Wednesday, November 17, 2010

On inflation, nominal expenditures, investment and QE2

David Beckworth has a great post up on QE2, where he basically takes Lee Kelly's side:

"A key objective of QE2 is to raise nominal spending. Given sticky prices and excess economic capacity, this increase in nominal spending would boost the real economy. Now the way QE2 can raise nominal spending is by raising inflation expectations. Higher inflation expectations creates the incentive for banks, firms, and households sitting on money to start spending them. In fact, the key problem facing the U.S. economy right now is an excess money demand problem. While it is true that rising inflation expectations may temporarily lower real interest rates too and thus stimulate interest sensitive spending, this effect is of second-order importance and is only temporary."

Robert Murphy recently mused on Keynesian reactions to QE2, and other people have assumed some nefarious plot at the root of the Keynesian embrace of QE2 as well (I have comments on both). Murphy, I think, is on the right track. Caplan is simply paranoid. Either way I don't think this is that hard to understand, and the Beckworth post frames it nicely.

Lee Kelly and Beckworth think that what Keynesians think is a first order effect is actually a second order effect.

I genuinely don't know how to weigh these things against each other, so I note with Keynes that investment demand is what's really ****ed up, so lowering rates to increase investment is the "first order" effect, but the inflation is nice for other reasons too.

And there's a range of Keynesians that sound more like Lee Kelly and Beckworth (DeLong seems to be on that end) or sound more like Keynes (Krugman seems to be on that end). But as long as I'm writing #1 by one effect and #2 by another effect and Lee Kelly is putting #2 by one effect and #1 by the other, of course you're not going to see that much clashing. We're in the middle of an economic depression. Academic quibbles like that are interesting, but simply not worth it.

We've got an incoming Congress that will be hamstrung, if not controlled, by a weird libertarian-populist hybrid of a Republican party and we finally have a public institution they can't touch (for now at least) willing to serve the American people with an action that leaves a lot of us shurgging our shoulders and crossing our fingers, but can nevertheless get an impressive cross-section of liberals and conservatives nodding their heads in agreement. Conservatives. Like - really live conservatives. I feel like I haven't seen one of them since 2006! Lately they've been morphing into populists or libertarians or populist-libertarian hybrids. But now Bernanke (aka "the Bernank") has managed to smoke out a few actual conservatives with a few sensible things to say that liberals and centrists can be a party to.

Who the hell cares that Lee Kelly and David Beckworth call second order what some of us might call first order? We finally have a growing chorus of people that don't descend into Don Boudreaux-style mangling of the equation of exchange or Jeffrey Tucker-style swooning at rising bond yields. It's about time.

UPDATE: I really want to emphasize that I am an almost exclusively self-taught Old Keynesian with lots of sympathy for the kinds of things Lee Kelly, Sumner, and DeLong have been saying. My graduate macro course did a lot of other stuff that has very little application to what we're going through now (waterzooi can attest to this... OLG models with that professor that late at night could be a harrowing experience sometimes). I really am not sure how to rank these effects, but I just figure it's important to point to the classic Keynesian effect of monetary policy since the blogosphere is pretty saturated with more monetarist and new monetarist understandings. They may well be right in terms of prioritization, but I think it's well worth pointing out the way the story has been told by others.


  1. There is an excess demand for money; one consequence of resolving that excess demand will be higher measures of nominal income (like FINSAL or NGDP). If the Fed's notes were redeemable in gold, then another consequence would be more redemptions at the Fed's gold window. Monitoring these type of consequences can help the Fed gauge when the excess money demand has been satisfied.

    The Fed increases the supply of base money by purchasing assets -- primarily government bonds. This increases the quantity of bank reserves and, when the seller is not a bank itself, increases aggregate cash balances. How this all shakes itself out in terms of inflation, interest rates, investment demand, etc. all depends on how the recipients spend the money, to what degree there is idle capacity in the economy, if there have been any shifts in the economies' time preference, etc. So long as nominal income hits the Fed's target (a 2-3% level target would be good, in my opinion), nothing else really matter.

    I might predict that real interest rates will fall, investment demand will pick up, etc., but if I am wrong and nominal income hits its target anyway, then was the policy a failure? No, because the particular predictions are not really important. Does that make sense? I have no idea if Beckworth agrees with me or not.

  2. It's interesting how the blogosphere debate has been crystalising recently. Participants are discovering more common ground, or at least mutual understanding -- it's getting more difficult to brand opponents as stupid and/or evil. Intellectual divisions that once appeared substantial now seem more like semantic quibbles, and the terms and emphases of different sides appears to be converging.

    The so-called quasi-monetarists like Sumner, Beckworth, and Woolsey have been good in this regard, because they borrow from multiple traditions. In one place they seem Austrian, another monetarist, and then Keynesian. Of course, I don't know what influence all this has had on the larger public debate ... though NGDP targeting was mentioned at a Fed meeting!

  3. I think that's very well stated. I suppose my repeated point is simply that fiscal policy takes a lot of the uncertainty out of your "how it all shakes itself out" point. My major concern with guys like Scott Sumner (and your analysis here) is that there's a gap between nominal expenditures and actual output and employment growth. I can see the logic of why it could work without a problem, but there seems to be a lot of "let it all shake out" finger crossing.

    It's interesting how the blogosphere debate has been crystalising recently. Participants are discovering more common ground, or at least mutual understanding -- it's getting more difficult to brand opponents as stupid and/or evil

    I think so. But I don't think it's just the discovery of common ground - it's also the creation of common ground. People are learning, and that is excellent.

  4. I should emphasise that, like Daniel, I am self-taught. I just comment on internet blogs. I don't teach or publish.

    By the way, why do you always write my forename? You established "Lee Kelly" right at the beginning, so why not just write "Kelly" thereafter? I am not complaining ... just curious.

  5. I don't know... it might have something to do with the fact that that's your user name. I used to refer to Jon and Mattheus as Catalan and von Guttenberg but that sounded weirdly formal... that, to a certain extent, explains why I dispense with "Kelly" but it doesn't explain why I don't say "Lee".

    Very odd... I never even noticed. I think only knowing you virtually probably makes it sound less weird to my ears than yours - it's just a handle, you know?

    Alright - Lee it is. That is weird and now you've brought it to my attention and it sounds weird to me.

  6. Check out Bill Woolsey's latest:

  7. I think I actually unlearned macro in that class. :)

  8. hey - some macro is worth unlearning :)


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