Friday, April 13, 2012

Andrew Bossie is skeptical of Barro-Gordon too

A couple posts back I vented about the Barro-Gordon model. It's pretty goofy stuff. It turns out that when you observe in the wild a assume the existence of a central banker that is completely clueless about the fact that people form expectations by inferring from past experience, that central banker will act on their ignorance, thus underestimating the dangers of inflation through its impact on inflation expectations. And when you discover make it your explicit modeling assumption that central bankers underestimate the consequences of inflation, it turns out you get the result that they over-inflate!

Shocking stuff, right!

You all are amazed, I know. Well get this. If you think assume the central banker will over-inflate, one solution might be to hire a curmudgeon who doesn't give a damn about the social welfare function so that his stodginess will at least partially outweigh his ignorance the modeling assumption of central banker ignorance, and get something closer to socially optimal monetary policy.

That is "conservative central banker theory" in a nutshell. I don't think I'm exaggerating but a Barro-Gordon defender can correct me if they so choose.

It turns out, Andrew Bossie has an issue with this stuff too. He recently brought my attention to an older post of his that makes a similar point w.r.t. Bernanke. He writes of the theory:

"I have never really liked the claim that you need an overly conservative central banker to maximize social welfare because your normal run of the mill central banker will over inflate. This always bothered me. There always seemed to me to be a fair amount of intellectual path dependence to that finding in the 1999 "Science of Monetary Policy" paper. I'm just not sure how you come to that conclusion without looking for it. Like, it seemed to me the New Keynesians were looking to prove the ad hoc finding that you need a central banker who's DNA tells him to keep inflation low."

This pretty much sums up our current political rock-and-hard-place in the context of a "conservative central banker":

"No securities, no stocks, no bonds, nothing but a miserable $500 equity in a life insurance policy. You're worth more dead than alive. Why don't you go to the riff-raff you love so much and ask them to let you have $8,000? You know why? Because they'd run you out of town on a rail."




Nothing teaches economics at Christmas time quite like It's a Wonderful Life.

2 comments:

  1. It turns out that when you assume the existence of a central banker that is completely clueless about the fact that people form expectations by inferring from past experience, that central banker will act on their ignorance, thus underestimating the dangers of inflation through its impact on inflation expectations.

    Are you sure that's right, Daniel? I never read the original paper(s), but the New Keynesian models we learned at NYU on this stuff all had rational expectations. There was nothing about a central banker being "completely clueless." He knew exactly how people would react to his actions, they knew his preferences and the policy rule, and it just so happened that it was a better equilibrium if he hated inflation more than the rest of society did. Your description doesn't sound right to me.

    ReplyDelete
    Replies
    1. So we worked from lecture notes. I looked through the paper this morning, and all our equations look the same as in the notes: http://www.sfu.ca/~kkasa/barro83.pdf

      I should probably look closer. Agents have rational expectations, that's true. I'll take a closer look at this later.

      Delete

All anonymous comments will be deleted. Consistent pseudonyms are fine.