Henry Brady's career is a testament to the virtues of big social science—big questions, big datasets, big toolkits, big institutions. While much of political science still functions as a cottage industry, he has played a leading role in the second generation of our discipline's post-war industrial revolution. Building on the varied contributions of such earlier stalwarts as Warren Miller, David Truman, Gabriel Almond, and Paul Lazarsfeld, Brady has contributed significantly to the theoretical, empirical, methodological, and institutional development of the field. With intelligence, energy, and good will, he has instigated and nurtured a remarkable variety of projects, partnerships, and programs to advance our understanding of important political phenomena.
2009–10 APSA President, Henry Brady, Class of 1941 Monroe Deutsch Professor of Political Science and Public Policy, University of California, Berkeley, Ph.D., MIT

Brady is the Class of 1941 Monroe Deutsch Professor of Political Science and Public Policy at the University of California, Berkeley. For the past decade he has also been the director of Berkeley's Survey Research Center. This summer he left that post to become the dean of the Goldman School of Public Policy, the school he first joined as a new assistant professor more than 30 years ago.
Brady's career trajectory is an object lesson in the long-term scholarly value of patient investment in intellectual capital. He studied math and physics as an undergraduate at Harvey Mudd College and urban politics, public finance, political economy, and econometrics as a graduate student in economics and political science at MIT. The result was a broad curiosity and a formidable set of skills; but skills do not translate automatically into first-rate research.
During his first six-year stint at Berkeley, Brady published—literally—nothing. However, his work during that period bore handsome fruit over the next five years, which he spent at Harvard (1985–87) and the University of Chicago (1987–90). His methodological work on scaling and dimensional analysis produced a series of important articles. His theoretical and empirical analyses of political psychology and choice resulted in two influential articles published in the American Political Science Review (Brady and Sniderman Reference Brady and Sniderman1985; Brady and Ansolabehere Reference Brady and Ansolabehere1989). He also formed a research partnership with Richard Johnston that would produce a steady stream of important publications in the years to come.
Similarly, during his years at Harvard and Chicago Brady invested substantial time and energy in ambitious projects that would bear fruit after he returned to Berkeley—this time in political science as well as public policy—in 1990. His work with Johnston, André Blais, and Jean Crête on the 1988 Canadian Election Study led to the publication in 1992 of Letting the People Decide: Dynamics of a Canadian Election. His work with Sidney Verba, Kay Schlozman, and Norman Nie on the 1990 Citizen Participation Study led to the publication in 1995 of Voice and Equality: Civic Voluntarism in American Politics. And his work with Cynthia Kaplan on political participation in what was then the Soviet Union led to a two-decade-long collaboration whose main fruit will be the forthcoming book Gathering Voices: Political Mobilization and the Collapse of the Soviet Union.
Brady's partnerships with Johnston, Verba and Schlozman, and Kaplan are just a few of the dozens of productive collaborations he has participated in over the years. Of the 93 books, monographs, articles, chapters, reviews, and reports listed on his curriculum vitae, 67—including most of the most recent, and most of the most influential—include the keyword “with.” Few of these are works Brady could have produced himself; but equally few are works his myriad collaborators (including each of us) could have produced without him.
Along the way, Brady's scholarly activities have put him, Zelig-like, at the scene of several key events in contemporary political history. He was in Moscow two weeks after the 1991 coup attempt and at the Kremlin two days before the collapse of the Soviet Union. His role in the 1993 Canadian election study provided a ringside view of the dramatic collapse of the Tory Party in Canada. In 1997 he was instrumental in persuading the state of California to mandate rigorous policy evaluation as part of welfare reform legislation. In November 2000 he was in Palm Beach, Florida, providing forensic analysis of vote returns in the midst of the contentious aftermath of the 2000 presidential election. And his subsequent work as an expert witness contributed to the outlawing of inferior punch-card voting systems in Illinois and California, in the latter case resulting in the temporary postponement of the 2003 gubernatorial recall election.
In recent years Brady's scholarly agenda has continued to expand without apparent limit, including new and forthcoming work on political identity, voting technology, the logic of causal inference, the structure of political cleavages, and the impact of welfare policies, among other topics. With so much ground to cover, we can only touch on highlights.
DIMENSIONAL ANALYSIS AND SCALING
In an era when political methodologists mostly recycled techniques from other disciplines, Brady published significant original methodological work both within and beyond political science. His research on dimensional analysis and scaling combined a deep theoretical grasp of measurement with the technical skill of an MIT-trained econometrician and the sensibility of an engineer. He illuminated important theoretical connections among alternative statistical models, clarified the statistical properties of familiar techniques, and developed new estimators tailored to deal with the distinctive features of important classes of data.
Brady's first major article, “Statistical Consistency and Hypothesis Testing for Non-Metric Multidimensional Scaling” (Brady Reference Brady1985b), appeared in Psychometrika. It singlehandedly provided a firm statistical foundation for what had long been a popular technique for exploratory data analysis.
As a thoughtful practitioner of survey research, Brady was acutely conscious of the statistical difficulties engendered by individual differences in translating latent judgments into numerical or categorical response data. In “The Perils of Survey Research: Interpersonally Incomparable Responses,” Brady (Reference Brady1985a) provided a patient, intuitive analysis of the problem of interpersonal incomparability, specifying the (quite precarious) conditions under which observed correlations are unaffected by individual differences in the interpretation of response options and providing a simple example of how statistical models can go wrong when those conditions do not hold.
In “Traits versus Issues: Factor versus Ideal-Point Analysis of Candidate Thermometer Ratings,” Brady (Reference Brady1990b) noted that political scientists use two quite different kinds of models to analyze choices. One, embodied statistically in factor analysis, assumes that the utility of any given alternative is a linear function of its unobserved characteristics. The other, embodied in spatial models, assumes that the utility of an alternative is a function of proximity to some unobserved “ideal point” in a hypothetical space. Brady drew on the relevant literatures in statistics, economics, and psychology to clarify the relationship between factor-analytic models and ideal-point models while firmly grounding both in more fundamental theories of choice. He presented a flexible new estimator for these models, demonstrated its statistical properties—a rarity—and showed how it could be used to test hypotheses about dimensionality and functional form, and even to adjudicate between the two different families of models.
In “Dimensional Analysis of Ranking Data” Brady (Reference Brady1990a) developed factor-analytic and ideal-point models and estimators for analyzing data in the form of rank orderings. In a related piece, “Factor and Ideal Point Analysis for Interpersonally Incomparable Data” (Brady Reference Brady1989), he cleverly transformed the apparent limitations of rank orderings into strengths by noting that they are impervious to some of the problems highlighted in his previous work on interpersonal incomparability. For example, “thermometer ratings” of candidates on a 0-to-100 scale seem more informative than ordinal rankings, but may be very difficult to interpret if a rating of “70” means very different things to different survey respondents. Rank orderings, in contrast, are unaffected by linear transformations of each individual's ratings, making them much less subject to statistical biases stemming from interpersonal incomparability. Thus, Brady resourcefully adapted his methods for estimating factor-analytic and ideal-point models with ranking data to serve as methods for estimating those models robustly with rating data.
ELECTORAL POLITICS AND PARTY SYSTEMS
Brady's interest in electoral politics began with the dynamics of choice in multi-candidate U.S. presidential primaries. From there, he looked north to Canada, where multipartism and a peculiar pattern of volatility coexisted with institutions that elsewhere forced bipartism and restrained electoral flux. In the course of studying these complex and fluid electoral settings, Brady and his colleagues developed and implemented powerful new methods for capturing campaign dynamics and produced important new insights into the connection between campaigns and party systems.
Brady's work on primaries harnessed detailed analysis of media coverage and shifting public opinion to broad systemic questions about the workings of the nominating process. He used the primary-season component of the 1984 American National Election Study (ANES)—a “rolling cross-section” survey in which new respondents were interviewed each week from January through June—to measure stability and change in public perceptions of the candidates' traits, issue positions, and chances of success (Brady and Johnston Reference Brady, Johnston, Orren and Polsby1987). Later he built macro-models of the nominating system and used computer simulations to explore the dynamics of alternative processes driven by name recognition and perceived viability (Brady Reference Brady1996).
