James G. Gimpel
Associate Professor
Department of Government
University of Maryland, College Park

POSITION STATEMENT

My field is political behavior, including the geography of voting and elections. I am particularly interested in how economic and social forces with clear spatial dimensions (re)shape the politics of places. On the theoretical side, I am working to understand the cognition of political environments. To this end, the work in social psychology and in the newer field of environmental psychology has been especially helpful.

Current Research

I have several projects underway, all of which integrate spatial and contextual analysis. One of my papers proposes to push the political participation literature in a new direction by using innovative methodologies to examine the geographic accessibility of precinct polling places. My argument is that commuting to and from precinct locations can be burdensome on potential voters, particularly on a busy weekday in congested metropolitan areas when many voters are pressed by the demands of everyday living: work, family and school. Some precinct locations are more accessible than others, and for the less accessible ones, at least some people will feel that the cost to get there outweighs any benefit they may reap in terms of personal satisfaction from having fulfilled a civic obligation.

Accessibility is a core concept in this study, but it is not normally a political science concept. The notion of accessibility originates mainly from the literature on transportation planning and engineering, and, to a lesser extent, economics. Accessibility is thought to be the reciprocal of the costs of moving people and goods between points in space. Travel costs are central because the less time and money spent in travel, the more places that can be reached within a certain budget, and the greater the accessibility. As defined by transportation planners, accessibility is a function of distance and impedance. Taken together, distance and impedance provide a measure of travel cost to and from a destination. And as these costs increase, turnout is hypothesized to diminish. Even after controlling for variables that account for the motivation of the precinct population, and the competitiveness of elections, I believe that accessibility will make a significant difference to turnout.

Spatial Data Analysis

For all of my ongoing work in geography, I use ESRI�s Arcview software, along with relevant extensions and scripts. In order to calculate distance and impedance measures, I will be using the Network Analyst and Spatial Analyst extensions, along with several scripts that are available for free on ESRI�s website for calculating distances between, say, points and polylines.

For spatial statistics and data analysis, I use Luc Anselin�s SPACESTAT. It is exceedingly user friendly. I make use of SPACESTAT calculated spatially lagged dependent variables in the regressions reported in my book, Separate Destinations: Migration, Immigration and the Politics of Places (University of Michigan Press, 1999).

I am making increasing use of Gary King�s ecological inference maximum likelihood technique, and its accompanying software. Given that so much spatial analysis is done using ecological data, I think it should go hand-in-hand with many spatial analytic applications.

In two books I have used Atlas GIS (also an ESRI product) to produce camera-ready chloropleth maps. (The publishers later ruined several of these through careless production mistakes). Although Atlas GIS is limited, it is simple to learn with minimum start-up costs. If you need to map your data fast, it is a good program.

Best Practices in Political Science

For best practices, the journal Political Geography is a good starting point, with mostly high standards. As for authors in political science, I think John Sprague and Carol Kohfeld (Washington University, St. Louis), and Robert Huckfeldt (Indiana University), and several of their students, are at the cutting edge. In the political science subfield of international relations, Michael Ward (University of Washington) comes readily to mind. There are others, and the number of us is growing.

Suggestions for Resources, Workshops

I think that training with the software is always a need, but often we can figure that out with help from local experts on our own campuses. I would like more effort focused on the theoretical underpinnings of empirical work with spatial data. For example, I could learn a lot from a workshop where environmental and social psychologists brought their insights to bear on matters relating to the cognition of political and social environments, with additional theorizing and hypothesis testing about how those cognitions are translated into attitudes and behavior. My interest in geographic and spatial patterns in my data far outpaces my ability to explain those patterns.

With the 2000 census data to be released within the next two years, it would be imminently reasonable to work up a series of conferences that use this data to highlight spatial patterns and processes.