This Urban Institute blog post by Kristin Blagg and Matt Chingos is the latest example of DC schools data being used to perpetuate a narrative of progress, while masking growing race and class inequities. In this study, “Does Gentrification Explain Rising Test Scores,” the authors’ definitive answer is probably not: “The bottom line is that gentrification alone cannot explain why student scores improved in Washington, DC, a conclusion that echoes previous analyses using publicly available data,” they write.
To make this claim, the authors calculated how much race correlated with rising test scores. They found the change in racial demographics did not match the rise in test scores. Thus, they concluded that gentrification alone did not cause the change in test scores. (Almost as an aside, the authors note that their analysis does not include income as demographic factor and that racial gaps remain).
Gentrification is an infusion of wealthy, often white people into an urban space, and the displacement of poor people and families–most of them black and brown–a complex interplay of race and class. I cannot imagine any serious quantitative analysis of gentrification excluding income. Full stop. This is a fatal error. Income is the primary definition and driver of gentrification! When I tweeted this question to one of the authors, he tweeted me a link to a chart that showed that in the period of the study, DC child poverty rose from 27 percent to 30 percent, the period that saw a rise in school test scores. He also said they found similar results when they controlled for parent education levels. This still does not settle the issue for reasons I will explain:
DC schools are severely segregated along lines of race and class. Any claims about what is or what isn’t affecting school achievement and amid changing demographics must by definition take that reality into account. An accurate look at the interplay of race and class on test scores would have to examine scores from schools where predominantly white and wealthy students are enrolled, as well as the few schools where their numbers are rising and trace the trajectory of the scores there. Then, it must look at the overwhelming majority of public and charter schools whose population is mostly black and mostly poor and trace the trajectory of the scores and enrollment patterns at those schools.
In D.C., we fund more than 60 different school districts, each with their own budgets, boards and policies and approaches to “reform.” There is nothing consistent about D.C. schools across the board except for one thing: the more white children, the more resources. This is largely due to segregation and because there is virtually no white child poverty in D.C. That said, given the sizeable black middle class in D.C., “black” does not necessarily mean poor and low parent education levels. And some high-poverty schools the (larger charter chains especially) have formidable private fundraising arms that are closing those resource gaps. One other other end of the spectrum, unlucky schools that have no libraries and are left to die on the vine. So it’s complicated. Much more complicated than the prevailing narratives suggest.
School segregation unfortunately makes taking a comprehensive look at policy intersections with race and class relatively straightforward. As a DC taxpayer watching this train wreck go by year after year, I really wish more economists did more work in this direction. I am careful not to make any empirical claims in the other direction of the Urban Institute. But it really epitomizes the poor quality of research that is driving the conversation on progress in D.C. schools. That in turn leads influential commentators like Matt Yglesias to make the truly magical leap to pronounce that “D.C’s controversial school reform is working.” Given the fast-widening gaps between the overwhelming majority of students enrolled who are black, and relatively small number of white students, that is sort of like saying, “D.C.’s controversial School reform is working because it is working for white people.”
To this, I will say empirically: It is not.