Date of Award
5-9-2024
Degree Name
Doctor of Philosophy (PhD)
Degree Type
Dissertation
Abstract
An individual's greatest political voice is often within America's smallest democracies. Through three papers, this dissertation explores three key dimensions of political participation within America's local governments: representation, evaluation, and coverage. The first paper explores the question of whether a more accessible local government is a more representative one. By leveraging the switch to online meetings induced by the COVID-19 pandemic, I analyze whether the shift to more accessible meetings promotes healthy democratic participation or amplifies pre-existing inequities in local political participation. To accomplish this, I first construct a novel dataset of public meeting minutes from 2018 until 2022 for county, municipal, and school board governments within the St. Louis region. I then combine public participation data from these minutes with existing political and demographic datasets. Through a time-series cross-sectional analysis, I find that virtual meetings significantly increased participation in some local contexts but did not affect or decreased participation in others. Importantly, I find evidence that while virtual meetings experience similar inequalities in who participates as in offline meetings, they garner greater participation from women in county and school board governments. Before individuals decide whether to voice their concerns to a local government, they must, in some way, evaluate its performance. In the second paper, I examine how individuals form these evaluations and whether and how an individual's partisan identity may shape their evaluations. I argue that for some services, for which parties hold ideologically distanced and distinctly homogeneous opinions, evaluations of local services will be biased toward their party's average position regardless of a service's actual performance. To test this theory, I examine individual's evaluations of their local schools and police. Through a cross-sectional analysis of national survey respondents matched to measures of their local conditions, I find that for polarized services such as policing, individuals have a systematic bias in favor of their party's position regardless of the service's objective performance. Additionally, this bias exists irrespective of the partisan control of state and local governments. Finally, in the third paper, I introduce a promising framework to address a problem often hindering our coverage of local political data: the transformation of text-based sources. Political and legal texts include many meaningful data we need as researchers to quantify; however, these texts are often too complex to transform automatically and require slow and costly hand coding to do so correctly. I propose a novel approach: a human-in-the-loop pipeline that leverages Large Language Models (LLMs) to crowdsource and transform text-based sources into quantifiable datasets. Through case studies in local and judicial politics, I demonstrate that LLMs are a cost-effective alternative, producing results comparable to professional coders while significantly reducing transformation time. By easing the costs to process text-based sources, this framework promises to improve the accessibility of research and the coverage of important political outcomes. Overall, this dissertation makes significant contributions to understanding political participation in America's local governments by examining representation, evaluation, and coverage. It not only sheds light on the complexities and inequities of local democratic participation within an increasingly nationalized context but also introduces innovative methods to expand our understanding of local political dynamics.
Language
English (en)
Chair and Committee
Betsy Sinclair
Recommended Citation
Boschelli, Lucas, "A Big Voice in a Small Democracy: Three Papers on American Local Politics" (2024). Arts & Sciences Electronic Theses and Dissertations. 3009.
https://openscholarship.wustl.edu/art_sci_etds/3009