Brady's interest in Canadian electoral politics began with an explicit comparison of the Canadian and American candidate selection processes (Johnston and Brady Reference Johnston, Brady and Perlin1987). It extended to general elections when Brady joined Johnston, Blais, and Crête in conducting the 1988 Canadian Election Study. The 1988 CES reflected and contributed to a wave of scholarly interest in campaign dynamics, media effects, and opinion formation and change. The critical contribution of the CES team was to take the design pioneered in the 1984 ANES primary study to its logical conclusion—a rolling cross-section survey covering the entire campaign period, including several question-wording experiments and supplemented with a post-election interview. Any survey that is in the field for an extended period will capture temporal heterogeneity; the rolling cross-section design simply converts this bug into a feature, so to speak. But it also means that data analysis must be rethought to take into account that heterogeneity, and here too Brady has been a pioneer (Johnston and Brady Reference Johnston and Brady2001; Brady and Johnston Reference Brady and Johnston2006).
The 1988 CES project changed the basic form of electoral research in Canada and became an influential model for election studies elsewhere, notably in New Zealand, the United Kingdom, Germany, and—through the National Annenberg Election Study—the U.S. It was also a significant catalyst for substantive and methodological work on media and campaign effects; Brady and Johnston's Reference Brady and Johnston2006 edited volume, Capturing Campaign Effects, is an influential sampling of that work.
It helped that the CES team's object of study, the 1988 Canadian election, was an election scholar's dream, with substantial priming of a single key issue, a dramatic and consequential debate among the party leaders, shifting electoral fortunes, and consequently shifting strategic imperatives. The analysis of that election produced by Brady and his colleagues, Letting the People Decide: Dynamics of a Canadian Election (Johnston, Blais, Brady, and Crête Reference Johnston, Blais, Brady and Crête1992), won the 1993 Harold Adams Innis Prize as Canada's best English-language book in the social sciences.
The authors of Letting the People Decide masterfully situated the 1988 Canadian election in its historical and institutional context. Returning his focus to the U.S., Brady (Reference Brady and Rakove2001) similarly situated the 2000 presidential election in American political history, analyzing how the parties' shifting positions on economic and social issues resulted in a striking inversion of the political alignment of 1896.
VOICE AND EQUALITY
Beginning in the late 1980s, Brady teamed with Sidney Verba and Kay Schlozman in a theoretically, empirically, and logistically path-breaking study of political participation in contemporary America. Their project resulted in a series of major articles (including three published in the American Political Science Review) and a classic book, Voice and Equality: Civic Voluntarism in American Politics (Verba, Schlozman, and Brady Reference Verba, Schlozman and Brady1995). Brady's work in the field has continued over the years, including the ambitious essays “Conceptualizing and Measuring Political Participation” (Brady Reference Brady, Robinson, Shaver and Wrightsman1999) and “An Analytical Perspective on Participatory Inequality and Income Inequality” (Brady Reference Brady and Neckerman2004).
The primary empirical basis for this prolific wave of research was the 1990 Citizen Participation Study, an innovative large-scale two-stage survey. In the first stage, a representative sample of 15,000 people was interviewed by telephone. In addition to providing baseline data, the telephone survey allowed the researchers to identify 2,500 respondents for intensive (nearly two-hour) face-to-face follow-up interviews. The second-stage sample was stratified by level and type of political participation (and by race and ethnicity), producing remarkably rich data from a large, diverse pool of political activists.
Verba, Schlozman, and Brady used these data to provide a wealth of detailed description of participatory activities, documenting (for example) how many campaign workers concentrated most of their efforts on local races (50%) and what proportion of those who participated in protests were recruited by people they knew personally (68%). More importantly, they embedded these myriad facts about participation in a fruitful theoretical framework, the Civic Voluntarism Model, which highlighted and integrated the effects of resources, engagement, and recruitment on patterns of political participation.
