EXISTING DATA: PROBLEMS AND POSSIBILITIES

Although these issues create difficulties in collecting local elections data, the institutional variation of municipal governments also provides a unique opportunity to study the relationships between electoral arrangements and a number of outcomes, including turnout, vote choice, candidate emergence, and the competitiveness of electoral contests. While some urbanists have pursued these questions, this area of inquiry has predominantly sat in the purview of comparativists and state politics scholars. The absence of a central, standardized database on local elections has played a role in this oversight. Unlike federal and state election returns, which are now readily accessible at the state, metropolitan, and county levels through publications like the Statistical Abstracts of the United States (U.S. Census Bureau 2010) and County and City Data Book (U.S. Census Bureau 2007b), election returns for local races have never been systematically compiled at any level of aggregation for any period of time. Consequently, most research on local elections is based on case studies and, apart from research on minority representation in local legislatures (which uses rosters rather than election data), a small set of studies that rely on larger samples.

The absence of systematic, longitudinal local elections data has serious implications for both empirical description and causal inference. First, not only do we have a very incomplete and potentially severely biased picture of electoral processes and outcomes at the local level, but existing studies' foci on the largest cities in the nation have limited both the generalizability of empirical findings and, more generally, the questions addressed by extant studies. Second, with only a few exceptions, existing data are based on cross-sectional designs that are ill-equipped for testing many of the theories central to debates within electoral studies and political science. Third, all existing local elections datasets examine only the final stage of the electoral process, ignoring primaries or general elections that do not produce winners, which, for many cities, are the most salient and competitive races. Fourth, almost no datasets include precinct-level election returns (but see Barreto Reference Barreto2007; Barreto, Villarreal, and Woods Reference Barreto, Villarreal and Woods2005). Fifth, only one study (Oliver and Ha Reference Oliver and Ha2007) has incorporated multi-level data to investigate cross-level inferences and the effects of context on individual-level attitudes and behaviors. Despite the nested nature of local jurisdictions (individuals–precincts–wards–cities) and the fact that cities offer the best opportunity for studying the impact of context on behavior, no existing dataset is designed to exploit and leverage these comparative advantages.

Beyond questions of local politics and elections, a local elections database with the ability to provide connectivity to several companion datasets would enable social scientists to test an even larger set of theories. Two prime examples include the Record of American Democracy (ROAD) dataset (King et al. Reference King, Palmquist, Adams, Altman, Benoit, Gay, Lewis, Mayer and Reinhardt1997) and the State Legislative Election Returns (SLER) project (Carsey et al. Reference Carsey, Berry, Niemi, Powell and Snyder2008). The ROAD data include precinct-level election returns between 1984 and 1990 for all federal offices, partisan statewide elections, and state legislative elections, as well as party registration and enrollment when available and some political and census data for precincts and slightly more aggregated units. SLER is a candidate-based dataset (N = 259,000) that includes information about general elections for state legislative seats from 1967 to 2003, as well as some primary elections between 1967 and 1988. Although SLER does not include variables measuring political or sociodemographic features of legislative districts, states, or chambers, it does include unique district identifiers (within states and chambers) that allow such data to be readily merged. At present, neither ROAD nor SLER include local elections returns, but their connectivity with local elections data would enable elections scholars to test an extraordinarily large and diverse set of research questions at various levels of aggregation.

Finally, a local elections database would offer an important complement to survey data, which tend to dominate the study of local political behavior. Existing surveys such as the Multi-City Study of Urban Inequality (MSCUI), the National Politics Survey (NPS), the Latino National Survey (LNS), the National Black Election Study (NBES), the Detroit Area Study (DAS), and the American National Election Survey (ANES) provide important information about potential voters' behavior, information levels, and attitudes, but they are not ideal for understanding either electoral behavior per se or the effects of institutional or contextual variables on political behavior. Moreover, the high costs of survey research have historically limited studies to a few cities and a single point in time. Finally, the focus on constituents renders surveys inapplicable to questions about candidates in elections, and thus, a full understanding of the dynamics of local elections cannot be garnered from survey data alone.

LOCAL ELECTIONS IN AMERICA PROJECT

Given this need for a local elections database, in July 2009, we organized a National Science Foundation (NSF)–sponsored workshop that brought together leading scholars, practitioners, and organizations committed to local elections, minority politics, and data collection. We asked participants to consider two questions: First, what are some of the distinguishing features of local governments and elections, and in what ways can these features shed new light on old questions and provide fruitful ground for the study of new or emerging questions? Second, what are the fundamental issues surrounding the compilation, organization, and maintenance of a local elections database? The workshop was extremely helpful in developing the Local Elections in America Project (LEAP), which received funding in 2010 from the NSF's Division of Social and Economic Sciences (Political Science) and its Office of Cyberinfrastructure CF21 Venture Fund, as well as from the Ken Kennedy Institute for Information Technology at Rice University. This funding supports the creation of a software tool to systematically collect, digitize, and disseminate data on municipal elections. Such an infrastructure will allow us to create a centralized, comprehensive, and cost-effective local elections database for past and future elections by automating as much of this process as possible. The LEAP database will therefore fill a critical void in the field of elections and make a substantial contribution to the fields of urban and minority politics.

