Home > foundations of econ, macro, Micro > Nick Rowe and the inflation fairy

Nick Rowe and the inflation fairy

Nick Rowe is one of my favorite bloggers.   On most days, my only complaint about Nick Rowe is that he’s not posting.   Every time Nick Rowe sends his thoughts into the aether, I feel as though I’ve learned something.

So you can imagine my disappointment in this post from Nick.    This post is wrong, I’m quite sure.    At issue is the post from Yglesias that I linked to yesterday (here).  Not only doesn’t Nick seem to think through Matt’s case, but his response is pure AS-AD model.   I find it unlikely that Matt doesn’t understand AS-AD.   In fairness, Nick is trying to push a pet idea about how we teach economics (the thought being that economics the “non-linear”–not in the mathematical sense, but in the “artistic” sense as in a story that doesn’t follow a linear plot).   “Artsie-nonlinear” you should, of course, read as “equilibrium”.

In addition to agreeing with Matt on this one, I’ve a couple specific problems with this.

  1. It ignores disequilibrium dynamics:   when I say that “signals require realizations”, I am talking about a disequilibrium phenomenon.   In models everyone is on “equilibrium path”, no one deviates.   In the real world, real economic actors need to observe a deviation–or at least observe enough information to predict the costs of deviation–before they know to avoid doing it.    Since Nick has harped on this himself, I would think that he would see it straight off.
  2. Inflation from where?   Macro types like to assume either a Phillips curve, or at least a central  bank reaction function which implies some inflation rate, keeping all other variables constant.   The CB waves its magic wand and inflation appears, it has no other cause.   As a theoretical matter, AS-AD is consistent with all inflation and no RGDP growth,  all RGDP growth and no inflation or anything in between.   In the real world, inflation is the sum of many individual decisions to a) raise prices b) raise wages and each of those decisions has a cause.

For point one, I might add a more general point.   It is not strictly the concept of “equilibrium” that is the problem here.   I would argue, instead, that the problem boils down to the form the equilibrium concept takes–almost trivially, there will be a “correct” equilibrium concept which approximates the real world… although that’s a subject for another post.

In macro, there is an implicit use of the “Bayesian Nash Equilibrium” (really perfect bayesian… but I digress) concept which means, in a nutshell, that everyone’s beliefs are formed through the same process.   In words, I could have the same beliefs as you… to the extent that I don’t, it is because it was not worth my time to look up information in your possession.    There is classic uncertainty, but only classical uncertainty.  We all make optimal use of the information available to the world.

In the real world, we form our beliefs through history, seeing things.   We know the bad guy for the bad guy when the bad guy busts some kneecaps.  We are not certain, until it happens, how the bad-y will act.   The signal badness requires the realization of bad actions.

There is another concept that better matches the real world process of expectations formation including “model” uncertainty:  “Self-Confirming Equilibrium“.   The idea for SCEs is that one’s beliefs on path only have to be consistent with one’s own information (see here, for how this relates to model uncertainty).   In short, signals require realizations.

  1. June 22, 2012 at 3:23 pm

    Sorry you didn’t like my post, but I hope I can remain one of your favourite bloggers?

    You are right that my post is just basic AD/SRAS. It only get’s past ECON1000 when: I discuss how an increase in expected inflation and/or an increase in expected growth will shift the AD curve right; I hint at RE stuff towards the end.

    And yes, this post was really just an excuse for me to play around with the Artsie non-linearity stuff, rather than a full response to MattY.

    I’ve since re-read MattY’s post, after reading your earlier post. I think you are maybe reading into MattY something that isn’t obviously there. But I may be wrong. On re-reading MattY’s post, I think I interpret him as saying that the AD curve is horizontal under inflation targeting, so if the SRAS curve is upward-sloping, you can’t get an increase in Y unless the central bank targets a higher rate of inflation?

    The stuff you are discussing in this post, and your previous, is i think far more sophisticated than what either me or MattY are talking about. And in your previous post, the example you use to illustrate your point is about the effect of expected inflation on the position of the SRAS curve. I (and I think MattY too) were just taking the slope and position of the SRAS curve as given, even though it’s not, of course.

    As to your points in this post, like whether people have to actually observe out-of-equilibrium play in order to form beliefs that help define the equilibrium path….when I was a young guy i used to try to get my head around questions like that. Because I think they are really important in discussing things like social institutions. (And monetary policy is a social institution). But it makes my head hurt.

    • BSEconomist
      June 22, 2012 at 10:47 pm

      First, all you have to do to remain one of my favorite bloggers is to keep saying smart things… no pressure!

      I may be reading more into Matt’s post than he intended, no doubts there. I do that quite a bit. Still, I think his story is consistent with my point. My reading of his post is just that the Fed can control total nominal spending, it can’t control how that is split–the act of employing previously unemployed resources will put upward pressure on prices so that both would rise together. So, yes, that would be an upward slopping AS curve. I don’t recall any direct implication about the AD curve–I read him as saying that the new equilibrium point must move NE or else the Fed’s reaction function is holding back growth. [On second thought, that would be a horizontal AD curve]

      Of course, my idiosyncratic views on this are not really necessary for the issue at hand… just a long winded way of saying that AS curves really do have to slope up. I suppose the point is to construct a quasi-Hayekian understanding of SRAS–I’d like to argue that inflation is related to the information flows. I’m not sure how successful I’ve been.

      Sorry to make your head hurt though.

  1. July 21, 2012 at 1:50 pm

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: