moving to opportunity and neighborhood effects: some sociological background
The New York Times posted a big feature yesterday on a couple of new papers by Harvard economists Raj Chetty, Nathaniel Hendren, and Lawrence Katz. The papers, like much of Chetty’s other work, use deidentified individual-level tax data to get at factors affecting income over time. In this case, they are interested in getting at neighborhood effects—in one paper, county-level effects on intergenerational income change, and in the other, the effects of the Moving to Opportunity experiment, which in the 1990s provided housing vouchers through a lottery system, on the same.
The big findings are that moving to a better neighborhood improves children’s income as adults, with the effects being cumulative. The experimental MTO data shows that each additional year of residence in the new neighborhood contributes linearly to an increase in adult income. However, the effects of a move zero out around age 13, after which they may be negative. The quasi-experimental data on non-MTO moves, which cleverly compares different-age siblings to get at length of exposure to the new neighborhood, points to substantial variation in mobility across counties for children at various income levels. The NYT visualization of this latter data, which is personalized by your location, is really terrific.
It’s some impressive work. But sociologists, of course, have been studying neighborhood effects for a long time. And while there is a lot of interest in the study, there is also a not-totally-unjustified sense of annoyance:
Big News: Economists discover a thing sociologists have been studying in detail for about a century. https://t.co/AxVkIfqsyw
— Mervyn Horgan (@simmelian) May 4, 2015
I’m fascinated both by the studies, sociologists’ reaction to them, and how the research is picked up and interpreted by the media. I tried to put all those things into one post, but it was getting way too long. So I’m going to break my reactions into a couple of chunks. Today, I’ll highlight some of the excellent existing work by sociologists in this area. Tomorrow, I’ll comment on why Chetty and Hendren’s work is getting so much more attention than related work in sociology. And later this week I’ll address how this kind of research gets covered in the media and is likely to be translated into policy conversations.
So for starters, some pointers to the sociology literature.
The main reason sociologists were annoyed by all this was the tone of the NYT coverage. It portrayed the Chetty et al. study as a huge breakthrough: “a large new study is about to overturn the findings of Moving to Opportunity,” which it called “deeply disappointing…haunt[ing] social scientists and policy makers.” While the article does quote sociologist David Grusky, who notes that the effects of neighborhoods on children is something “which social science has been grappling for decades,” the general tone is that these new studies are a massive advance; a qualitatively different, definitive piece of research that puts to bed old doubts and, as Chetty himself puts it, “shows we can do something about upward mobility.”
This is the fault of the NYT article, not Chetty et al. While economists are, often fairly, criticized for ignoring work in other social sciences, that can’t be said about these papers. They cite a ton of sociology. Bill Wilson, Rob Sampson, Patrick Sharkey, and Jacob Faber make the first paragraph of one; Susan Clampet-Lundquist, Doug Massey, Kyle Crowder, Scott South, and Geoffrey Wodtke are on the first page of the other. These two articles have possibly the highest percentage of sociology citations I’ve ever seen in an economics paper. So, credit where credit (for giving credit?) is due. It’s also worth noting that Justin Wolfers’ companion piece in the Times on why these studies matter more directly acknowledges past work in sociology.
Nevertheless, this is as good a place as any to call out some of that work on neighborhood effects by sociologists, which has been going on for 20+ years or a century, depending on how you want to count. This is not my area, so I am no doubt overlooking important research. Feel free to add pointers to additional work in the comments.
There are numerous good reviews of research on neighborhood effects. A recent Annual Review of Sociology piece by Sharkey & Faber emphasizes the heterogeneity of neighborhood effects, and the need to focus on the specifics of how, when, and for whom neighborhoods matter. A 2002 ARS by Sampson, Morenoff & Gannon-Rowley covers the first wave of work from the 1990s.
Chetty and Hendren’s work builds most directly on relatively recent research that conceptualizes neighborhoods as having cumulative effects over time. The Sharkey & Faber review gets at this, but see in particular Crowder & South’s 2011 Social Science Research article on how duration of exposure to a disadvantaged neighborhood affects high school graduation, and in particular Geoffrey Wodtke and colleagues‘ recent work, which uses PSID data to look at the cumulative effects of exposure on teen parenthood and high school graduation.
A lot of the Moving to Opportunity evaluations were done by economists—the big evaluation came out in 2011, though more studies have since been produced—but AJS had a 2008 symposium on the study that’s worth a read, featuring pieces by Clampet-Lundquist & Massey, Ludwig et al., and Robert Sampson.
Finally, I would be remiss if I didn’t mention at least some of the interview-based and ethnographic work on neighborhoods that deeply informs sociology—including the research of Mario Small, Xavier de Souza Briggs and colleagues, and David Harding, among many others. Small and Jessica Feldman have a nice piece that gets at some of the epistemological debates underlying the neighborhood effects literature, and that lays out some ways ethnographers might contribute directly to it.
Tomorrow: Why this work hasn’t received the attention Chetty et al. are getting, and how it might get more.