Multiple Time Series Models

Many analyses of time series data involve multiple, related variables.  Multiple Time Series Models presents many specification choices and special challenges.  This book reviews the main competing approaches to modeling multiple time series: simultaneous equations, ARIMA, error correction models, and vector autoregression.  The text focuses on vector autoregression (VAR) models as a generalization of the other approaches mentioned.  Specification, estimation, and inference using these models is discussed.  The authors also review arguments for and against using multi-equation time series models. Two complete, worked examples show how VAR models can be employed. An appendix discusses software that can be used for multiple time series models and software code for replicating the examples is available.

Key Features

• Offers a detailed comparison of different time series methods and approaches.
• Includes a self-contained introduction to vector autoregression modeling.
• Situates multiple time series modeling as a natural extension of commonly taught statistical models.

Patrick T. Brandt is an Assistant Professor of Political Science in the School of Social Science at the University of Texas at Dallas.  He has published in the American Journal of Political Science and Political Analysis.  He teaches courses in social science research methods and social science statistics.  His current research focuses on the development and application of time series models to the study of political institutions, political economy, and international relations.  He received an A.B. (1990) in Government from the College of William and Mary, an M.S. (1997) in Mathematical Methods in the Social Sciences from Northwestern University, and a Ph.D. (2001) in Political Science from Indiana University.  Before joining the faculty at the University of Texas at Dallas, he held positions at the University of North Texas, Indiana University, and as a fellow at the Harvard-MIT Data Center.  

John T. Williams was Professor and Chair of the Department of Political Science at University of California, Riverside. He taught time series analysis at the Inter-university Consortium for Political and Social Research Summer Training Program for over ten years. His work uses statistical methods in the study of political economy and public policy. He co-authored two books: Compound Dilemmas: Democracy, Collective Action, and Superpower Rivalry (University of Michigan Press, 2001) and Public Policy Analysis: A Political Economy Approach (Houghton Mifflin, 2000). He published over twenty journal articles and book chapters on a wide range of topics, ranging from macroeconomic policy to defense spending to forest resource management. He was a leader in the application of new methods of statistical analysis to political science, especially the use of vector autoregression (VAR), Bayesian, and event count time series models. He received a B.A. (1979), an M.A. (1981) from North Texas State University, and a Ph.D. (1987) from the University of Minnesota. Before moving to Riverside in 2001, he held academic positions at the University of Illinois Chicago (1985-1990) and at Indiana University, Bloomington (1990-2001).