This symposium brings together the collective wisdom of many of the workshop participants and sets an ambitious research agenda for political scientists whose interests span the fields of campaigns and elections, racial and ethnic politics, federalism, and state and local politics.

In the first article, Karen Kaufman and Antonio Rodriguez make a compelling case for the study of local politics and elections by focusing on the changing demographic composition of the United States. As they point out, these trends have been manifest in American cities for some time. For example, non-Hispanic whites already comprise less than 50% of the populace in 62 of the 245 largest cities (U.S. Census Bureau 2002). Moreover, in 2006, approximately 80% of elected officials of color served on municipal councils or school boards (Hardy-Fanta et al. Reference Hardy-Fanta, Lien, Sierra and Pinderhughes2007). Similarly, in his article, Baodong Liu zeroes in on the distinctive racial and ethnic characteristics of cities, examining the issue of multiracial coalitions and the conditions under which they form. Using county-level data from the 2008 presidential elections, he illustrates how the support of African American, Latino, and Asian voters varied according to the racial/ethnic context.

The articles by Paul Lewis and Brian Adams shift the focus away from issues of race and ethnicity to address different substantive areas associated with local politics and elections. Lewis investigates the issue of scale—the number of inhabitants (or, alternately, constituents or voters)—and its relationship to the behavior of local politicians and voters. He considers explanations for why political scientists have tended to overlook this issue and lays out several intriguing areas for future research. Adams looks at the issue of campaign finance, a well-studied topic for political scientists, yet one that has been virtually ignored in the context of local elections. His article explores how various features of local politics and elections provide new avenues for inquiry and additional leverage for understanding more commonly examined relationships. Finally, in the last article of this symposium, Christine Kelleher Palus returns to the issue of data collection, sharing personal experiences and words of wisdom based on her own efforts at compiling a large-N sample. In addition, she maps out additional avenues of research that could be fruitfully exploited using local elections data.

All of the contributions to this symposium highlight theoretical and empirical advancements that could be made with the creation of a local elections database. The breadth and depth of the institutional, demographic, and candidate features that characterize the approximately 85,000 local governments in America provide fertile and as yet untapped resources for political scientists. We believe that the old adage that “all politics is local” rings as true today as it has at any period in our history. Indeed, as the nation continues to move toward a multi-racial, multi-ethnic reality, the lessons learned from local politics and elections today will pave the road for a better understanding of state and national politics. As the articles that follow document in greater detail, the study of local politics and elections represents a looking glass into the future.

Author Bios

Brian E. Adams is an associate professor of political science at San Diego State University whose research focuses on democratic practices in local politics. His publications include two books—Citizen Lobbyists: Local Efforts to Influence Public Policy (2007) and Campaign Finance in Local Elections: Buying the Grassroots (2009)—as well as articles in the Journal of Urban Affairs, Urban Affairs Review, and Public Administration Review. He can be reached at badams@mail.sdsu.edu.

Karen M. Kaufmann is an associate professor and associate chair of the department of government and politics at the University of Maryland. Her research interests include urban politics as well as Latino and black politics. Her work has been published in such journals as the American Journal of Political Science, the Journal of Politics, Political Research Quarterly, Urban Affairs Review, and the Du Bois Review. She is also the author of The Urban Voter: Group Conflict and Voting Behavior in American Cities (University of Michigan Press). She can be reached at kkaufmann@gvpt.umd.edu.

Paul G. Lewis is an associate professor of political science in the School of Politics and Global Studies at Arizona State University. In addition to numerous articles on local politics and public policy, he is the author of Shaping Suburbia (University of Pittsburgh Press, 1996) and coauthor of Custodians of Place (Georgetown University Press, 2009). He can be reached at Paul.Lewis@asu.edu.

Baodong Liu is an associate professor of political science at the University of Utah. He is the author of The Election of Barack Obama: How He Won and a coauthor of Race Rules: Electoral Politics in New Orleans, 1965–2006. He can be reached at baodong.liu@poli-sci.utah.edu.

Melissa Marschall is an associate professor of political science at Rice University in Houston, Texas. Her research interests include immigrant and minority incorporation, urban politics, education policy, race and representation, and political behavior. She can be reached at marschal@rice.edu.

Christine Kelleher Palus is an assistant professor and director of the master of public administration program at Villanova University. Her research interests are urban politics and public management. Her work has been published in such journals as Urban Affairs Review, the British Journal of Political Science, and the Journal of Public Administration Research and Theory. She can be reached at christine.palus@villanova.edu.

Antonio Rodriguez is a graduate student in the department of government and politics at the University of Maryland. His current research pertains to Latino politics in the United States. He can be reached at anrodriguez@gvpt.umd.edu.

Anirudh Ruhil is an associate professor at the Voinovich School of Leadership & Public Affairs at Ohio University. His work spans urban politics (with an emphasis on race and representation), public health research on survey-based incidence and prevalence rates, K–12 educational research (value-added analysis for accountability and school improvement purposes), and college access and retention in rural America. He can be reached at ruhil@ohio.edu.

Paru Shah is an assistant professor of political science at Macalester College in St. Paul, Minnesota. She received her Ph.D. in political science from Rice University. Her research areas include urban politics, education policy, race and ethnic politics, and research methods. She can be reached at shahp@macalester.edu.