Urban schools don’t inspire much confidence these days. Politicians and policy leaders routinely bemoan their quality. And media outlets regularly run stories of “failing urban schools.”
Middle- and upper-income parents have expressed misgivings, too. But they’ve done it much less volubly. With relatively little fuss, they’ve simply picked up and moved—departing from city school systems at ever-greater rates. Among expressions of no-confidence, this has arguably been the most significant, because it has reshaped district demography. Each year, it seems, urban schools serve larger concentrations of poor students, racial minorities, and English-language learners. As higher-income families depart, resources go with them, and schools are faced with the daunting prospect of doing more with less.
If such departures are driven by good information about school quality, one can hardly blame parents with resources for acting in the best interests of their children.
Yet what if the information people are acting on is inaccurate or misleading?
Thanks largely to No Child Left Behind, the public has access to performance data for all public elementary and high schools. The data collected and reported, however, largely consist of student standardized test scores. As George W. Bush, who as president signed the act into law, put it, “We measure. We post the scores. We look at results.” Today, over 15 years after NCLB first went into effect, test scores are commonly used—by policy leaders, parents, and the general public—as a measure of school quality, often in the total absence of other information. A New York Times feature, for instance, produced a set of charts for prospective suburban homebuyers using only two inputs: “home price data from Redfin … and school quality data based on test scores.”