A while back, while reading Daniel Hertz’s recent article about Chicago in City Observatory, I came across the work of Robert Sampson, an urban social psychologist at Harvard. The work is over a decade old, but I still find it compelling and relevant.
Sampson’s basic thesis is that perceptions of disorder, i.e. a chaotic or blighted neighborhood, have little to do with actual measurable disorder in a neighborhood. Rather, our reactions about which neighborhoods are ordered or disordered are colored by race and previous assumptions. He calls this the “enduring neighborhood effect.”
Here’s one of the charts from his 2004 Social Psychology Quarterly article. It’s based on research he did in Chicago:
Here’s how Sampson describes the above chart, from his (highly technical!) social psychology article titled “Seeing Disorder: Neighborhood Stigma and the Social Construction of “Broken Windows,” co-authored with Stephen Raudenbush:
The results so far support the hypothesis that neighborhood racial context helps shape perceptions of disorder, with controls for carefully observed disorder. Does this tendency depend on the resident’s race? Do the effects of an individual’s race on perceptions of disorder vary randomly across neighborhoods? We tested this specification; once again, black residents reported less disorder than whites (coeff. = –.13, t = –3.11), an effect that varied randomly across neighborhoods. More interesting, however, the contextual effect of racial composition is largely independent of the observer’s ethnicity.13 Specifically, blacks were not significantly more or less likely than whites to view predominately black neighborhoods as high in disorder, with controls for observable disorder and other covariates.
A notable exception was the interaction between Latino ethnicity and block-group percent black. Perceptions of disorder increase as a function of percent black for members of each ethnic group, but this tendency is significantly pronounced for Latinos. This relationship is graphed in Figure 1; predictors not displayed are held constant at their means. In neighborhoods less than 25 percent black, whites and Latinos essentially do not differ in their perceptions of disorder. At roughly 25 percent black, however, a threshold suggested by prior research as particularly salient (e.g., Schelling 1971), Latinos begin to diverge sharply from whites: when neighborhoods reach 75 percent black or more, Latinos perceive substantially more disorder than do whites (also see Charles 2000).
The authors’ key conclusions are that perceptions of neighborhoods are “central” to how we make decisions about neighborhoods. They call this “stigma” and argue that it’s a big reason for urban inequality.
Social structure proved a more powerful predictor of perceived disorder than did carefully observed disorder. The data suggest that in shaping perceptions of disorder, residents supplement their knowledge with prior beliefs informed by the racial stigmatization of modern urban ghettos (Wacquant 1993). These beliefs, we suggest, may be incorrect but not necessarily “irrational” or reflections of simple racial prejudice. The rational basis of these beliefs lies in a social history of urban America, which links geographically isolated ethnic minority groups with poverty, economic disinvestment, and visible signs of disorder (Massey and Denton 1993). Skin color is not only visually but also psychologically salient in a society with a long history of slavery, segregation, and racial conflict (Loury 2002).
One key point that they make is that these perceptions hold across racial groups, so that African-Americans are equally likely to view certain neighborhoods as being “disordered” because of racial composition. To make a very long story short, our perceptions are largely shaped by our assumptions, at least for neighborhoods with more extreme homogenous concentrations. Persistent segregation and stigma is as much a social psychological problem as it is a structural problem.
For more on this, check out Daniel Hertz’s recent article in City Observatory. I wonder if much has changed on this since the study was first published?
[Read the whole technical study here.]