The Krugosphere is hostile to the macroeconomics of the University of Minnesota. I understand that. Krugman has used the term “Dark Age Macroeconomics” to describe what took place between the late 1970s and today. I understand that, too.
But what happened in Minnesota could have stayed in Minnesota. Instead, Stan Fischer and Olivier Blanchard gave MIT’s blessing to DSGE models and vector autoregressions. To me, those two are the real villains.
Had Fischer taken his cues from, say, Clower and Leijonhufvud, rather than from Sidrauski and his ilk, macroeconomists might have spent the last 30 years working on interesting issues and gaining some better understanding of the economy. Instead, they spent the last thirty years diddling with fancy unverifiable equations and pouring a few globs of macro data into the VAR immersion blender.
I’m not sure I follow your hostility to Vector Autoregressions (VAR). VARs are time series models to describe statistical relationships. They are agnostic about economic relationships and specifically, cause and effect.
Dr. Kling, your beef with VARs is that people use them to assert cause-and-effect economic relationships, isn’t it? When in reality VARs allow no such thing (unless you bring in heavy cavalry such as Structural VARs, which are economically meaningful but also giants of clay).
No, my beef is that they are working with imprecise data that is subject to structural change and short-term peculiarities, and they ignore that. Any competent statistician wants to start with clean data. The people who use VAR’s do not give a darn.
What are your favorite papers of Clower and Leijonhufvud that I (in a job that gives me some latitude to take some risk) might take a cue from? I’ve always wondered what the best alternative (to DSGE) is and I think you’re the first to give me a hint!
Why visitors still use to read news papers when in this technological globe all
is accessible on web?