Mayors and Test Scores
To the Editor:
It goes without saying: The long-term goal of school reform is the improvement of student performance. The best, if imperfect, way to measure academic performance is through standardized tests. Drawing a direct connection between particular interventions and educational outcomes, however, is a hazardous endeavor. That is especially true with governance arrangements.
Thus, any reader who came away from the essay by Kenneth K. Wong and Francis X. Shen in your Oct. 14, 2009, special pullout “Leading for Learning” with the impression that the implementation of mayoral control in city schools is likely to elevate test scores needs to take a closer look at the evidence.
Kenneth Wong is a well-regarded education researcher with whom I have had the pleasure to collaborate on more than one occasion. In his 2007 book The Education Mayor, written with Mr. Shen and others, and on which their Education Week essay was based, they collected mounds of data from thousands of schools to find a correlation between school governance and student performance. These are interesting findings that invite further inquiry. But these results need to be interpreted with caution.
Unfortunately, despite heroic attempts to fine-tune their methodology, Messrs. Wong and Shen overstate the evidence that governance hikes achievement, and their comparisons are problematic. Notwithstanding the close attention they paid to reconcile state data from more than 100 districts over a four-year period (1999-2003), there remain too many intervening variables at play among districts that had operated under mayoral control for different lengths of time. As we learn in basic statistics, correlation is not the same as causation.
First, it is questionable whether Baltimore, Oakland, Calif., and Philadelphia should have been counted among the mayor-led districts, since their mayors shared authority and had relatively little power over schools. New York City, also counted among the mayoral cities, did not practically implement mayoral control until the end of 2002.
Messrs. Wong and Shen also ignore data from the National Assessment of Educational Progress indicating that there is no correlation between mayoral control and either better or improved student performance. The slight advantage that they observed in state-test results is overshadowed by a lack of progress in closing the achievement gap between minority and white students. This doesn’t mean that mayoral control of schools is a bad idea. It just means that there is insufficient evidence to build an argument for it on the basis of student performance.
New York is a case in point. In 2007, as it was considering the renewal of mayoral control for the city, the leadership of the state Assembly asked the city’s public advocate to appoint a Commission on School Governance to study the issue, for which I served as executive director. Members of the panel were well aware of rising scores on state reading and math tests that coincided with mayoral control. At my urging, the panel decided not to speculate on whether these improvements had resulted from the governance change, or, more importantly, whether mayoral control was responsible for any real improvement in student achievement.
City scores on the NAEP tests had remained relatively flat. Urban school districts throughout the state that did not have mayoral control also were recording improvements on the state tests. Under mayoral control, state and local spending on New York City schools had increased by more than 40 percent. Under mayoral control schools became a higher priority. As in many other cities that have adopted mayoral control, implementation itself was an indication that education had already become a higher public priority at the state and local levels.
In the end, the commission recommended the renewal of mayoral control with some changes, most of which were adopted by the state legislature. Like Messrs. Wong and Shen, we pointed to the direct line of accountability under mayoral control that was previously absent. Based on our own assessment of the experience with mayoral control in New York City, and on a series of papers we solicited from leading experts who had studied mayoral control in cities around the country (including Mr. Wong), we concluded that the major advantage of mayoral control is that it increases the institutional capacity for change. That is no mean accomplishment. That capacity increases with the amount of authority given to a local chief executive, but so do the risks and the need for checks on that very power.
Mayoral control needs to be implemented cautiously. The growing experience with it needs to be assessed carefully. We should not jump to conclusions.
Vol. 29, Issue 10, Page 27
Vol. 29, Issue 10, Page 27
Get more stories and free e-newsletters!