University of Georgia
In a career spanning four decades, Keith Poole's contributions to Political Methodology, indeed to Political Science and the study of American politics, have had enormous impact. The work is widely read and cited (north of 14,000 Google cites, which probably understates greatly given that many use NOMINATE data without citing their source), and his contributions have revolutionized the political methodology and provide a foundational bedrock for American institutional research. He is best known for his research on measurement models, especially the NOMINATE model co-developed with Howard Rosenthal that revolutionized the manner in which political scientists measure and think about ideology. One can say perfectly correctly, and without any hyperbole: the modern study of the U.S. Congress would be simply unthinkable without NOMINATE legislative-roll-call-voting scores. NOMINATE has produced data that entire bodies of our discipline—and many in the press—have relied on to understand the U.S. Congress. His Voteview website provides both academics and journalists and the general public a way to interrogate his measures of ideological placement of U.S. politicians across time.
NOMINATE (short for "Nominal Three-Step Estimation") is a procedure Keith began working on in the 1980s to estimate with computational efficiency the spatial structure of Congress, placing both members and roll-call votes in multidimensional space. Using technology available in 1988-89, Keith was able to compute two-dimensional D-NOMINATE scores for the first 100 terms of the U.S. House of Representatives and the U.S. Senate. He subsequently worked to create revised versions of the NOMINATE method and now has several forms of these data available. Keith's data—not only NOMINATE scores but also comprehensive archives of congressional roll call votes—have been freely available online since August 1995. In this way, the nature of Keith's work in the late 1970s foreshadowed many of the things that we now take for granted in the methodology community – a world in which political scientists would be skilled in statistical computation, with estimates from that computation accessible and replicable by anyone. Notably, even at a time when many were not willing to share data when asked, Keith had his data on display online and made it freely accessible for any interested person to download. In building the NOMINATE enterprise, he wrote software that was distributed freely to the community, allowing researchers to port these techniques to other settings. He also prepared and distributed roll call data from each U.S. Congress, while producing estimates of DW-NOMINATE scores every two years as well as replication code to reproduce those estimates. The upkeep of the data, the refining of the methods, and the maintenance of the website have all been as invaluable as the method. Over the past three decades, NOMINATE has been a model research agenda in political methodology, spawning a range of substantive and methodological extensions. On the methodological side, researchers have extended the NOMINATE model to incorporate Bayesian estimation techniques, to applications involving Big Data, and to produce valid ideal point estimates across time and institutions. On the substantive side, other scholars have ported these methods to different settings and data sets, such as the Supreme Court and the European Parliament. The substantive contribution that NOMINATE is perhaps best known for is one in which Keith has published extensively himself – the study of political polarization in American politics, not only one of the most important academic topics in American Politics but also one of the most discussed among the media and public as well.
The sheer volume of work that Keith Poole has contributed to the discipline is impressive in its own right: eight books, one monograph, and 68 articles. The body of this work has advanced our discipline's understanding of many fundamental questions in American democracy such as: how inequality contributes to ideological polarization and, with it, becomes a major problem for the nation; in economic policy, and how developments in infrastructure were key to American economic development and how poor economic policy can lead to financial crises. His methodological book and article contributions most centrally surround measurement theory. In these publications, Keith has shown the value of applying scaling techniques to the analysis of parliamentary voting, created new measurement techniques, and importantly has shown readers how they can apply the techniques in their own research. By developing standalone software, R packages, and example code for his books, he has fundamentally advanced political scientists' ability to conduct better research with original data using advanced measurement techniques.
What is less commonly known about Keith is that he has also been a tremendous mentor to a number of political methodologists. Keith has taught, mentored, advised, and/or worked closely over the years with many students and scholars junior to him, including John Londregan, Nolan McCarty, Jeff Lewis, Royce Carroll, Adam Bonica, James Lo, Chris Hare, Jamie Monogan, Ryan Bakker, Michael Lynch. On more than one occasion, he was known to go extraordinarily beyond the normal call of academic good citizenship to assist and help guide students and junior colleagues. One story recounted how Keith actually “changed some of the underlying code in W-NOMINATE [at that time] and compiled a new binary to help [a first-year graduate student just recently met at a workshop with his] project.” Keith subsequently hired that graduate assistant (even though he was located at another school) to work on extensions of W-NOMINATE he was then undertaking with Jeff Lewis & Howard Rosenthal; he worked closely with and mentored this student, James Lo, ultimately producing five papers and three R packages from their collaborations. Another recounted tale is of a junior colleague trained as an economist approaching Keith with a question whether economic performance influenced autocrats’ tenure similarly to how it was understood to affect the re-election fortunes of democratic governments. Keith guided the colleague to some relevant literature, and the two realized that they could make a positive contribution, resulting in a series of papers on the causes of coups d'_etat, that this substantive application in comparative politics and analysis of macro-level panel data being quite distant from Keith’s work on scaling and on the American Congress deterring him not in the least; that colleague, John Londregan, says of the experience: “This research project with Keith is one of the primary reasons my research agenda shifted from economics to political economy.” Keith’s engagement with younger scholars has exerted an enormous benign influence beyond his advisees and junior colleagues as well: Consider the huge number of scholars (a short list would include Josh Clinton, Jeff Lewis, Adam Bonica, John Londregan, Nolan McCarty, Simon Jackman, Douglas Rivers, Sebastian Saiegh, Marc Ratkovic) whose research builds on or depends on Poole's scaling methodology.
The conclusion from the nomination letter signed by 12 other nominators is entirely fitting: “Throughout his career, Keith has pushed the field of methodology light years ahead beyond boundaries that most practitioners could not even have conceived and found limiting without his work, and repeatedly demonstrated how methodology done well can surmount them and in so doing enrich our understanding of political phenomena. He is simultaneously an intellectual giant, a deeply creative political scientist, and a methodologist par excellence. His work has set a standard for all political methodologists to follow. He has influenced generations of young political scientists not only as a standard bearer for the profession, but also as a mentor and friend.” The award committee (Rob Franzese (chair), Lonna Atkeson, Wendy Tam Cho, Kosuke Imai, and Simon Jackman) unanimously and enthusiastically concur in awarding Keith Poole the 2016 Career Achievement Award of the Society for Political Methodology.
Selection committee: Robert Franzese (Michigan, chair), Wendy K. Tam Cho (Illinois), Lonna Atkenson (New Mexico), Kosuke Imai (Princeton), Simon Jackman (Sydney)