Christopher H. Achen


Several times in his academic career, Christopher “Chris” H. Achen, the Roger Williams Straus Professor of Social Sciences, and professor of politics, had the good sense to be in the right place at the right time. After an undergraduate education at the University of California-Berkeley, his pursuit of a Ph.D. led him to Yale University’s political science department at the height of its intellectual dominance of the discipline. His faculty appointments took him first to the University of Rochester, briefly back to Yale, back to Berkeley for a decade, and after five years at the University of Chicago, on to the University of Michigan, the primary center for the study of political behavior and the uncontested incumbent leading survey research in the second half of the 20th century. Chris joined Princeton’s Department of Politics in 2004, during a period of rapid expansion and rejuvenation that saw the department recruit its current leaders and a large crop of assistant professors.

Far from just following the intellectual stream of his times, Chris made the departments where he taught throughout his career the right places to be—no small feat in big political science departments or, in fact, in the field of quantitative political science. Arguably the most famous political methodologist of his generation, and a member of the American Academy of Arts and Sciences, he has long been a leading advocate for developing the field of quantitative methodology in the discipline. He is also a shining example of the generation of genuine generalists, contributing to every major subfield of political science, publishing not only in behavior and methodology in the American context, but also on comparative political systems, rational deterrence theory in international relations, and the empirical foundations of political theory.

During his graduate education, influenced by course work with Gerald Kramer, he specialized in quantitative methodologies, particularly Bayesian inference when its development in political science was only at its very beginnings and when political science had yet to establish political methodology as a subfield. As a faculty member, he taught a variety of methods classes and topics. From 1983–1985, he was the founding president of The Society for Political Methodology (and the methods section of the American Political Science Association).

Over the years, Chris published three influential monographs on political methodology: on regression analysis, threats to causal inference using observational data, and ecological inference. His long involvement with survey research, first at the budding Survey Research Center at Berkeley, then at its much larger parent-in-spirit, Michigan’s Institute for Social Research, added another applied element to his experience with quantitative study of politics.

His contributions to political science combine a sophisticated use of methods with a keen understanding of substance. Among the topics he has written about are the organization of ordinary people’s political belief systems, voter turnout and decision-making, representation, and accountability (or its lack) in contemporary democracies. While a good deal of his research focuses on the American context, he has also engaged in a number of international collaborations, co-directing book-length projects about decision-making in the European Union and voting behavior in Taiwan.

His classic papers remain staples of graduate syllabi. One of them is his first publication, “Mass Political Attitudes and the Survey Response,” which appeared in the discipline’s flagship American Political Science Review. There, he confronted Philip Converse’s famous account of public attitudes, offering the first and still most powerful critique. Where Converse diagnosed widespread absence of coherent and systematic thinking about politics in the mass public, Chris saw blunt measurement instruments and sloppy statistical interpretation of data. The “nonattitudes” Converse believed he found were in fact much more stable attitudes hidden by a whole lot of error in measuring them. One of the first applications of measurement error models to political science, his first published work has had profound influence on how we understand variability in people’s survey responses.

Chris’s early work on representation provides another illustration of how appropriate methodological tools can make a marked difference for understanding of substantive problems. In “Measuring Representation: Perils of the Correlation Coefficient,” published in 1977 in the American Journal of Political Science, he dismantled the conventional wisdom that members of Congress were more responsive to constituents’ civil rights positions than to their opinions about foreign policy. Taking appropriate account of the fact that the variance of civil rights positions was considerably higher, Chris showed that legislative responsiveness did not in fact differ between the two policy domains and recommended that for future research, “correlations should be abandoned in the study of representation.” Another article on the same topic, published soon after, draws on normative political theory to build a framework for quantitative analysis of the relationship between legislators and constituents, emphasizing the distinction between “proximity” (how close the legislator’s preference is to the constituents’) and “responsiveness” (how much the constituents’ preferences seem to affect the legislator’s preference).

In a pair of highly regarded articles on voting behavior and party identification (“Breaking the Iron Triangle: Social Psychology, Demographic Variables and Linear Regression in Voting Research” in 1992 and “Parental Socialization and Rational Party Identification” in 2002), Chris brought a more precise, formalized approach to a major reconceptualization of how people form and update their attachment to political parties.

The deepest intellectual bond Chris formed may have been with Larry Bartels, the Donald E. Stokes Professor in Public and International Affairs, Emeritus, and professor of politics and public affairs, emeritus, at Princeton. Academic companions, friends, and co-authors, the two overlapped only intermittently at the same institutions. They met during Chris’s year on the Yale faculty, when Larry was a first-year student there looking for political science classes that didn’t require him to get up early. Chris, a well-known night owl, taught his course “Quantitative Methods in Political Science” at the right time on the right topic for Larry. A few years later, Chris taught Larry again, when Chris was on the faculty at Berkeley and Larry was in the Ph.D. program in political science there. They were faculty colleagues at Princeton from 2004–11. Research conducted during this period led to their book Democracy for Realists: Why Elections Do Not Produce Responsive Government, published in 2016 and already cited more than one thousand times. In this landmark and wide-ranging account of voter decision-making, they not only popularized the study of shark attacks in political science (to understand the political consequences of random events), but also challenged the fields of public opinion and elections by comprehensively debunking sunny claims about the meaningful and reflective input citizens provide in an electoral democracy.

Known for his careful feedback, usually delivered modestly from the back of the room after others had asked their questions, Chris has urged many cohorts of students (and assistant professors) to think before they estimate. His 2002 article “Toward a New Political Methodology: Microfoundations and ART” put some of this advice in print, cautioning against the mindless inclusion of large numbers of independent variables in regression models and recommending instead a deliberate and precise substantive foundation for statistical estimation. Illustrating the impact of his methodological advice on data analysts, another paper, “Why Lagged Dependent Variables Can Suppress the Explanatory Power of Other Independent Variables,” has gathered more than a thousand citations without even being published.

His colleagues and graduate students appreciated this feedback, and we all feel fortunate to have been in the right place with Chris at Princeton. He prioritized and took great pride in his graduate student mentoring, which, not surprisingly, was recognized by Princeton, with a Graduate Mentoring Award by the McGraw Center for Teaching and Learning in 2017. His Princeton students raved about his tireless and empathetic mentorship. He spent equal times training them in the skills of the discipline, and patiently and sensitively helping them overcome doubt and personal hardship. As one of his graduate students wrote in the University citation, “[Achen] made it clear to me that he had faith in me, not just to finish the Ph.D. program but to excel as a teacher, a speaker, and a scholar. That confidence has helped me when I have faced challenges outside of academia as well.” Chris also received a lifetime achievement award for graduate mentorship from the University of Michigan.

He has led efforts to promote greater diversity at Princeton and in the discipline more broadly, including articles he has written to the Political Methodology Society urging greater efforts at inclusiveness within the field.

To know Chris is to know someone with a spectacular array of interests and talents. We will miss him in department meetings, but we look forward to many more engagements, ideally over a cocktail at one of Princeton’s popular watering holes, with conversations ranging from Plato to the habits of rural Vermonters during the Civil War to the ongoing plight of the Michigan football team.