Saturday, November 17, 2012

Ryan and Gene on RBC

Here and here, respectively.

I think they both make important points, although it looks like the post Gene had about it being a theory of the error term is down... I'd disagree with that a little.

So Gene is right that it doesn't really make sense to call it a cycle theory - it's really just a random downturn theory. But that's true of Keynesians too, and Austrians for that matter. The old accelerator-oscillator is actually a "cycle", but I think most business cycle theories aren't cycle theories. Kind of an interesting wrinkle, I guess.

I do think RBC gets a bad rep in terms of their ideological content. It's precisely because they've been genuinely dedicated to explaining the data that they've (over time) incorporated the frictions and things that Ryan refers to (and as I've mentioned recently - they include more realistic microfoundations, such as including home production in labor supply decisions).

I think one of the biggest weaknesses of RBC is its reliance on calibration. It makes it very hard to assess whether the model is a good way of thinking about the underlying data generating process or whether they have just smashed a round peg into a square hole with a big hammer. Contrast this with even a naive Keynesian model that has some interesting out of sample predictions about some stylized facts, for example, at the zero lower bound - I am much more convinced by that than by repeated ex post calibration. When those naive Keynesian predictions are reaffirmed in New Keynesian models that incorporate all the (in Ryan's words "shitty") microfoundations that an RBC has to offer, I wonder even more what RBC adds.

The point that RBC is just a theory of the error term is both right and wrong. It's wrong no a theoretical level. A shock and a random error are really two different things, I think. But on an empirical level Gene has a point. It's one thing to theorize about how real shocks or technological shocks will impact the economy. It's another thing to not have an actual variable to measure that shock and just assume any perturbaitons are precisely that sort of shock.

This gets into the application of RBC to real life as well. As many have pointed out: exactly what technological shock occurred to throw millions of people out of work in the last five years? It's hard to take that seriously. We have, after all, a massive, obvious, front-page news story demand shock. Why mess with RBC, then?

12 comments:

  1. "With four parameters I can fit an elephant, and with five I can make him wiggle his trunk."

    -attributed to John Von Neumann

    ReplyDelete
  2. I thought the difference between RBC and Keynesian models were the frictions.

    In other words, if you add a demand shock and staggered price setting to an RBC model, you will get a New Keynesian Model.

    I could be wrong, since Macro wasn't my best subject. But if that isn't the difference, what is?

    ReplyDelete
    Replies
    1. I think this is more or less correct, which is why I don't understand some of the fist shaking.

      Delete
    2. To point out it is not a cycle theory is not "fist shaking."

      Delete
    3. You have a strange definition of "cycle theory" (as do the people you cited). The empirical relationship that we seek to explain is increases in unemployment and stagnations in growth. Whatever can theoretically explain those two things are cycle theories. The methodological peculiarities you pointed out are present in all these models. The reason RBC is wrong is because it is wrong empirically, not because it is not a genuine cycle theory.

      Delete
  3. it's really just a random downturn theory. But that's true of Keynesians too, and Austrians for that matter.

    You'd better hope Greg Ransom doesn't see this. He is not as forgiving as I am.

    ReplyDelete
    Replies
    1. So I would think for a "cycle" you'd need something that actually cycles back and forth and back and forth. ABCT explains how a shock can propagate in a boom and then a bust, but then it's sort of over.

      You'll find some people who will tell you that'll just keep up as a result of frau... I mean... fractional reserve banking. Sure, I'll grant that they are making a cyclical argument. But I think the way most people think about ABCT, it is an explanation how you can have an upswing and then a downswing, not a continuous cycle.

      Delete
  4. I think Austrians and Minsky, for instance, have genuine cycle theories, and my colleague and I are going to demonstrate this.

    This does NOT mean they are better for it! If there really IS a business cycle, then a theory that says there really is is better. But if there ISN'T a business cycle, then a theory that says there isn't is better!

    ReplyDelete
    Replies
    1. Minsky I see as a cycle theory. Austrians that think fractional reserves are inherently destablizing, I can see as a cycle theory too (if that's what you mean). I'll take a look at that paper you sent.

      Of course we could also have an actual cycle, but still have a lot of the events we care about (Great Depression, early 80s, 2008, etc.) caused by non-cyclical causes.

      Delete
    2. ABCT, in my understanding, is based upon the effect of the interest rate on the structure of production, not necessarily FRB or even QE, but rather only mentioning their influence upon the rate of interest as compared to the structural reality of production and the movement, use, and exchange of goods over time. Certainly, it mentions money, but it is primarily a macro theory of production over time, and how this is coordinated and/or discoordinated.

      I'm not an economist, but that is how I've always seen it (a macro-time theory of production). More important is that it is entirely consistent with economic first principles and laws.

      Delete
    3. Joseph -
      Right, but some people talk about fractional reserve banking itself as distorting from what you call the "structural reality of production". Not all do, of course.

      Delete
    4. I think some ABCT theories are cycle theories and some aren't. Lots of them rely on the central bank pulling the interest rate down to start the boom and subsequent bust. Those are only cycle theories if the central bank's actions are treated as endogenous, i.e. if the central bank is considered within the theory. If not then it's really a shock type theory.

      I prefer to think about it as a shock type theory than a cycle type theory.

      Delete

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