The public and the media commonly refer to members of the United States Congress and other legislatures as "liberals" and "conservatives". The fundamental concern of this project is to provide accurate measurement of liberal/conservatism. As a result of our development of the NOMINATE scaling algorithm in the 1980s, political scientists and economists have used our measurements of Congress in scores of professional publications and used our algorithms as measurement techniques for dozens of other legislatures, ranging from state legislatures in the United States to foreign parliaments to the United Nations General Assembly. The measurements are frequently used in teaching American political history and in introductory courses on American politics.
This project increased the accuracy of measurement for Congress and updated the database current. Moreover, the work on other legislatures has suggested that the decision-making model that underpins NOMINATE does not always extend to other legislatures. Left-right (or liberal-conservative) preferences may be more sharply defined than in Congress. Abstention, as well as Yea and Nay, may be a meaningful choice. The project also provided work on estimation methods that allow for a more general approach to legislative behavior.
A unique feature of our work is that its dynamic feature permits comparisons of liberal/conservatism over time. For example, others have used these measurements to study the evolution of slavery as an issue in congressional politics in the nineteenth century and trade legislation, including reciprocal tariffs and fast track, throughout American history. This dynamic feature allows us to work on models with time-dependent preferences and on supercomputing techniques for large-scale computation.
The broad objective of this award was to increase understanding of legislative behavior and organization over the course of US history and to improve infrastructure for conducting such research. Our specific objectives were to increase the statistical rigor of the widely-used DW-NOMINATEand W-NOMINATE empirical models of roll call voting and also to develop geographic boundary files for every Congressional district in US history. In combination, these objectives allow scholars, journalists, students, and the broader public to study the geographic distribution of support for legislation as well as the geographic distribution of ideological variation in the Congress over time. Work done under this award has also demonstrated that more features of the mapping between legislators' ideologies and roll call voting behavior can be inferred from their roll call voting choices than had previously been understood. Results and products of this award have already and will continue to have an impact on the study of the US Congress in both contemporary and historical settings. The new software has been download over 1700 times and used by a large number of scholars leading to new research and publications. The digital Congressional district boundaries produced under this award may ultimately have the greatest impact by allowing Historians, Geographers, and Economists, as well as Political Scientists to link increasingly available geocoded historical information to data on members of Congress and their voters. More broadly, it is the ultimate objective this research to use the techiques and data developed and refined under this award to allow journalists, students, and the public to study, analyze and visualize the structure of Congressional voting over the insitutions entire history. By allowing the public to better understand Congress's current organization and function in greater historical perspective, the work arising from this project contributes the to the public's understanding of and appreciation for our democracy.
