Structural change, individual change, and four-story walkups

In this post, Alex Madva discusses his recent article in Journal of Applied Philosophy on the importance of an empirically-grounded approach to analysis and remediation of social injustice.

Should we “focus on structuring the social context, rather than changing the beliefs or values of individuals?”

No: Debates about the priority of social-structural versus individual change are confused, or so I never tire of arguing (see, e.g., these papers, and other contributions to this issue). The important questions are which kinds of individual and structural changes to pursue, and how best to think about individuals and structures in tandem. Which changes in individuals are most conducive to bringing about large and durable structural reforms? And vice versa? In “Integration, Community, and the Medical Model of Social Injustice,” I call for epistemic humility in these conversations. Before confidently asserting what’s required, we need to spend more time heeding, and producing, rigorous evidence.

Arguing that we need structural change is easy. (At that level of generality, liberals, conservatives, libertarians, socialists, and anarchists should all agree!) Which structural change we need—that’s hard. Consider Crul, Steinmetz, and Lelie (2020). They examined how the architecture of apartment buildings might increase or decrease ethnic tensions in Amsterdam’s diverse, working-class neighborhoods. Specifically, the researchers observed less tension among people living in four-story walkups than in larger suburban structures, which feature more semi-public spaces like inner courtyards.

Why? For one thing, children run around and make noise (as children do) in the courtyards. The courtyard walls amplify the noise, leading residents to complain about how “those people” don’t control their kids. Since the walkups offer fewer shared spaces, the children play outside in parks. Overall, interactions with one’s neighbors are more voluntary in the walkups. In the suburban buildings, the “architectural context that forces residents to share and, in some occasions, compete for the many semi-public spaces may amplify… negative feelings.” (And one can imagine how involuntary neighborly contact might amplify ethnic tensions during a pandemic!)

Studies like these highlight how 30,000-feet-up debates about “underlying structural” causes can be misleading. Is the most effective strategy for promoting justice to incentivize different groups to live, learn, and work together in shared neighborhoods, schools, and businesses? Well, what if the answer depends on unnoticed architectural details of the spaces people are moving into?

Structural Reforms for Racial Justice

An age-old racial-justice debate is between integration (bringing different racial groups together) and community development (building up power within racially oppressed spaces). Two recent contributions are Elizabeth Anderson’s The Imperative of Integration and Tommie Shelby’s Dark Ghettos: Injustice, Dissent, and Reform. My first encounter with Anderson’s book persuaded me that integrationist interventions are essential. Then I read Shelby’s response to Anderson, and I swung the other way.

Eventually I dug deeper. Making my way through their bibliographies (and their bibliographies’ bibliographies…), I sensed a disconnect between the state of the evidence and what Anderson, Shelby, and many others claim. This disconnect was apparent in their empirical predictions, but also in their normative claims, about the harms and wrongs of various structural reforms—about what’s required, permissible, and impermissible in the pursuit of racial justice. For example, Shelby criticizes Anderson’s recommended interventions as disrespectful restrictions on the autonomy of the oppressed.

Yet both place too much emphasis on too few studies, including a handful on residential integration. Integrationists overgeneralize from these studies’ successes while their opponents trumpet their shortcomings. But there simply aren’t enough studies to make generalizations about Residential Integrationism In General. (There are tons of studies, but very few with replicated comparisons between experimental and control conditions, with large data sets, etc.) The most we can say is that we should learn more about when these interventions do versus don’t work, and when they represent Paternalistic Social Engineering versus Respectful Expansions of Autonomy.

“Are Housing Vouchers Just?” = Poorly Formed Question

Policies for facilitating movement to advantaged neighborhoods often founder because so few participants can take advantage of them. For example, discriminatory practices in wealthy spaces exclude the entry of the less well-off, even when the latter have vouchers to help obtain new homes. But shortly after my paper appeared online, Peter Bergman, Raj Chetty, and colleagues made waves with a novel housing-voucher study. Dylan Matthews has a nice write-up about the study, which gave some participants more than a voucher:

“… They would also be given information on which neighborhoods promise the most opportunity for their kids… They’d also be assigned “navigators” whose job it was to walk them through the apartment application process, and receive additional financial assistance with down payments if necessary

… the additional support raised the share of families moving to high-opportunity neighborhoods from 14 percent to 54 percent. “This is the largest effect I’ve ever seen in a social science intervention,” Chetty said…
(To learn more, see Lecture 3 of Chetty’s online course.)

Such findings highlight that details of implementation make all the difference toward any intervention’s success. But Chetty is no diehard integrationist. He readily acknowledges the limits of voucher programs: “it’s not a scalable solution to just move everyone.” We “[u]ltimately need… policies that improve low-mobility neighborhoods rather than moving low-income families.”

Consider how Chetty frames this study’s significance:

“Something like 1 in 5 low-income families moves houses in a given year… this is not so much that we’re going to take a family that was going to stay in a given place permanently, uproot them and move them somewhere else. But rather, at the point that they’re moving for whatever other reason, let’s potentially encourage them to make a more informed choice…”

So is this Paternalist Technocracy or Respect for Autonomy? Is “Integrationism” about restricting or expanding options? And how can we know in advance whether so-called “Community Development” efforts actually restrict or expand the autonomy of the oppressed? (Gentrification, anyone?)

Or is the very debate between Integration and Community Development—pitched at high levels of abstraction—itself a piece of ideological bluster serving to reinforce the status quo rather than unravel it? What we really need are fair-minded examinations of concrete interventions, which thus far reveal how much we still don’t know.