Verba, Schlozman, and Brady also analyzed the implications of observed patterns of political activity for what they referred to as “participatory distortion”—discrepancies between the political preferences and concerns of the general public and those conveyed to policymakers through various participatory channels. Their nuanced analysis incorporated differences in the nature and extent of inequality in various forms of participation (the most affluent one-tenth of Americans accounted for about 15% of all contacts with public officials, but more than half of all campaign contributions), differences in the specificity of messages conveyed by various forms of participation (letter writing conveys much more specific messages than voting), and differences in the political concerns that impel various forms of participation (taxes and the economy were mentioned as motivating concerns by more than half of all voters but by only 8% of those who contacted federal officials).
In a symposium titled Voice and Equality published in the American Political Science Review, Verba, Schlozman, and Brady (Reference Verba, Schlozman and Brady1997) graciously concluded their remarks by “look[ing] forward to the research that will enhance what we have done, correct it, and inevitably—but, we hope, not too soon—supersede it.” They need not have worried about being superseded too soon. Voice and Equality remains a landmark work for students and scholars of American political behavior and representation. In 2007 the authors received the Philip Converse Book Award from APSA's Elections, Public Opinion, and Voting Behavior organized section; and according to Google Scholar, the book has been cited more than 2,500 times in the 14 years since it was published.
WELFARE REFORM
Brady has a keen sense of the significance of political and policy events and the opportunities they create to advance knowledge. His contributions to the evaluation of welfare reform provide just one example of his ability to see these opportunities and use them to build a broad and inclusive infrastructure for scholarship and applied policy research.
In the mid-1990s Congress passed legislation repealing the six-decade-long federal entitlement to public assistance for American families, devolving substantial policy and administrative authority to state welfare systems. Due largely to Brady's leadership, the state of California created one of the most sophisticated approaches in the country for collecting and analyzing data to evaluate the consequences of these momentous policy changes.
Early in the 1990s Brady began working with the research arm of the California Department of Social Services on a common problem in public service agencies: they collected a massive amount of data on their clients, but these data were kept in multiple, incompatible data systems and the state had no capacity to link client records or use the data to track client outcomes and evaluate policy changes. Brady drew on his extensive experience with large data projects to think through the problem and the possibilities. He put together a team that connected the data wonks from Berkeley's UC DATA archive with the data management and research staff from the state Department of Social Services. Under Brady's direction this team developed one of the first protocols for building analyzable, longitudinal administrative datasets at the state level, linking individual-level client data over time and across agencies. Tapping the capacity of Berkeley's Survey Research Center, Brady expanded the collaboration with the Department of Social Services to field telephone surveys linked to longitudinal administrative data.
The novel approaches that Brady developed for building and analyzing administrative datasets and linking them to individual survey data have served as models throughout the country. They have been described in academic papers and policy reports (Luks and Brady Reference Luks and Brady2003; Brady, Nicosia, and Seto Reference Brady, Nicosia and Seto2002; Brady, Buchler, Jarvis, and McNulty Reference Brady, Buchler, Jarvis and McNulty2001). The data from the California project have been widely used in studies of welfare policy and outcomes by Brady and others. Brady's own publications using these data reflect his methodological expertise, most notably in an in-depth study of disability and welfare use that was the first to employ hazard modeling to correct estimates of welfare program exits for simultaneous entry to disability assistance programs (Meyers, Brady, and Seto Reference Meyers, Brady and Seto2000).
Brady's work with the California Department of Social Services began well before the major federal welfare reforms of the 1990s, and the data infrastructure he created positioned the state better than most states to conduct a rigorous evaluation of the impact of these changes. Nevertheless, it took Brady's active leadership at the state level to make this happen. In 1997 he brought his formidable energy, commitment, and intellectual power to bear in a successful effort to pass legislation to require and fund an evaluation of the CalWORKs program in California. As with so much of his work, Brady's efforts created the infrastructure that has allowed many subsequent scholars, graduate students, and professional research staff to conduct the highest quality scholarship and policy analysis.
