Thomas Sargent and Christopher Sims were awarded the Nobel Price in Economics today.
When Robert Lucas won the Nobel Prize in 1995 I wondered if Professor Sargent and Professor Sims could win the prize. Certainly, their contributions were important. Professors Sargent and Sims, along with Professor Lucas, published in the Macroeconomics field, an area I was studying. While I did not research the Nobel committee’s search and decision process, I decided that Sargent and Sims were unlikely to win. Most likely, the Nobel committee would find Sargents’ and Sims’ work too technical or too hands on. It lacks the general applicability to broad human goals of, say, Lucas’s “On the Mechanics of Economic Development”.
I was wrong, and I am happy to be wrong. I have been a student of Professors Sargent and Sims work since about 1992. I still re-read certain articles of theirs. I also benefit from their contributions in my daily work as a forecaster of the United States economy and as professor of economics.
One of the many contributions of Professor Sargent’s that I use in my work includes suspicion of the Phillips Curve as a useful empirical device. See his early article here. I have always felt that the Phillips Curve attempted to measure a relationship that was too complex to measure with just one linear equation. Factors that drive the unemployment rate include but are not limited to: population dynamics, household labor force choice, the business cycle, sectoral changes, productivity, and macroeconomic policy choices. Factors that drive the inflation rate include but are not limited to: commodity markets, expectations, demand and supply pressures, technology, anticipated monetary policy, and unanticipated monetary policy.
When I build my forecast models, I do not explicitly link unemployment and inflation in a Phillips Curve. I try to provide a reasonable measure of the factors listed above for each of these indicators independently. This implies that sometimes there is a tradeoff and sometimes there is not. While it is more often the case that the resulting forecast maintains a tradeoff but there are times where the tradeoff does not exist. I remind the reader than for much of 2008, inflation was rising and the unemployment rate was rising.
As an economic forecaster it is impossible to do my job without using one of Professor Sims’ contributions, the Vector Autoregression model, or VAR. This model-building technique, one that was forcefully brought to macroeconomists’ attention by Sims in 1980, has become an important competitor to the main alternative method of forecast model-building, the structural approach. The VAR approach is in some ways easier to implement and perhaps more importantly, is an approach that “lets the data do the talking” more than the typical structural approach. The VAR approach is agnostic to model structure, attempts to capture any and all inter-related forces among all indicators, and provides a convenient testing and diagnostic apparatus for examining the properties of the model economy.
I use the VAR approach for some of my work, and I teach the VAR approach as an important tool in our M.S. in Economics program here at California Lutheran University.
While I am happy for Professors Sargent and Sims, perhaps I am the biggest winner, as I benefit from their insights any day that I am in the office.