Christopher Sims

Bio/Description

Christopher “Chris” Sims, the John J. F. Sherrerd ’52 University Professor of Economics, transitioned to emeritus status in July 2021. Chris was born in 1942 in Washington, D.C., and grew up there and in Germany, Virginia, and Connecticut. He graduated from Harvard College in 1963, where he majored in mathematics. He began his Ph.D. studies in economics at the University of California-Berkeley, but returned to Harvard to finish his Ph.D. (and meet his wife Cathy). Chris remained at Harvard University for two years as an assistant professor from 1968 to 1970. He and Cathy then moved to the University of Minnesota, where Chris taught for the next twenty years. They moved to Yale University in 1990 before joining us at Princeton in 1999. 

Chris’s most well-known work focuses on determining the dynamic effects of economic policies on the macroeconomy. A leading example involves quantifying the effects of monetary policy actions of the Federal Reserve on employment and inflation. The key challenge in determining the size of these effects is that, while the Fed’s actions affect employment and inflation, they are also affected by employment and inflation. Persuasively sorting out the direction of causality and isolating the dynamic effects of Fed actions requires analyzing historical data using just the right mix of economics, institutional detail, and statistical theory. Chris showed how this could be done. He developed a statistical framework, the Structural Vector Autoregression (SVAR), that has become the standard for estimating dynamic causal effects in macroeconomics. He was awarded the Nobel Prize in 2011 for this fundamental contribution. 

In his early work, Chris experimented with single-equation models (distributed-lag regressions) to estimate the dynamic effect of one variable (say, the money supply) on another variable (say, employment), using the timing of changes in the variables to help sort out cause and effect. These experiments suggested complicated patterns of causality requiring dynamic models with multiple equations, and Chris began experimenting with vector autoregressions (VARs). These high-dimensional statistical models presented novel challenges. Standard frequentist statistical methods performed poorly when applied to these large models, and that led Chris to develop practical Bayes procedures for analyzing VARs. His work on Bayes methods extended well beyond VARs and was instrumental in educating economists more generally about the power of Bayes methods. These are now standard tools in empirical economics. 

Chris’s work showed that VARs (often when combined with the Bayes methods he proposed) provide a practical and flexible way to characterize economic time series data. This led to three distinct strands of research. The first used VAR models for forecasting. In the early 1980s, Chris developed a VAR forecasting model for the Federal Reserve Bank of Minneapolis that became a template for related models developed by other public and private entities that needed macroeconomic forecasts. Chris’s VAR model not only provided a forecast for each of the economic variables in the VAR, but also provided quantitative measures of uncertainty of these forecasts. 

In the second strand of research, Chris showed how VAR models could be used to sort out cause and effect by linking policy actions to the forecast errors computed from a VAR. Here, economic theory and institutional knowledge is needed to describe how policy actions lead to forecast errors, but once this done, the resulting Structural-VAR can be used to compute the dynamic effect of the policy on all of the variables in the VAR. Countless researchers have subsequently used versions of Chris’s method to estimate policy effects. 

The third strand of research follows from Chris’s demonstration that VARs provide good descriptions of the dynamic patterns in economic time series data. VAR models are now the benchmark by which other dynamic models in economics are evaluated – a researcher who develops a model that fits the data “nearly as well” as a VAR can claim success. While Chris’s work on VARs is his most influential, he has made other fundamental contributions. Two that stand out are his work on the role of government debt for determining the price level and his work on rational inattention in decision making. 

Beyond his own research, Chris has made other important contributions to the economics profession. He rightly has the reputation as one of the most interesting discussants at economics conferences. He has served on countless committees, advisory groups, editorial boards, etc., and served as president of two of the leading professional organizations, the Econometric Society in 1995 and the American Economic Association in 2012. 

Here at Princeton, Chris has been an extraordinary advisor to both Ph.D. students and undergraduates. He has twice served as the Department of Economics’ director of graduate studies, and for seven years he served as co-chair of the Griswold Center for Economic Policy Studies. 

While our students miss Chris in the classroom, we are happy to report that he otherwise continues his usual work schedule in the department. 

Written by members of the Department of Economics faculty.