Using Ideas As My Maps: Tracing the Flow of Ideas Through the Legislative Process

Ideas are an often-ignored part of the political science literature on legislation. Most studies of Congress focus on institutional concerns, party considerations, gains from exchange, and other unideated concepts. But the exchange of ideas in the writing of bills represents the fundamental realization of public policy in practice, and any story of major legislation that does not adequately deal with where the ideas in said statutes come from is missing a vital part of the story. Particularly, the role in Congressional hearings as an information exchange is often given only prima facie concerns in legislative expertise models, but what exactly is being transmitted is almost never addressed directly.

In the first section, I study how think tanks position themselves strategically to influence Congressional behavior. Using a dataset of think tank citations from Congressional floor speeches and committee testimony records, I compare the influence of think tanks based on a new measure their ideology, and in doing so, show that think tanks engage in strategic ideological positioning to maximize their influence. An additional hypothesis examined is the relationship between think tank members' previous work experience in government with the organizations' overall prominence. By treating think tanks as strategic actors in the complex constellation of legislative politics, this chapter makes the argument that think tanks need to be considered in even the most cursory account of the policy making process.

In the second section, I introduce an approach to idea tracing that is better able to answer the questions: whom does Congress listen to and when during the legislative process do they listen? Using machine learning and text-as-data approaches, I establish a more direct framework to evaluate the impact of testimony on the language of legislation through the development a text reuse approach to map specific ideas from congressional hearings to bills. Specifically, I focus on the language of Dodd-Frank and systematically trace the development of ideas with calls to testify in Congress from experts. In doing so, I provide the first account of who is being listened to during the legislative process.

The final substantive section expands on the previous section and applies the technique developed more broadly, to multiple pieces of important financial legislation. Using a topic model to identify major policy areas, I establish a more direct causal framework to evaluate the impact of testimony on the language of legislation. Utilizing a modified text-reuse method which combines a localized alignment text-reuse algorithm for identifying reused ideas, I compare the rate of idea reuse between hearings and final bill language for four major bills, spanning four Congresses. This allows for an effective way to test which speakers are being listened to during hearings, and under what conditions expert testimony has an impact on bill writing. By focusing on financial legislation, this paper can directly compare the work of people in multiple Congresses and, by discussing multiple bills, see when these speakers matter and when they don't, and in doing so, model a network of influence of external speakers in Congress.