COUNTING ALL THE VOTES
November 2000, Brady was one of a team of six political scientists who served as expert witnesses in legal proceedings focusing on the notorious “butterfly ballot” in Palm Beach County, Florida. Large numbers of Palm Beach County voters believed they had mistakenly voted for Pat Buchanan rather than Al Gore because they misread or misinterpreted the complex two-page presidential ballot. Their belief was bolstered by strong circumstantial evidence: Buchanan received a much larger share of votes in Palm Beach County than elsewhere in Florida, and there were also many more multiply-punched (and therefore invalid) ballots in Palm Beach County than elsewhere. Several lawsuits were filed claiming that the confusing “butterfly ballot” was illegal and seeking a county-wide revote. Given the nature and magnitude of the anomaly and the closeness of the overall vote in Florida, it is very likely that a revote would have resulted in Gore winning the state and the election.
Brady and his colleagues described their experience in Florida in the week following the election in an engaging piece published the following March in PS (Brady, Herron, Mebane, Sekhon, Shotts, and Wand Reference Brady, Herron, Mebane, Sekhon, Shotts and Wand2001b). As it turned out, their statistical analyses were irrelevant to the resolution of the legal dispute, which turned on the “clear and unambiguous” language in the U.S. Constitution requiring that presidential electors “be elected on the same day throughout the United States.” Nevertheless, they provided the most careful and widely cited analysis of the actual impact of the “butterfly ballot” on the outcome of the 2000 presidential election.
This work was reported in more detail in a piece by the same authors published later that year in the American Political Science Review (Brady, Herron, Mebane, Sekhon, Shotts, and Wand Reference Brady, Herron, Mebane, Sekhon, Shotts and Wand2001a). While the article employed state-of-the-art statistical methods, it relied most heavily on an impressive variety of straightforward comparisons of voting patterns across U.S. counties, across Palm Beach County precincts, across election years, across offices, and between Election Day and absentee ballots. (Absentee ballots in Palm Beach County did not employ the confusing “butterfly” layout, and they did not display the anomalous Buchanan vote observed in the Election Day returns.) This “strategy of triangulation,” as the authors called it, provided very persuasive evidence that Buchanan's vote in Palm Beach County was significantly inflated—at Gore's expense—by the design of the ballot. It also provided a beautiful example of how careful attention to the substance of a research problem can bolster the persuasiveness of statistical analysis by ruling out potential alternative interpretations of observed patterns in the data.
Having been thrust into the field of election administration, Brady proceeded with characteristic energy and intelligence to examine the strengths and weaknesses of alternative voting systems. In Counting All the Votes: The Performance of Voting Technology in the United States (Brady, Buchler, Jarvis, and McNulty Reference Brady, Buchler, Jarvis and McNulty2001), he and his coauthors analyzed county-level election returns from most of the U.S. in order to evaluate the comparative performance of voting systems using punch cards, optical scans, and other technologies. They also examined changes in performance in places that switched from one type of system to another. Their analysis showed that punch-card systems, which were used by one-third of U.S. voters in 2000, are distinctly inferior to optical scans and electronic systems, producing higher overall rates of invalid votes and larger “racial gaps” between non-whites and whites in the frequency of invalidated votes. Work of this sort has helped to spur and guide the modernization of U.S. voting technology in the wake of the 2000 election debacle.
RETHINKING SOCIAL INQUIRY
In 1995, Brady published a brief review of Gary King, Robert Keohane, and Sidney Verba's Designing Social Inquiry in The Political Methodologist newsletter. That brief review was the starting point for a project that culminated, almost a decade later, in a major volume coedited by Brady and David Collier, Rethinking Social Inquiry: Diverse Tools, Shared Standards (Brady and Collier Reference Brady and Collier2004).
