One of the first papers of Tom Sargent I read was his little “Note on the Accelerationist Controversy” published 40 years ago in 1971. It showed how commonly-used statistical tests rejecting the vertical long run Phillips curve were flawed because they did not take expectations into account properly. Then his 1975 Journal of Political Economy paper with Neil Wallace, which showed that monetary policy was ineffective in models with rational expectations and perfectly flexible prices, made it clear to me that we had to find a tractable way to put sticky prices into rational expectations models. Tom’s 1978 paper with Robert Lucas “After Keynesian Macroeconomics” pointed out many of the problems with Keynesian approach to economic policy. I recently found that our current policy experiences, 30+ years later, confirm that view. Tom’s emphasis on “cross equation restrictions” in rational expectations models set new standards for empirical estimation as he showed in his 1980 paper “Formulating and Estimating Dynamic Linear Rational Expectations Models” with Lars Hansen. Tom has also made his technical research accessible to economics students. His book Macroeconomic Theory published in 1979 is an early example, and last spring we used his more recent textbook with Lars Ljungqvist in the first year Ph.D. program at Stanford. His 1986 book Rational Expectations and Inflation made the technical subjects accessible at a non-technical level.
Chris Sims introduced the use of vector auto-regressions into macroeconomics in his1980 paper “Macroeconomics and Reality.” This work has had a deep and pervasive effect on macroeconomics which persists today. I first used his methodology in a paper published that same year (in the same journal and issue as the Hansen-Sargent paper mentioned above) to demonstrate that the stochastic dynamics of the business cycle in all the major industrial countries could be explained by a combination of a monetary reaction function and a particular form of staggered price setting, revealing a trade-off between output and price stability. Work I did in 1985 on nominal GDP targeting used estimated and theoretical impulse response functions as suggested by Sims. That research indicated that nominal GDP targeting had several flaws and pointed the way to a different kind of policy rule.
Both Chris Sims and Tom Sargent are of the school that you should evaluate policy proposals rigorously with estimated and theoretically well-founded models, rather than just speculate on how a policy would work or did work, and that is also an important model to follow.
Congratulations and thank you, Tom and Chris.