Wednesday, July 31, 2013

Sumnerians talk funny

I've been having a conversation on my post about Evan Soltas with "aaron". He's commented here before, but infrequently such that I didn't quite remember where he's at - what his views are, what his level of understanding of economics is, etc.

And at first in this conversation I got the impression he just didn't know what he was talking about. He was acting as if I had been arguing that low interest rates indicate a loose monetary stance, which of course they don't (if anyone doesn't understand why, please let me know). After a lot of back and forth I realized he is not dense... he's just a Sumnerian. And Sumnerians talk funny and it leads them to make these wild claims about Keynesians and Wicksellians generally that don't make much sense unless you assume that Keynesians and Wicksellians are speaking Sumner's language (which of course they aren't).

Anyway, I can't be the only one that's come across this sort of thing from Sumnerians, so I thought you might be interested in my penultimate response to him:
"You've gone around and around on this enough that it's clear you're coming from Sumner's perspective. If that's the case then you need to understand that most of this is semantics and you shouldn't be running around accusing people of not understanding that the level of interest rates is not the same thing as the tightness or looseness of policy.

Some people do genuinely make this mistake. Evan [Soltas] doesn't and I don't, but I bet you fifty bucks the Wikipedia page on "monetary policy" does. So it goes.

Now, if we set those people aside for a second there are two groups of people: most people who think in Wicksellian terms, and a small amount of people who think in Sumnerian terms.

The Wicksellians say that what determines the stance of monetary policy is the relation of the interest rate to the natural rate. You can have situations like right now where the interest rate is quite low, but it's high relative to the natural rate and so monetary policy is tight. We would like to have looser policy to close the gaping output gap.

Then there is this small cadre of Sumnerians who mash up the Wicksellian point with the NGDP point. This, I think, is a bizarre way to talk about things particularly if the degree of efficacy of monetary policy is an open question. It amounts to determining whether we've taken an action based on whether that action was successful, which is pretty stupid IMO.

If you want to make the case for talking like this, fine. But don't accuse a Wicksellian of not understanding that interest rates alone can indicate the stance of monetary policy. It makes you look like you don't understand the nature of the conversation."

11 comments:

  1. I have noticed too that market monetarists seem to misunderstand/ignore Wicksell and also credit view ideas that focus on the unique role that banks play in the transmission mechanism. My criticism goes straight to the top, as I've seen Sumner totally butcher Wicksell when paying him lip service. For example he thinks that the natural rate is synonymous with real interest rates.

    This is a big turn off to me, since I am a Wicksellian/Austrian/free banker that is influenced by Gurley & Shaw and Tobin.

    ReplyDelete
  2. First, I do understand the difference in semantics, but I also think that getting the semantics wrong makes the entire rest of the discussion sub-optimal.

    I legitimately do not understand how you can say, "It amounts to determining whether we've taken an action based on whether that action was successful, which is pretty stupid IMO," and talk about the natural rate of interest or the output gap in the same section of text. At least we can "see" NGDP, or the price level or something even better than either of those, and thus do something vaguely scientific based on this viewpoint. You used to talk about phlogiston not being bad science, and it wasn't, until it was replaced by something that could actually be seen. To insist on still talking about chemistry using the language of the phlogistonite would be absurd.

    P.S. MS in Economics, professional financial economist, broadly Market Monetarist (though closer to Christensen than Sumner himself), and broadly BHL. You could have just asked.

    ReplyDelete
    Replies
    1. I am not saying don't talk about NGDP - I am saying don't use NGDP to state what the monetary authority is doing. Use the behavior of the monetary authority to do that.

      There was no particular need to ask, my only point was that with unfamiliar commenters I tend to learn about them as the conversation progresses, that's all. I don't have a need to keep a file on them or anything.

      The only reason I brought it up is that your initial point seemed so out of left field I didn't know where you were coming from.

      Delete
    2. So here's a question - could I use those same NGDP figures to say that fiscal policy is too tight? If not, why not?

      Delete
    3. I think so, yeah.

      I'm not sure what fiscal policy being too tight would mean, though. Since monetary policy has won every time it had gotten into a fight with fiscal policy, I'm not sure why you would even talk about fiscal policy being too tight or too loose. I think of it like as something not unlike a Stackelberg response function, fiscal policy does something and monetary policy then accepts that as a given and determines the level of "money"/credit/etc needed to get to a certain level of NGDP. Ideally, this is conducted by a market-based mechanism (Free Banking or the Chuck Norris effect). Either way though, fiscal policy has no actual impact on NGDP unless monetary policy is asleep or negligently bad (coughDraghicough).

