Policy makers, public managers, and program professionals face pressing questions: Is my policy successful? Is my organization effective? Does my program work? They are likely to place a high priority on obtaining answers that are unambiguous and positive. To be relevant to mainstream policy discourse, the policy research community is drawn to answering such questions in specific policy or program contexts, using a variety of methods, including randomized experiments, quasi-experiments, and cost-effectiveness analyses.Of less general political interest are questions concerning how public governance and management contribute to governmental outcomes: Do the designs of policies and programs affect their performance? Does management matter and, if so, how? Are particular organizational structures better than others for accomplishing certain goals and, if so, for whom, when, and where? Answers to these kinds of questions, which are the subject of this symposium, allow practitioners to assess how they might replicate success (or avoid failure) at other times, in other places, or under different circumstances; or how poorly performing programs or agencies might be brought up to the standards of the most successful. Thus, such answers are of practical importance to public managers because they aim to provide insights beyond the context of a specific policy or program.The answers to such questions, however, are more complicated, take longer to explain, and have more qualifications than answers to the question, "Does my program work?" (which itself, of course, may be difficult to answer). Producing these answers is also a challenge to scholarship: Convincing and valid insights require an understanding of complex relationships in human, social, and organizational affairs. The use of theory in identifying these relationships is critical, but theories must be carefully tested rather than assumed or asserted to be true.The contributors to this symposium are members of an expanding network of researchers interested in questions of governance and public management using the theories, models, methods, and data of the social and behavioral sciences.1 To promote rigorous research designs and the prospects for cumulating knowledge about 1 One or more of the authors of each paper also participated in the workshop "Models, Methods, and Data for the Empirical Study of Governance and Public Management," held at the University of Arizona in May 1999. Contributions to that workshop were collected in Heinrich and Lynn (2000). The papers in the current symposium explicitly build on intellectual foundations established at that workshop. Thus, they constitute a second generation of contributions-employing more refined models, methods, and data-to an expanding project on public governance.