The Mismeasurement of Partisanship in the United States: Nonpartisans as False Negatives

Professor Andy Baker, University of Colorado Boulder

An important and unresolved debate in scholarship on American politics is how to classify Independent leaners—that is, survey respondents who, upon first query, self-identify as political Independents but who, in a follow-up question, reveal a leaning toward one party. The debate boils down to which approach and question wording locates the boundary between partisans and nonpartisans most accurately, and it remains unresolved because both sides have lacked the data and a convincing methodology to determine which approach does so. I propose to use item-response theory (IRT) methods on data collected via a question-wording survey experiment conducted on the 2017 CCES. Item-response theory is a psychometric methodology that can estimate which items from a set of question items produce more certain distinctions between groups. For my purposes, I intend to experimentally vary the wording of different partisanship questions and then apply IRT techniques to determine which ones produce the clearest distinctions between partisans and nonpartisans.  If experimental findings I have gathered in three foreign countries (Brazil, Mexico, and Russia) hold in the U.S., then the common practice of explicitly inviting U.S. respondents to report nonpartisanship is akin to reading aloud a “no opinion” option, which is the easy-out practice that is a well-known taboo in questionnaire design. If my expectations hold, then I would provide compelling evidence for the longstanding, but in practice ignored, argument to reclassify Independent leaners as partisans.  


Making Deliberation Work: Understanding the Effects of Deliberative Structures

Christina Ladam, Ph.D. Candidate, University of Colorado Boulder

This project focuses on understanding the ways in which people process information in a deliberative setting. Though research shows that people are biased information processors motivated by partisan-driven goals, I argue that a deliberative setting promotes accuracy-motivated reasoning when using certain design features. Throughout the fall of 2017, I am conducting small-group deliberative experiments with students on the CU-Boulder campus focusing on the effects of rules-giving and the role of a moderator on individuals' political behavior. These sessions are recorded, allowing for direct observation of participants' behavior. A pre- and post-test also allow for measures of opinion change. The findings of this project help to unpack the black-box of deliberation, allowing for greater understanding of what actually happens when people are brought together to discuss political issues. This is part of a larger dissertation project on deliberative institutions.


Tribal Institutional Design and Community Well-Being

Josalyn C. Williams, Ph.D. Candidate, University of Colorado Boulder

This research project is investigating the influence of political institutional design within American Indian communities on tribal well-being. American Indian communities have higher rates of poverty, lower education rates, higher crime rates, and lower general health rates as compared to the general U.S. population. Currently, the majority of the 566 federally recognized tribes have their own unique, formal, political institutions. A key component of this research project is the development of a comprehensive data set measuring detailed components of tribal governance, political institutions, tribal services and goods provisions, and tribal well-being. Using this data set, I plan to evaluate the effect of institutional legitimacy, perceived equality and diffusion of political power, and systems of policy enforcement/diffusion on community well-being.