Spatial Analysis in the Social Sciences

The effects globalization exerts on domestic policies are widely debated in political discussion and political science research. Despite an extensive and cumulative portfolio of literature we still lack systematic and comprehensive comparative analyses of globalization processes and their political impact. Focusing on policy areas which are to a greater or lesser extent assumed to be affected by internationalization we try to establish if and to what degree globalization effects can be observed in highly developed industrialized democracies by means of aggregate data analysis. For this purpose globalization is conceptualized as a diffusion process that can take different form and thus allows for the gathering of more nuanced conclusions about the effects of globalization on domestic policies. Apart from the prominently discussed diffusion effects which stem from competition on world markets we also analyze other diffusion processes and their respective impacts (e.g. learning and emulation, trans- border contagion, coercion or imposed takeover and erosion of domestic economic assets). In this talk I discuss how the different diffusion effects are conceptualized and operationalized in our project, and describe some results from implementing these spatial lags into existing data sets.

Understanding the environmental and social context in which violence occurs has been an increasingly important area of research with broad applicability in public policy bearing on violence mitigation reaching far beyond purely research issues. Increased theoretical understanding of the spatial/environmental context of violence, such as theories relating alcohol outlets to violence, have gone hand-in-hand with improvements in spatial techniques and computing power, transforming this area of research. Further, although there has been research on violence related to alcohol outlet type and density there has been less research attempting to include measures of drug arrests and to more fully include spatial features such as characteristics of adjacent geographic areas on target area violence. In this presentation, by way of example, we will examine the relationship between alcohol outlets, drug markets (approximated by arrests for possession and trafficking) and violence in Boston as well as presenting related research in the city of Flint. An important emphasis for this talk will be on the process of doing spatial/geographical research in a public health setting, examining how to conceptualize and implement such analysis, including developing data layers, mapping and spatial/temporal approaches.

Recent scholarship has applied multilevel models and survey data to examine how income-based voting varies across states. Due to limitations of survey data, we know little about sub-state variation in income-based voting. New, geocoded registration records representing 73 million voters and a data set of 165,000 precincts-level election returns permit us to analyze income-based political behavior within small geographies. We find sub-state differences in income-based voting are explained by measures of racial diversity. Within homogenously non-Black areas, the wealth of the area bears no relationship to the income-partisanship correlation. The correlation between income and partisanship is strong in rural, heavily black areas of the Old South and weaker nearly everywhere else, including in modernized areas of the South. State-level differences in income-based voting are explained not by state-level factors, but by differences across a small number of racially diverse areas.

One justification for both American federalism and its heavy reliance on geographically defined legislative districts holds that Americans’ political interests are spatial, and are thus affected by their place of residence. Yet research on the effects of local context has primarily considered only a small subset of potential contextual influences, such as local political or racial demographics. This paper aims to expand the examination of contextual influences on political attitudes by developing hypotheses about the contextual attributes that are more or less likely to prove influential. It then analyzes public opinion on several issues with clear spatial implications. Among these issues, the paper examines the influence of living near coastlines or farmland on attitudes related to climate change; the influence of proximity to a nuclear power plants on support for nuclear power; the influence of living in a heavily polluted area on attitudes toward environmental regulation; the influence of living in a high-crime area on anti-crime attitudes; and the influence of proximity to wolf populations on attitudes about wolf hunting. In keeping with models emphasizing partisanship and symbolic politics, the central finding is of little spatial influence, although local crime rates do correlate with attitudes when crime is a nationally salient issue.

A large literature has looked at the determinants of the diffusion (or spread) of policy without analyzing why state policymakers innovate (create unique policy) or emulate (copy existing policy) in the first place. I argue that innovation and emulation are spatially-defined concepts and theorize about why policymakers in the legislature choose to innovate versus emulate. I illustrate my theory using data from the adoption of renewables portfolio standards over the last three decades.