Michael Belongia on Econtalk
This interview is required listening for anyone interested in the Fed. A few comments:
1. Michael starts off with some interesting anecdotes about how Fed chairmen often abuse their powers. In one case he discusses how Greenspan used to do an end run around the FOMC, so that he could get his preferred fed funds target (and make it look almost unanimous), even if a number of FOMC members were initially opposed. Another example involves Volcker vetoing the choice of monetarists like Lee Hoskins and Jerry Jordan for the position of regional Federal Reserve Bank presidents. The view from the inside is not pretty.
2. He then does a really nice job of explain why fed funds targeting often ends up being procyclical, pouring fuel on the fire.
3. Much of the piece has a political economy slant, which I think will appeal to people who lack knowledge of and/or share my disinterest in highly technical monetary theory. In these areas Russ Roberts also does very well. He is more sympathetic to Belongia’s views than he is to my views, and this allows for a very smooth conversation where Russ is able to contribute a lot of insights into the public choice aspects of Fed. I won’t summarize all of Michael’s views on what the Fed should be doing differently, they link to one of his papers, and I also have this earlier post that discusses some of his views.
4. I was especially interested in his discussion of the flaws in the current monetary aggregates. I admit to not paying a lot of attention to this issue, as I prefer futures targeting to any form of monetary aggregate targeting. However most economists favor some sort of backward-looking Taylor Rule, and Michael points out that all of the arguments raised against the monetary aggregates in the 1980s, equally well apply to the Taylor Rule (or rules, since its proponents can’t even agree on what version is best.)
5. Not surprisingly, the one area I disagreed was the NGDP target discussion. But first let me indicate where I agree, and then where I diverge.
a. I agree that the Fed cannot hit two targets with one instrument.
b. I agree that the current mandate of stable prices and full employment causes great mischief. As I argued a couple days ago, it leads to a situation where at some times the hawks have the upper hand and at other times the doves dominate policy. This leads to an erratic policy. It also gives the Fed a ready-made excuse for their failures. Like a magician using sleight of hand, whenever one policy goal is doing poorly, they can say “but look over here, we need to loosen or tighten because of this other objective.” They need a single goal.
So where do we disagree? I have the feeling that Michael thinks NGDP is still implicitly straddling two goals, as the NGDP growth rate is the sum of inflation and real growth. A couple responses:
a. NGDP is still a 100% nominal variable, fully under Fed control. The Fed can (in principle) move NGDP over a range of zero to infinity, just like the price level.
b. The problems associated with inflation volatility (Michael’s preferred target is inflation) are more accurately described as problems that arise from NGDP instability. What are those problems? Inflation instability is alleged to create business cycles. For instance, if wages are sticky it may lead to suboptimal real wage movements. But this is even more true of NGDP instability. Inflation instability is alleged to distort financial markets, to redistribute wealth between lenders and borrowers. But George Selgin has argued (and I agree) that this argument better fits NGDP instability. (Although Selgin uses a slightly different target from NGDP.)
c. Michael mentions an example where Milton Friedman once asked (when someone proposed a 6% NGDP target) what would the Fed do if inflation was 7% and real growth was 1%? [Edit, I meant minus 1% growth.] The implication was that the high inflation showed that we needed to tighten policy, but the NGDP target would be signaling “all is well.” Surprisingly, I think the Fed should and would pay more attention to NGDP than inflation, even in that case. The NGDP target would keep wage growth and core inflation well behaved, and the transitory supply shock would not have any long run impact on inflation, and wouldn’t have any impact at all on long run inflation.
d. This italicized point is of course debatable, but I am pretty confident in making this assertion. If the long run real growth rate is 3%, then (under a 6% NGDP target) for every year with 7% inflation and -1% real growth, there would have to be another year with 7% real growth and -1% inflation. Surprisingly, those sorts of fluctuations in inflation would not cause any of the problems that inflation is alleged (correctly) to cause. It would cause less RGDP instability than an inflation target, and it would be fair to lenders and borrowers.
e. What about Michael’s concerns about how unanticipated inflation can disrupt long term decisions? After all, in theory you might get a decade or two of 7% inflation and -1% RGDP growth. In theory yes, but not in practice. Even the worst decades for supply shocks have shown surprisingly steady RGDP growth rates. I seem to recall that even the 1970s saw close to 3% RGDP growth. Now I happen to think that for all sorts of reason the current decade may be the worst in my lifetime. But that could occur even with 2% RGDP growth.
So I have two basic arguments, the problems we typically associate with inflation instability are better described at reflecting NGDP instability, and second, that over any extended period of time NGDP targeting will deliver a fairly stable inflation rate, as higher than desired inflation during one or two years will be offset by lower than desired inflation in the following few years.
Despite my quibbles about NGDP targeting, I agree with 90% of what Michael Belongia has to say. And he says it very well. He has worked at the Fed and you almost immediately sense that this is someone who knows what he is talking about. With no disrespect to Mississippi, it really bugs me that people this knowledgeable in both monetary theory and the real world practice of monetary policy are not more highly valued at the elite schools any longer. At the margin he would add far more value to an Ivy League program than some young hotshot that will churn out more DSGE models that no one will pay attention to in another 10 years (or in many cases even 10 months.)
Tags:
12. January 2010 at 09:27
Scott,
I don’t consider nominal GDP targeting to be a right or left wing issue. However, I am unaware of any economists on the left side of the spectrum who have come out in favor of GDPn targeting. Can you name any?
12. January 2010 at 10:44
Richard A. No, and I think that is quite revealing. The proposal obviously seems to be to the left of inflation targeting. I think I might know some reasons, but will think about it a bit more before making any comments
12. January 2010 at 11:09
It’s interesting that Belongia’s comments about recent Fed policy very much agree with yours in that it has been tight for years. This interview gives some insight into how the way the Fed actually operates makes it vulnerable to just the kind of mistake you think it’s making now.
So, the question would be, if this is why monetary policy is currently too restrictive, why doesn’t the FOMC know this? And, why isn’t this discussed more widely in the media? Unfortunately, given Bernanke’s public statements and Obama’s nomination to the FOMC, it seems that they actually like their current policy position, apart from any mistakes.
I also find it interesting that a Fed chairman can supposedly set monetary policy on his own. This suggests that merely changing the chair could make all of the difference, so long as the new chairman knows what to do.
So, all we need is a populist movement to have congress pass a law to target NGDP soley. Somehow I’m not optimistic. Nonetheless, I’ve already contacted my representatives about NGDP targeting and the White House and Senate about you being Fed chairman. I hate to act so autocratically in pulling these levers, but what good is power if never used?
12. January 2010 at 11:12
Concerning Friedman’s rhetorical question: the problem isn’t NGDP targeting; it’s simply that 6% NGDP growth is way too high! Try and frame a similar rhetorical question about a target NGDP growth rate of, say, 2%. You can’t. Instead, the lower target turns the tables on Friedman, who would find himself arguing (as some monetarists indeed don’t shrink from arguing) that the Fed needs to tighten money growth in the face of adverse supply growth, and thereby reduce NGDP growth below 2% (or whatever lowish number) so as to inflation stable–a recipe for disaster.
I assume, by the way, that in the example “1%” should read “-1%.”
By the way: I am _not_ claiming that I or anyone else could have actually won an argument with Friedman using the logic I’ve suggested here. I know from hard experience that Friedman was a much better debater than that!
12. January 2010 at 11:39
“Michael mentions an example where Milton Friedman once asked (when someone proposed a 6% NGDP target) what would the Fed do if inflation was 7% and real growth was 1%”?
Let’s assume here that inflation is measured by the GDP price deflator. A 7% inflation rate and a 1% real growth rate means current nominal GDP growth rate is 8%. The current real growth rate of 1% is less than the average real growth rate 3%. This implies that the economy got use to a GDPn growth rate higher than 8%.
What the Fed should do is let the economy get use to a GDPn of 8%. This would be when GDPr approaches 3% and inflation is near 5%.
Once the economy becomes use to a GDPn of 8%, the Fed could began reducing the growth rate of GDPn to 6% over a couple of years.
12. January 2010 at 13:42
Richard A – That’s an interesting point about leftish economists not supporting NGDP targetting. I agree with you that its not obvious why not.
Might it come down to one of the basic differences in values between left and right? The left generally wants to use politics to reduce suffering, whatever the cause. The right focuses more on making policy itself fair and letting people look out for themselves.
So the left tends to look at the downside of tight monetary policy as being unemployment, because that’s an obvious form of suffering, whereas the right sees the problem as being undermined expectations. Similarly the left sees the downside of slack policy as being the erosion of spending power, where the right sees it as … undermined expectations.
So for the right the problem looks symmetric, so NGDP targeting looks like a reasonable solution. For the left isn’t – unemployment is much more direct, immediate and (apparently) unavoidable form of suffering that the erosion of savings. Even if limited to monetary stimulus only, leftish economists would probably want to risk above-expectations inflation even during recessions with real causes, in order to keep employment up. And of course in general many leftish economists prefer fiscal stimulus anyway because its effect on employment is more direct.
As an aside, I don’t think its quite correct to call the view I call “right wing” above by that name, although I think its more or less Scott’s (and mine). It seems more of a soft libertarian viewpoint. There seem to be people (a long way) to Scott’s right who would prefer to risk deflation over allowing any inflation at all – the only explanation I can see for that view is that they see the role of monetary policy as promoting profits for banks and bondholders.
12. January 2010 at 16:30
George Selgin wrote —
“I assume, by the way, that in the example “1%” should read “-1%.””
That would mean the Friedman question should read, “Milton Friedman once asked (when someone proposed a 6% NGDP target) what would the Fed do if inflation was 7% and real growth was -1%?”
A GDPr of -1% implies that the growth rate of GDPn has just been recently slowed to 6% and the economy is in the process of adjusting to the lower growth rate of GDPn. In the long run, a 6% growth rate of GDPn will produce a GDPr of 3% and price inflation of 3%.
Now, let’s assume that the economy has gotten use to a GDPn growth rate of 6% and price inflation (P) is 3%. Suppose there is a price shock that causes P to increase to 4%. The Fed could reduce GDPn to 5% to try to bring P back down to 3% (inflation targeting), or it could simply maintain GDPn at 6%.
If the Fed were to target inflation by reducing the growth rate of GDPn,it would also be reducing the demand for labor. If instead the Fed maintained a constant growth rate of GDPn, no such reduction in labor demand would occur.
12. January 2010 at 20:17
To paraphrase what I read in another blog, oil prices rise, which pushes up inflation to 7%. The nominal income target is 3%, so the Fed engineers a recession, causing real output to shrink 4%. This is too much! The Fed instead should allow nominal income growth to rise, perhaps to 9%, resulting in a slow down in real income growth to 2%. What? Don’t you care about the unemployed?
The alternative perspective (which I think Sumner counts as right wing,) is that oil production falls, so the real product of the economy falls 4%. With the nominal income target of 3%, this generates a 7% inflation rate. Put this baldly, then why not slow nominal income growth to -4% and leave inflation unaffected?
I think the fancy approach is the suggest that the trade off in the social welfare function between inflation and output variation is not necessarily proportional, and so some fancy rule optimizing the social welfare function is better than stable growth of nominal expenditure. (And since we all have rational expectations, there is no benefit to a simple rule like a stable growth path for nominal expenditure vs. adjusting the federal funds rate to maximize a social welfare function on inflation variability and the output gap.)
Suppose the demand for oil is unit elastic. Then the higher price and lower quantity of oil leaves expenditures in that market unchanged. And with nominal income targeting, there is no change in the expenditure in the rest of the economy. Now, suppose oil demand is elastic? Or inelastic? If nominal expenditure in the rest of the economy is to be stablized, then nominal income must change. Of course, normally, there should be changes in expenditure in the rest of the economy, signaling the shift of resources between industries. Does the change in nominal expenditure in the rest of the economy, given a nominal income rule, just free up the resources that should be shifted to oil exploration and production? With careful analysis, surely we can find the optimal response function… better than simplistic old nominal income targetting. Right?
13. January 2010 at 09:26
This podcast was great. Also, I would recommend Belongia’s working paper entitled, “The evolving role and definition of the federal funds rate in the conduct of U.S. monetary policy”. The paper does a good job of discussing what the fed funds rate actually is and is not (a hint for Keynesians Old and New, it is not an instrument).
The link for the working paper is here:
http://mpra.ub.uni-muenchen.de/18970/
13. January 2010 at 11:42
[…] third objection is the most important (and is also mentioned by Bill Woolsey in a recent comment to this post.) How likely is it that the optimal coefficients on price level and output fluctuations are […]
13. January 2010 at 17:37
Mike, Bernanke should know this because he made exactly the same observation about Japan in 1999.
I’m not convinced Bernanke has as much power as Greenspan had back in the late 1980s and early 1990s. Remember that we now need unconventional policy moves, whereas Greenspan was merely manipulating the discount rate.
George, thanks, I corrected that. Yes, I agree the 6% figure is too high, but I answered usingt that number so that I couldn’t be accused of changing the subject.
Richard A. You are right but I made a typo. It should have been negative 1% real growth.
Simon K, I think it is even stranger than you do, as inflation targeting is viewed as being more “right wing” than NGDP. I have a post today on this issue.
Bill, I agree, and have a new post up today on this issue.
Josh, Yes, I liked that paper, especially the discussion of “instruments.”
14. January 2010 at 06:52
I think you’re missing a critical part of why left wing economists use Keynesian modeling and oppose NGDP targeting. Their goal is to increase the size and scope of government, not stabilize the economy. In times of recession, they call out for fiscal stimulus, which becomes a permanent increase in the size of government when the recession fades. Krugman (as a symbol for the type) understands that monetary policy is more effective, but he argues for fiscal policy anyway, because he would want a larger government whether or not there was a recession.
I realize that in the interest of being polite, you would want to address their arguments at face value. When you have addressed all of someone’s stated reasons for doing something and they still disagree, it’s time to look at their unstated reasons for believing/advocating it. James Buchanan’s “Democracy in Deficit” is more thorough coverage of what I’m trying to get at.
To address Bill Woolsey’s point, I agree that the optimal weights for inflation and output are probably not 50/50. However, in a demand induced/nominal recession, I view Fisher’s debt deflation to be a major source of the contagion in the banking sector. If companies expect to earn x and they only earn .9 x, they won’t be able to pay their debts. They don’t care how many goods they have to sell to earn x, as long as they get it. If the total sales in the economy remain stable, then companies on average will be able to pay back their loans (unless there is some collective irrationality, which I don’t believe in).
Supply shocks might be a solid theoretical reason to leave some discretion for monetary policy, but you have to ask yourself, is it worth risking the abuse of that power by the monetary authority?
14. January 2010 at 07:10
Sorry, I clicked on the link from “Why liberals should support NGDP targeting” and posted on the wrong comments section.
15. January 2010 at 14:20
azmyth, I agree with your points. Although I think it is a mixture of motives, not quite as underhanded as you imply. But mostly I agree.