How Mediated and Political Contexts Shape State Lawmaking in Wisconsin

• In Event: Eleventy Million Observations: Applications in Text Analysis & Machine Learning

Abstract

Despite the considerable attention to an increasing partisan divide within both American state legislatures and the public more generally, little attention has been directed at how social and political contexts might encourage greater or lesser political polarization at the state level. We propose to link our database of Wisconsin legislative behavior from 2010-2019 (bill introductions, roll call votes, bill sponsorships, and the text of the bills themselves with our database of state politics news coverage from 27 newspapers across Wisconsin about Wisconsin politics (totaling about 300,000 articles) to explore the relationship between the volume and tone of news coverage and legislative behavior among both individual legislators and geographic clusters of legislators. Using a combination of topic modeling and time series analysis techniques, we generate dynamic regression models to explore whether newspaper coverage precedes or follows legislative activity on contentious political topics. We will also focus on whether party representatives respond differently to state-level news coverage on particular issue domains, especially when their geographically proximal lawmakers are of their own party.

Authors

• Jordan Foley, University of Wisconsin-Madison

• Josephine Lukito, UW Madison

• Dhavan Vinod Shah, University of Wisconsin

• Michael W. Wagner, University of Wisconsin, Madison