What Alan Blinder and Mark Zandi Don't Get

What the Blinder-Zandi paper does is explore the properties of a macroeconometric model. The economics profession abandoned those models thirty years ago, so the tool they are using is like a fossil, frozen in time. Of course, there have been many tweaks over the years, but my guess is that if I had access to the full model I would feel like I was returning to what I first worked on when I worked for Blinder thirty-five years ago.

The model assumes a Keynesian world, in which labor is a variable factor of production that responds to incremental increases in aggregate demand. That might be an excellent assumption for 1910, when 73 percent of the work force was blue-collar. By 2000, 73 percent of the work force was white-collar. See Wyatt and Hecker. In today's Garett Jones economy, labor acts more like a fixed factor. Blinder and Zandi do not know this (they may know it, but I doubt that it is incorporated into the model). So they do not know about jobless recoveries, breakdowns in Okun's Law, the high ratio of permanent job losses to temporary layoffs, etc. Instead, at best they are living in 1970, with some add factors thrown in to get the model to track recent data.

They do not know about lots of new research into labor dynamics, which shows that the rates of job creation and destruction are enormous relative to the net gains or losses in employment. See Davis, Faberman, and Haltiwanger. Again, Blinder and Zandi may know about this research, but I can almost guarantee you that their model does not.

If macroeconometrics were a viable paradigm, we would have seen major efforts to try to bring this sort of model up to date from its 1975 time warp. However, for reasons I have documented, the profession has decided that this macroeconometric project was a blind alley. Nobody bothered to bring these models up to date, because that would be like trying to bring astrology up to date.

On the one hand, you're saying that Blinder and Zandi's model could be better, but on the other hand, you're discrediting macro models altogether. It's a bit of a contradiction.

Link to "reasons I have documented" is broken. Needs a http://

As to the content: oh, please. Apply that level of rigor to your pet theory of Garrett Jones labor, which comes from a tweet. Empirics is for other people, clearly. But who needs econometrics when you already have a narrative to suit your political conclusions!

It looks like he's referring to this essay on that last link http://arnoldkling.com/essays/macroeconometrics.doc

It's kind of funny, kind of pathetic, to see an old model that was abandoned being used in current debates. As if those hundreds of really smart people let the model die even though it worked.

Seriously, if we're going to be all "we know these models are wrong, why are we still using them?" please realize that this destroys virtually everything. Yes, even your own pet right-wing RBC or neoclassical or whatever models. Everything - from Solow to general SMD equilibrium to whatever - we know these are wrong. We can't even pretend that they might be right, we actively know that they have to be wrong.

So when you propose a higher level of rigor, make sure your own pet theories meet them first!

Read Full Article »
Show commentsHide Comments

Related Articles

Market Overview
Search Stock Quotes
Partner Videos