      To the first reply:
      I'm not sure what you mean, don't use NGDP to state what the monetary authority is doing. I will try to craft a response based on what I think you mean. Ideally, a monetary authority would just a set a rule (and ideally a good rule) and then the monetary authority is effective when it hits that goal and ineffective when it doesn't. Without an explicit rule, the monetary authority has to do a lot more work and they are effective when they hit what that goal should be and ineffective when they don't hit what that goal should be. The Fed should stabilize aggregate demand (read: Nominal GDP), the fed is effective when it stabilizes NGDP and ineffective when it doesn't. The easiest way to get there would be to just promise to get there and do anything you need to make sure the market knows you mean it. If you're not going to do that, if you insist on being ad hoc about it all, then you do whatever you need to do; buying mortgage bonds seems to work okay - the Fed has been able to keep annual NGDP growth stable at 4%. Buying mortgage bonds may only be able to deliver 4% NGDP growth though, and if so, the Fed should do something else. I'm not terribly picky about the mechanism if the result is correct (though I have a pretty good idea about what the best mechanisms would be).

      Delete
    4. re: "deally, a monetary authority would just a set a rule (and ideally a good rule) and then the monetary authority is effective when it hits that goal and ineffective when it doesn't"

      OK, but this is really the heart of my point: "effective"/"not effective" and "loose"/"tight" are two entirely different things. This sentence sounds fine to me. What's odd is when Sumnerians talk as if unsuccessful monetary policy is untried monetary policy.

      This is quite different from, say, the Taylor Rule, which is an actual rule about what the right monetary policy consists of. An NGDP rule defines goals, not policy.

      It would be like me using the goal of publication in the AER as an indicator of whether I wrote today. Writing today may or may not get me a pub in the AER. It would be great if it did. But I wouldn't say "Daniel didn't write because he didn't get in the AER". That is confusing the activity with the desired result of the activity.

      Delete
    5. The problem with the metaphor is that the Fed can get something they wrote published in the AER without actually writing it.

      The Fed recently tightened money significantly just by Bernanke giving a speech.

      Delete
    6. I don't follow. Offering guidance (tight or loose) is as much part of monetary policy as messing with interest rates through bond markets. How is this not like writing?

      He could have tightened money without it impacting NGDP in the way that we all hope it will. That's the point - he could have loosened monetary policy without it having the desired result.

      The point is you and Sumnerians generally really ought to separate out the action from the result and stop pretending they're the same thing.

      Or at the very least don't treat people like they're misunderstanding things because they don't talk in your weird way.

      Delete
  3. "Then there is this small cadre of Sumnerians who mash up the Wicksellian point with the NGDP point"

    I didn't read your exchange with aaron but I think Summner's view would be that if you keep NGDP growrh stable then you will allow the rate of interest to gravitate towards its natural rate. I think this is consistent with Wickselsian (and Hayekian) theory. When Market Monetarists say things like "low interest rates are often indicative of tight money" they don't mean that the natural rate is negative and even a zero rate is high. They mean that tight money has caused NGDP to fall below trend and affected expectations so much that even at zero interest rates no-one wants to borrow to invest. The solution is to commit to increasing the money supply sufficiently so that NGDP is back on trend. If this is done correctly then expectations will be restored and the interest rate will return back to a level consistent with the natural rate and the economy will be out of recession

    ReplyDelete
    Replies
    1. I do think in the end it's all fairly consistent - that's why I say in the comment it's largely a matter of semantics. But clearly it's different enough that aaron can find himself under the impression that I'm saying low rates mean loose policy!

      re: "They mean that tight money has caused NGDP to fall below trend and affected expectations so much that even at zero interest rates no-one wants to borrow to invest. The solution is to commit to increasing the money supply sufficiently so that NGDP is back on trend. If this is done correctly then expectations will be restored and the interest rate will return back to a level consistent with the natural rate and the economy will be out of recession"

      Right and this is precisely what Keynesian mean too. The only difference is that we are not so sure of the omnipotence of the Fed in practice that we don't want to put all our eggs in one basket. But this is not inconsistent with the Wicksell-Keynes way of talking.

      Delete
  4. "You can have situations like right now where the interest rate is quite low, but it's high relative to the natural rate and so monetary policy is tight."

    The "natural rate" is tossed around a lot with different meanings. I assumed in the past that you meant "in the absence of Fed intervention" or "if the economy was healthy". Neither of those definition work with the quote above, which always confused me. Now I get it.

    For those wanting a short summary of Wicksell's definition of natural rate:
    http://research.stlouisfed.org/publications/mt/20050301/cover.pdf

    ReplyDelete

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