Rethinking Social Inquiry is consciously framed as a response to King, Keohane, and Verba's volume, which Brady and Collier characterized as a “deservedly influential and widely read … claim that quantitative methods can solve many problems faced by qualitative researchers.” Some of the chapters disputed this claim by critiquing King, Keohane, and Verba's “quantitative template,” as Brady and Collier called it. Among these is Brady's own original review, which criticized King, Keohane, and Verba for offering a narrow account of causal inference and for ignoring most of the rich social scientific literature on measurement.
Additional chapters provided examples of “diverse tools,” including case-oriented analysis, process tracing, and “triangulation” combining qualitative and quantitative approaches. The volume concluded with two chapters authored by the editors and Jason Seawright. The first reconsidered the various threads of criticism of King, Keohane, and Verba's “mainstream quantitative” approach from the perspective of statistical theory; it also underlined the importance in practice of tradeoffs between (for example) maximizing the number of observations in an analysis and maximizing conceptual validity and causal homogeneity.
The book's final chapter offered an alternative view of methodology highlighting diverse sources of leverage in causal inference and emphasizing the inferential value of “causal-process observations”—facts and insights that provide information about context or mechanism even if they do not happen to fit into a rectangular data matrix. In a brief appendix, Brady applied the distinction between “data-set observations” and “causal-process observations” to a far-fetched but attention-getting statistical analysis of the impact of television networks' premature call of the 2000 election in Florida on turnout in the Florida panhandle. He drew on “causal-process observations”—most notably, the fact that the networks called the race 10 minutes before the polls closed—to suggest that the assumption-laden regression estimate of 10,000 lost votes was probably off by two order of magnitude or more. Even in the data-rich domain of political behavior, Brady concluded, causal-process observation “provides valuable inferential leverage that supplements, and in this instance contradicts, the conclusions based on the analysis of data-set observations.”
In the years following the publication of Rethinking Social Inquiry, Brady and Collier have been rethinking social inquiry on an even broader scale, most notably in teaming with Janet Box-Steffensmeier as editors of the Oxford Handbook of Political Methodology (Box-Steffensmeier, Brady, and Collier Reference Box-Steffensmeier, Brady and Collier2008). The editors' introductory chapter applied Brady and Collier's arguments in favor of eclecticism in research methods to an apt case: the rise of “causal thinking” in contemporary political science. The handbook as a whole richly illustrates the virtues of methodological eclecticism, not least in Brady's own chapter “Causation and Explanation in Social Science,” in which he described the logic and methodological implications of four distinct approaches to causality—“neo-Humean regularity,” “counterfactual,” “manipulation,” and “mechanisms and capacities”—and provided an insightful discussion of the epistemological strengths and difficulties of the influential Neyman-Rubin-Holland model of causal inference.
POLITICAL MOBILIZATION AND THE COLLAPSE OF THE SOVIET UNION
In the late 1980s, Cynthia Kaplan and Norman Nie conceived the idea of replicating Verba, Schlozman, and Brady's work on political participation in the Soviet Union. Brady and Kaplan were both then at the University of Chicago and Kaplan invited Brady to become part of the project. Little did they know that their panel survey would capture attitudes and behaviors before and after the collapse of the USSR. Although their initial intention was simply to describe political participation in the Soviet Union, the goal of the project expanded to nothing less than explaining the Soviet collapse. Their sustained pursuit of that goal will culminate, two decades later, with the publication of Gathering Voices: Political Mobilization and the Collapse of the Soviet Union (Brady and Kaplan Reference Brady and Kaplanforthcoming).
Brady and Kaplan's multi-level analysis of elite conflict and social movement competition highlights the role of the mass public in the disintegration of the Soviet state. They use detailed events data collected from the Estonian and Russian language press to track the course of political mobilization in Estonia and the competition for popular support between groups calling for the restoration of an independent Estonian republic and those calling for reform within the Soviet Union. A formal model captures the dynamics of this framing competition, and an empirical fitting of it shows that the more radical “restorationist” movement beat out the “reform” movement by emphasizing threats to Estonian culture and ethnic identity.
