This last weekend Midwest Econometric Group members gathered for their annual meeting at the John Cook School of Business at St. Louis University in St. Louis, Missouri. Midwest Econometric Group is one of the few regional groups dedicated to theory and methods of econometrics in the USA and unlike Econometric Society gatherings its annual meetings are not too crowded, so one actually meets and chats with other fellow researchers and econometricians.
The sessions were organized in a way that micro or macro sessions did not have any conflict. There were a number of sessions on Time Series and Panel data with one session dedicated to finance. The papers were very high quality and the authors emphasized that paper is 70% or 60% complete. Despite this fact most of the papers were motivated by some original ideas. I noticed that there are a growing number of econometricians who dedicate their research agenda to the study of non-parametric models in econometrics. Those researchers who work on Bayesian applications and methodology in econometrics are seeking economic interpretations for different Bayesian concepts and methods. Overall MEG 2007 leaned toward theory more than application. It was a true educational experience.
Peter C. B. Phillips from Yale University was keynote speaker. Many, your correspondent included, enjoyed his talk immensely. He said: “Economic trends are like Hamlet, you do not know what they will do next.” He added:
“We never know what is coming next. Terms such as “if the current trend continues” or “long term trends” do not have any factual meaning. They just mean: if there is some continuation of current points.”
He highlighted how far practitioners have come in their econometrics works. Using Picasso’s famous quote on art he said: “Econometric models are lies that seek to reveal the meaning within data.” I made sure that I wrote down that sentence.
MEG 2008 is going to be in Kansas University. This is truly a great conference to attend.