When differences divide

In ethnic and racial terms, the U.S. is growing rapidly more diverse. Based on the theory of racial threat and the many studies that advance it, we should expect these changing demographics to lead to increased political contention between ethnic groups. Yet it remains unclear to what extent racial threat applies to the contemporary U.S. or to groups other than African Americans. This dissertation, comprised of an introduction and four essays, provides an alternative explanation of how inter-group political contention varies over time and space. Specifically, its theory of politicized change contends that local ethnic and racial cleavages are likely to be a source of political division under two conditions, both of which reflect the information environment. One is rapid local ethnic change, since sudden changes draw attention and destabilize shared expectations. The second is the availability of frames that politicize the demographic change, frames that often come from the media and from national politics. Neither of these factors has been emphasized in past accounts. Yet unlike many theories of inter-group proximity, this approach offers an explicit mechanism for how social cleavages come to be politicized. It also provides a dynamic component to theories of contextual effects. In this view, even seemingly local debates have national origins. To test its core hypotheses, the dissertation draws on several methods and on a wide range of data sources from 1950 to the present. After an introduction that highlights key theoretical claims, the first essay confronts the potential selection bias by using the post-Katrina migration as an exogenous, unanticipated shock to local demographics. Drawing on a new survey of southern residents and on statistical techniques including matching and differences-in-differences, it demonstrates that the contextual effect of living near the evacuees hinged on the available media frames. In places where the media linked the evacuees to public benefits, for example, support for anti-poverty spending declined markedly. The second essay provides additional evidence of how media frames can shape local encounters. It uses ten surveys from 1992 to 2006 to show that living in communities with growing immigrant populations made native-born Americans more anti-immigrant. But those effects only appear in the mid-1990s and in 2006, when anti-immigrant frames were nationally salient. The theory of politicized change can explain local political outcomes as well. The interaction of local demographics and available frames predicts where and when communities consider anti-immigrant ordinances, and also predicts spending in U.S. cities. Using a new, comprehensive data set of U.S. cities' spending since 1950, the third essay shows that beginning in the 1960s, diverse localities spent more of their budgets fighting crime. This remains true into the 1990s, even when taking crime rates into account. Contrary to past work, there is little evidence that other areas of spending have been affected by diversity in recent decades. These findings make sense given the prominence of racially charged frames about crime during precisely this period. The fourth essay presents a case study of votes on Massachusetts tax proposals. It reinforces the claim that changes in diversity are a powerful predictor of reduced public investments.