Extending their analysis beyond the Baltic states, Brady and Kaplan use events data for the entire Soviet Union and voting data from the Congress of People's Deputies of the USSR to show that the votes of representatives in Moscow were strongly related to levels and types of mass protest in the areas they represented. The strongest challenges to the Soviet state came from republics with extensive ethno-national protests calling for sovereignty and independence.
In Russia, the electoral realm and public opinion proved more decisive than social movements in swaying the elite competition between Mikhail Gorbachev and Boris Yeltsin. Brady and Kaplan's painstaking analysis of over 80 opinion surveys from the crucial 1989–91 period relates shifts in public support for these leaders to changing attitudes about economic and political reforms. They also connect Yeltsin's ability to draw support from across the political spectrum—including both democrats and nationalists—to the story of mass protest in the Baltic states by relating Yeltsin's 1990 proclamation of Russian sovereignty to earlier sovereignty proclamations by the Baltic republics.
Intrigued by the role of ethnicity in social mobilization in the Baltic, Brady and Kaplan developed and tested new measures of social and political identity. In “Categorically Wrong? Nominal Versus Graded Measures of Ethnic Identity” (Brady and Kaplan Reference Brady and Kaplan2001) they considered why individuals with the same nominal identity sometimes respond to political appeals uniformly, while at other times exhibiting stark differences. And in “Conceptualizing and Measuring Political Identity” (Brady and Kaplan Reference Brady, Kaplan, Abdelal, Herrera, Johnston and McDermott2009) they proposed an array of measures of social identity focused on relational qualities, constitutive norms, and cognitive models, demonstrating how history, demography, surveys, primary source materials, content analysis, and discourse analysis can address different conceptual aspects of ethnicity and elucidate different types of research questions.
Brady and Kaplan's joint work powerfully merges his data-analytic skills and longstanding interest in the pivotal role of the mass public in social change with Kaplan's expertise in comparative politics and longstanding interest in the role of ethnicity and social identity. It exemplifies rigorous testing of theory through the application of multiple methods employing a wide range of evidence. And it provides an impressive example of how intellectually adventurous and sustained collaboration can produce research that goes well beyond the capabilities of any individual scholar.
INSTITUTION BUILDING
Big social science takes more than brains and hard work—even the combined brains and hard work of diversely skilled collaborators. It also takes money, infrastructure, and institutions. While Brady's professional life has encompassed a remarkable array of path-breaking scholarly projects, he has also devoted prodigious energy and skill to raising the money, developing the infrastructure, and building the institutions that he and other social scientists need to do first-class scientific work.
In his brief stints at Harvard and Chicago, Brady directed the Harvard Data Center and the Center for the Study of Politics and Society in Chicago's National Opinion Research Center. At Berkeley he spent a decade directing the Survey Research Center, a venerable interdisciplinary facility with a dozen resident faculty and scores of students and professional staff. He also served for 17 years as director of the UC DATA archive, where he created the California Census Research Data Center. And he was a co-principal investigator (with Paul Sniderman and Philip Tetlock) in two iterations of the Multiple Investigator Common Vehicle, the precursor of the long-running Time-Sharing Experiments in the Social Sciences (TESS) project.
Brady was a leading intellectual and organizational figure in the early development of political methodology as a subfield, including serving as the fifth president of the Society for Political Methodology. More recently he has served as a principal investigator for the ambitious NSF-sponsored project on Empirical Implications of Theoretical Models (EITM) and host of the 2005 (and forthcoming 2010) EITM summer program. He has also played a leading role in national efforts to improve cyber infrastructure in the social sciences, and in a major NSF-sponsored training program for graduate students in politics, economics, psychology, and public policy at Berkeley.
Brady's previous service to APSA includes stints as a member of the editorial board of the American Political Science Review, president of the Political Methodology section, co-chair of the 2003 annual meeting, treasurer, and vice president.
In all of these roles Brady has been a tireless and effective proponent of scientific rigor, professionalism, and public responsibility. We have no doubt that he will be equally tireless and effective in sustaining these estimable values as APSA's president.

