moving to opportunity and neighborhood effects: why the attention now?

So yesterday I offered some pointers to the neighborhood effects literature, which is relevant to the new research on social mobility that is receiving extensive coverage in the NYT and elsewhere. Commenter Robert Park (it’s good to know that earthly departure does not preclude keeping up with orgtheory) mentioned additional work worth highlighting (links added by me):

On the efficacy of residential mobility programs, I’d note the work of Rosenbaum and Rubinowitz “Crossing the Class and Color Lines” which studies the Gautreaux program, on which MTO was based, and more recently, “Climbing Mt. Laurel” by Doug Massey and colleagues and the work of Stefanie Deluca.

I know I have not mentioned many significant scholars, which I hope will be read not as a slight but as a reflection of how deep and rich this literature is in sociology.

But today I want to add a few comments on the question of why Chetty and Hendren’s work is getting so much attention when related work in sociology has not.

First, I want to be clear about something. It’s true that I’m a little obsessed with economists. But it’s more an intellectual interest than a how-come-they-get-all-the-attention interest. Professionally, I’m interested in how we organize the production and use of knowledge. My first book was about the natural sciences; I’m now studying economics, as the most influential of the social sciences.

“How come they get all the attention” is, though, a question that interests lots of sociologists (and management scholars, and political scientists, and education researchers…). I’ve written about the general success of economics as a discipline relative to sociology before. But here are a few thoughts on why it’s playing out this way in this particular case.

1. The straight science reason.

The papers are significant. They use a new and previously inaccessible type of data—individual tax records—to creatively test the effects of neighborhoods in a way that partially resolves a major puzzle around neighborhood effects: how could so much observational data show that neighborhoods matter, while the Moving to Opportunity experiment demonstrated no economic effects, despite some benefits to physical and mental health. The Chetty and Hendren study shows, for the first time, significant economic effects from MTO—and they’re long-term. Particularly since we’ve got an RCT fetish (too strong?) at the moment, this is a big contribution.

Still, it’s hard to imagine these papers getting the same kind of attention had they been published by leading sociologists in AJS. That leads us to…

2. The media reason.

Economists have become the go-to social scientists for the media on a range of issues, only some of which are strictly economic. This predates the whole Freakonomics phenomenon—now a decade old—as can be seen in this chart of NYT mentions of various disciplines.


But the creation of The Upshot—can you believe only a year ago?—is worth singling out as important. The Upshot was a replacement for FiveThirtyEight, but under the leadership of onetime economics writer David Leonhardt has become much more economic than FiveThirtyEight ever was. Not exclusively, of course—its core conceit is data analysis, and its contributors come from a relative range of disciplines—but it’s clearly tapped into what’s going on in the discipline of economics in a way it is not tapped into, for example, sociology.

Finally, the media timing is “good” with the recent protests in Baltimore. People are, at the moment, concerned about cities in a way they were not a year or two ago.

3. The policy infrastructure reason.

Economists are much better integrated into both politics and policy than sociology is, outside of a few specialized domains like the Census Bureau. The NYT mentions that Chetty “has presented the findings to members of the Obama administration, as well as to Hillary Rodham Clinton and Jeb Bush.”

Sociologists occasionally manage to be heard in elite policy circles—Obama’s take-up of Sara Goldrick-Rab’s free community college plan comes to mind—but this is much rarer and less routine.

But beyond the obvious benefits of being Washington players, or at least having access to the players, I’d point to a second advantage of being part of the policy world: data.

Part of the breakthrough of this study—as well as Emmanuel Saez and collaborators’ work on top incomes, which really put income inequality back on the map—comes from having access to IRS data. Individual-level tax data has only recently been made available, to the point where Science wrote a whole article on how the heck they got access to it. The article suggests Alan Krueger, then chief economist of the Treasury Department, was responsible for sparking IRS openness to academic research. I don’t know if that’s true, but being connected in Washington has opened up a huge new research terrain that isn’t yet accessible to sociologists (or most economists, for that matter).

Finally, linking back to the media reason, when economists don’t have access to government data, they have access to national platforms that allow them to call for it, as Susan Dynarski recently did with student loan data in (again) The Upshot.

4. The political reason.

This is the most subjective, and a little conspiratorial, but I think it’s real. Chetty & Hendren’s work is politically friendly. It’s about mobility, not inequality. Democrats, and especially Hillary Clinton, are too dependent on Wall Street and wealthy donors to really want to make an issue of inequality. Chetty’s work—both these studies and the earlier ones on intergenerational mobility by location—shift the focus back to the safer idea of helping people move up the economic ladder, rather than asking why so few have so much in the first place. Even the name of Chetty’s center—The Equality of Opportunity Project—points away from a focus on inequality.

And this new work on neighborhoods, in contrast to much work in sociology, implies an individualistic set of solutions. As Vox puts it, “Want to help poor kids? Help their parents move to a better neighborhood.” No one, you will notice, is talking about actually improving the bad ones.

I’ll leave the question that raises—how research gets picked up and translated into particular policy solutions—for another post.


Written by epopp

May 6, 2015 at 3:47 pm

5 Responses

Subscribe to comments with RSS.

