Showing that Neighborhoods Matter

A long-standing contradiction has posed a critical puzzle to urban sociologists: On the one hand, a wealth of ethnographic studies make the point that the kinds of neighborhoods people inhabit profoundly shape their fortunes; on the other hand, statistical analyses of “neighborhood effects” have often failed to confirm that claim. From the earliest studies of the Chicago School, through classic works such as Whyte’s (1943) Street Corner Society and Suttles’ (1968) description of defended turfs, up to, for example, Harding’s (2010) recent interviews of young boys in Boston, the message, repeatedly, is: neighborhoods have consequences. Yet, ruling out individual-level explanations for variations among places has proven difficult; econometric studies find that experiential and behavioral differences across neighborhoods largely derive from the different kinds of people who “select in” to them. Some scholars have therefore rejected the idea that neighborhoods matter (e.g., Mayer and Jencks 1989) and even the idea’s defenders struggle to find the evidence (e.g., Sharkey 2012). The question of whether contexts such as neighborhoods matter matters to sociologists far beyond the students of the city. Rob Sampson, in his magisterial book, Great American City: Chicago and the Enduring Neighborhood Effect, is fully aware of this controversy’s broadest implications for sociology (see, e.g., Ch. 15; pp. 435–6) and his book is a major contribution to that debate.1 I will discuss how after reviewing the state of play in neighborhood effects. (This contribution to the symposium is thus more focused on basic theory and research than are those of my fellow contributors.) Sociologists assume that contexts—workplaces, schools, organizations, nations, families, and personal networks, as well as neighborhoods—substantially shape individual experience and action. The general public also believes this. Parents, for example, try hard to access the “best” contexts they can find for their children, particularly the “best” neighborhoods, schools, and peer groups. Sociologists have increasingly focused their attention—in part, thanks to Sampson’s Chicago project itself—on whether and how neighborhoods really matter. The number of studies on neighborhood effects grew exponentially since 1990.2 But urban sociologists’ and the public’s assumption that neighborhoods matter may be wrong or may be right only in small ways. Variations across places in individuals’ experiences and actions may be satisfactorily explainable by the personal tastes, resources, and habits that the individuals bring with them. The first direct assault in sociology on the contextual assumption was, I believe, Robert Hauser’s 1970 article on whether the gender composition of schools affects students’ aspirations. He concluded that the “contextual interpretation [in general] is . . . speculative,