient in math and reading. As such, schools may allocate efforts to students at the margin of proficiency while reducing effort given to very high- and very low-achieving students.
While more work is necessary to better understand the impact of decentralization policies on school performance, we must also attend to the ways in which school leaders employ their new autonomy. A clearer understanding of these and other issues related to decentralized school control will shed light on whether such a policy offers promise for improving schools.
The full study can be found in Steinberg, Matthew P., “Does Greater Autonomy Improve School Performance? Evidence from a Regression Discontinuity Analysis in Chicago,” Education Finance and Policy, Winter 2014, Vol. 9, No. 1, Pages 1-35.