  1. I work in this field and have read both the economists and sociologists which you have cited. I think you are missing a crucial distinction between the two fields: emphasis on statistical rigor. Very very few people, even within economics and more so in every other social science discipline, can come close to matching the degree of statistical sophistication of Chetty and his coauthors. Each discipline emphasizes different methodologies, and economists decided a very long time ago that they were going to be number one when it came to statistical analysis, at the expense of every other methodology. Robert Sampson is very sophisticated when it comes to empirical analysis, but I honestly do not think he would have been able to do what Chetty did even if he had had access to the data. Chetty et al. are beyond reproach when it comes to the pure empirical analysis, and economics as a whole has built its reputation around empirical research. The media eats it up. Being able to put a concrete number on the effect of spending on additional year in a particular neighborhood is a very powerful rhetorical device. Numbers with dollar signs attached to them are very simple to understand and resonate with both policy makers and every day people. They are misleading and often hide half the story, but that is how the attention game works. Do you disagree? I am interested in your take on this dimension. Thank you.


    Johan Uribe

    May 6, 2015 at 8:25 pm

  2. Thank you for this.

    I do think that chart is misleading, because the term “economist” is more common than “sociologist” for people with those disciplinary roles. I reran it with “professor of X” and “X professor” and it’s a little better for us:

    That said, the fact that “economist” means more than “sociologist” is of course part of what you’re talking about.

    Liked by 1 person

    Philip N. Cohen

    May 6, 2015 at 11:28 pm

  3. @Johan, thanks for the comment. I don’t disagree with you at all about economists being more methodologically sophisticated — certainly on average, and probably at the most elite levels (e.g. Sampson v. Chetty) as well. And this work in particular is impressive beyond its use of administrative data. I also agree that putting a number on spending a year in a neighborhood is a powerful attention-getting device.

    But I am less convinced it’s methodological sophistication itself that causes the attention, when there’s so much unsophisticated research that gets written about because it produces interesting numbers. Off the top of my head, I’m thinking about work on the payoff to various college majors, which is generally pretty weak, but makes headlines every time something new comes out. Or lots of stuff produced by think tanks.

    Now, I would buy that there’s a two-step process through which math skills get converted to disciplinary status which then gets converted to media attention. But given that the correlation between “high quality research” and “research that gets a lot of media attention” isn’t generally that high, and my read is that while the methods are impressive and go beyond existing work in this area, they’re not a quantum leap, I’m not sold on the idea that sophisticated methods directly explain the level of attention.



    May 7, 2015 at 12:53 am

  4. Interesting posts Fabio. I agree with epopp that the quality of the work shows no signs of being correlated with media attention, so I don’t think that’s it.

    One note though is that I’m puzzled by the trash talk about Sampson v Chetty as methodologists. For my money, Sampson (AJS 2008) still stands as far and away the best treatment of what field experiments (RTCs etc) can and cannot do and especially on the issue the central issue that Fabio raises– i.e,. that individual-level interventions (moving families across neighborhoods) is quite a different matter than neighborhood-level interventions (e.g., introducing a new approach to managing a particular social ill in a neighborhood). Note his opening on the two different kinds of neighborhood effects; this seems crucially important to keep in my mind (as do several important points that put the MTO experiment in context).

    Also, I just glanced at Chetty et al. I agree that they cite sociologists more than your average economist, but ahem, you might think that they would give Sampson more credit for basically giving them the framing of their paper. Here is a quote from Sampson 2008:

    “It is likewise important to appreciate the implications of the fact that at baseline, MTO adults and their children had for the most part grown up in high-poverty neighborhoods, which raises a developmental question about life-course timing and the durability of neighborhood effects. If the effect of disadvantage is cumulative, lagged, or most salient early in life, as recent evidence suggests with respect to cognitive ability and adolescent mental health (Wheaton and Clarke 2003; Sampson, Sharkey, and Raudenbush 2008), then moving, while still potentially important, does not bear on perhaps the most critical neighborhood influences in early child- hood. In this sense, the MTO experiment may be inconsistent with the- oretical perspectives that emphasize early brain development and critical periods when contextual effects get “locked in” (Shonkoff and Phillips 2000, chap. 8). ”

    As far as I can tell, this should have been the jumping-off point in the paper. (And it would still be missing several crucial points in Sampson 2008).

    I’m an outsider to this literature, and just glanced at this, so more than the usual caveats apply.



    May 8, 2015 at 1:18 am

  5. That chart is really interesting. If you’ll allow me to speculate a bit…

    It looks like economists seized prominence in the 1930s when they finally overtook historians as, at least by this metric, the leading policy analysts. And that’s no surprise really, since the great depression was not just a huge economic topic–hence the rise of economists–but a historically unprecedented one–hence the relative fall from grace for historians. It wasn’t the first depression–for the US the 1870s are considered the first recession–but it was historically unprecedented in its depth, length, and impact since for the first time most of the population lived in cities and couldn’t subsistence-farm as in the past. And that might be the key to why economists have been so dominant ever since–the movement to cities has only continued (80% of US lives in an urban area now), and unlike agrarian life, city life is dominated by commerce. Fundamentally, everyone living in a city wants to know first and foremost what’s happening to prices and employment.



    May 15, 2015 at 10:56 pm

Comments are closed.

%d bloggers like this: