David Beckworth did a podcast with Tyler Cowen, and there was an extensive discussion of recent Fed policy. Here’s one interesting comment:
Cowen: We did as well as we did coming out of 2008-2009 because of that work. Probably, we should have done more, as you and others have argued, but I give them a lot of credit for that. And I think next time around, if we have a crash like that crash, we will do more and tolerate a 4% inflation rate in a way that we didn’t in 2009, and things will be much better. That’s when the real payoff of what Scott has done and you have done, Bennett McCallum earlier, and others, next time around we’re going to say that we can live with 4% inflation, we can get it back down later on in a painless enough way, and in the situation like the next 2009, we’re going to do better. I strongly believe that. And all of the Fed talk in the meantime, it’s Straussian. The real question [is], would Congress put up with the Fed doing 4% inflation next time around? I think it’s yes.
After mid-2008, the recession of 2008-09 was an aggregate demand shock—a severe decline in nominal spending. If the Fed were to do NGDP level targeting during that sort of crisis, then you would not even need 4% inflation. Most of the excess NGDP growth during the recovery period would show up as real growth, not inflation.
[Technically, the recession began in December 2007, and there was high inflation during the first 6 months of 2008 due to rising commodity prices. But the deep slump began in mid-2008, and was caused by a negative demand shock with falling prices.]
In addition, I’d argue that a repeat of 2008-09 would only happen in if Fed policy had no credibility. But in that case the make-up policy would likely be ineffective, just as in 2009. The point of level targeting is not to get a quick recovery when there’s a deep slump, the point is to prevent a deep slump.
(Covid was one of those once in a century exceptions, when a deep slump was inevitable. But Covid is not a good example to think about when designing a system.)
When I make this argument people ask me “OK, but what if there is a deep slump, what then?” It’s clear to me that they don’t really believe level targeting will work. My answer is that you try to get back to the trend line if you’ve announced that as your policy, but I honestly don’t think it would work in that case. If markets don’t expect success, why should I? If an NGDP futures market showed a big drop in future expected NGDP, then you’d try to reverse that. But the market forecast already incorporates the likely policy response. So if markets are pessimistic then I’m also very pessimistic, even if the Fed claims to be following my preferred policy.
I thought the Fed had adopted average inflation targeting back in 2020. The markets saw through those claims before I did, and of course the markets were correct.
Cowen: Scott himself has said that it’s the worst time to have nominal GDP rules. In a pandemic, I think all rules get tossed out the window, whether or not they should be, they probably should be, but they will be. Then your price data… So, if I wanted to see a movie in May 2020, is the price infinite? Is the price still $14? That’s quite arbitrary. So, how you define the price indices for services, to me, becomes arbitrary. Nominal GDP targeting just becomes seat of the pants, which to be clear I’m fine with, but it’s not really nominal GDP targeting, because the quality of the price data is gone.
I’m confused by this. If movie theaters are shut down and price data is ambiguous, that’s a stronger argument for NGDP targeting. During Covid, the price of movies was ambiguous, but the contribution of movie theaters to NGDP was precisely zero. One reason that NGDP is superior to inflation is that NGDP is a relatively clear concept, whereas the price level is a very vague concept that’s never been adequately defined.
It’s true that I didn’t think it made sense to maintain steady NGDP growth during the shutdown, but there’d be no problem in continuing to target one or two-year forward NGDP along a 4% trend line during a pandemic. And as a practical matter, it’s impossible to stabilize “spot NGDP”, so NGDP targeting has always been about stabilizing future expected NGDP. The only thing that changes during a pandemic is that you might briefly wish to target NGDP a bit further out (say two years instead of one.) If the Fed had done so, we would have avoided the inflation fiasco of 2021-22. With 4% NGDP level targeting, inflation would have averaged 2% during 2019-23, even if headline inflation briefly reached 4% or 5% for a few months in 2022.
So perhaps once every 100 years there’s a shock that leads the central bank to target two year forward NGDP instead of one year forward NGDP. That’s doesn’t seem like a large concession.
Cowen: I don’t think that we can have a Fed bound by rules, as much as I would like to do that. The old Carl Schmitt idea, “he who is sovereign decides the exception.” I just don’t think you’re going to get around that. I don’t think the Fed would ever accept rules or recommend them to Congress. What you want is a Fed imbued with NGDP ideology, that at critical key moments in history, is willing to take some chances in the right direction. We got that. We’ve won, maybe temporarily, but for now-
I’m not sure what Tyler means by a “NGDP ideology”, but it can’t be anything like NGDP level targeting. The Fed followed up a disastrous NGDP shortfall in 2008-09 with a wildly excessive level of NGDP growth in 2021-23, with no attempt to revert to the previous trend line. So I’m not sure who has “won”, but it clearly is not market monetarists.
In case this post sounds too grouchy, I do recognize that the Fed now has a better understanding of the need for make-up policy. But being half way to enlightenment is very dangerous when an institution has a lot of discretionary authority and is subject to political pressure.
I added an update to my recent post entitled Nemo judex in causa sua, where I linked to a book by Peter Boettke, Alexander Salter and Daniel Smith. They do an excellent job showing the dangers of discretionary policy. Tyler might be correct that discretionary policies are inevitable. But in that case, there’s no reason to be optimistic about future Fed policy.
READER COMMENTS
spencer
Dec 28 2023 at 8:46am
The belated recognition of the demand shock, i.e., due to the prior deceleration in N-gDp, exacerbated the 2008 downswing. Failure to act has latent consequences.
Thomas L Hutcheson
Dec 29 2023 at 11:08am
TIPS expectations reacted very quickly. Apparently the Fed was not paying attention.
Todd Ramsey
Dec 28 2023 at 9:47am
What would the NGDP securities market look like in actual practice?
Would a new market open every quarter (ending Fed participation in the “old” market) predicting NGDP for the two-years-out quarter? Would a one-year-out market would be too volatile during an emergency comparable to COVID?
Would it be possible or desirable to instead open a new market every week or month (simultaneously closing the “old” one) using the Atlanta Fed Nowcast or something comparable?
When you’re sitting around bored someday, I hope you will consider a post outlining for lay people how your NGDP securities market would work. Thanks for all you do!
Scott Sumner
Dec 28 2023 at 11:54pm
I’ve already done many posts explaining how we could use NGDP futures markets to target NGDP. My online book explains the concept in Chapter 5:
https://www.themoneyillusion.com/wp-content/uploads/2023/03/Sumner_AlternateApproachesMonetaryPolicy_v1a.pdf
Todd Ramsey
Dec 29 2023 at 9:55am
What I may be too obtuse to find in either the online or The Money Illusion books is how an NGDP securities market would concretely be structured.
Would the security be a debt-funding tool for the U.S. Government ala Treasuries? Would the Fed be buying and selling in only one contract at any one time? How far out would the contract terminate: One year? Two years? A new contract every day? (every week? every quarter?) How big would the market be relative to the size of the Treasuries market?
If you can’t dumb the explanation down to at least my level, it will be hard to gain traction amongst the public.
You may have identified the “end of history” for macroeconomics. If we can only get it implemented!
Please?
Scott Sumner
Dec 30 2023 at 11:36am
Would the security be a debt-funding tool for the U.S. Government ala Treasuries? Would the Fed be buying and selling in only one contract at any one time? How far out would the contract terminate: One year? Two years? ”
No
Yes, one contract.
I’d prefer one year, but two years would be acceptable.
There could be a new contract every quarter, although other systems are possible. I don’t know how big the market would be–that’s not important.
Todd Ramsey
Dec 28 2023 at 9:50am
“So perhaps once every 100 years there’s a shock that leads the central bank to target two year forward NGDP instead of one year forward NGDP.”
Does such a policy substitute discretion into an otherwise rules-based policy? Are the Fed governors trustworthy arbiters of when discretion should override the rules?
Please don’t misinterpret. I’m a huge fan of the NGDP securities market idea. I’m just seeking a better understanding of how it would work.
bb
Dec 28 2023 at 2:34pm
Not grouchy enough.
Let’s be honest, he just likes say Straussian. Personally, I think the Fed has been winging it and changing their story along the way to make it look like they have a plan. Calling it Straussian is giving them way too much credit. Tyler also doesn’t seem to understand your fairly easy to understand views.
Rajat
Dec 28 2023 at 5:45pm
Sorry, another long comment. I think there’s a few things going on here. First, I take market monetarism to incorporate four ideas: (1) targeting NGDP instead of inflation; (2) targeting market forecasts; (3) level targeting; and (4) ‘Whatever it takes’ concrete steppes. Tyler early on in your blogging and since frequently refers to the first; but I have seldom heard him emphasise the other three. (Maybe he did in this discussion?) By “NGDP ideology”, I presume he is only referring to the Fed not getting hung up on offsetting negative supply shocks. I’d say 90% of popular commentators (by whom I include Tyler) principally associate you with NGDP and barely think of or understand the other elements. I was like that myself for the first few years of reading your blog. The significance of forecast targeting for macroeconomic stability requires a major conceptual leap. It requires one to truly accept – contrary to 99% of market commentators – the EMH and that central bankers don’t have magical insights into the way economies work, such that policy should in a sense be ‘led’ by markets. The stabilising effect of level targeting requires another leap, because just about everyone (Tyler clearly included) believes that even notionally independent central bankers are subject to time-inconsistent preferences. Sticking to a level target after a bout of over- or under-shooting is really hard, as we saw with the abandonment of FAIT (and, in a similar way, with the loose commitment to inflation targets in the 2010s by the ECB, BoJ (even under Kuroda) and the Fed (with the old, ‘is 2% a ceiling?’ debate). If you listen to David Beckworth’s conversation with Ed Nelson about Ben McCallum, even McCallum wasn’t sold on level targeting in spite of his acceptance and adoption of rational expectations. And you concede yourself that once in a century, it might make sense to extend the period of returning to trend. But it’s easy to say after the fact that the Fed should have targeted 2-year ahead NGDP instead of 1-year ahead. How could investors have known this in real time? What if there had been no vaccine? What if the death rate had been 10 times higher? What if affluent people had chosen to refrain from circulating and consuming services for 5 years instead of 2? All of that uncertainty undermines confidence in an announced level target. “Whatever it takes” magnifies the importance of that uncertainty. Most macroeconomists think of monetary policy at the ZLB in terms of long bond yields, as Friedman did (again, according to Ed Nelson’s book). Very few (again, I expect including Tyler) would have heard of the Ben Bernanke thought experiment about the BoJ. Both level targeting and the thought experiment require an a good understanding and real acceptance of backwards induction. The ‘Chuck Norris effect’ is a good metaphor, but in reality, investors are always testing the resolve of the central bank. The ‘chainstore paradox’ accords with most people’s intuition of monetary policy a lot more than the Chuck Norris effect. None of which is to say that we haven’t made good progress. Investors saw in 2009 that the Fed won’t allow another Great Depression, so NGDP might be allowed to fall 5% but it won’t fall 30%. That offers some comfort and confidence. But I think it’s telling that even one of the earliest and longest and smartest fans of your blog (Tyler) associates you and David Beckworth almost exclusively with NGDP and barely at all with the other concepts central to market monetarism.
spencer
Dec 30 2023 at 8:40am
re: “How could investors have known this in real time? ”
You can’t miss big changes in M*Vt (where N-gDp is a proxy). The distributed lag effects of monetary flows are mathematical constants, not long and variable.
As Dr. Richard G. Anderson posited: “Reserves are driven by payments”.
You must extrapolate required reserves as the monetary base. I.e., you exclude the currency component. Then the money multiplier is the money stock divided by legal reserves.
Link: Dr. Daniel L. Thornton, May 12, 2022:
“However, on March 26, 2020, the Board of Governors reduced the reserve requirement on checkable deposits to zero. This action ended the Fed’s ability to control M1.”
Scott Sumner
Dec 30 2023 at 11:49am
It’s unfortunate that people equate NGDP targeting with market monetarism, as that would imply that people like Larry Summers and Michael Woodford are market monetarists.
You said:
“The ‘Chuck Norris effect’ is a good metaphor, but in reality, investors are always testing the resolve of the central bank.”
I don’t see this as being an important problem. It’s not like the Fed intended to engage in level targeting and then abandoned the goal when tested by markets—they never had this sort of goal. I agree that other economists worry about policy effectiveness, but all I can do is keep pushing back against this sort of nonsense.
“But it’s easy to say after the fact that the Fed should have targeted 2-year ahead NGDP instead of 1-year ahead. How could investors have known this in real time?”
I said this at the time. They’d know it because the Fed would announce it. Given the sorry record of actual monetary policy, I don’t see how a once in 100-year adjustment is a big problem. Under FAIT we have almost yearly discretionary adjustments that catch the market off guard. So why not adopt something 100 times better? Why demand perfection?
Thomas L Hutcheson
Jan 1 2024 at 9:03pm
“Under FAIT we have almost yearly discretionary adjustments”
I disagree. I think the target has remained 2% but “flexibility” implies engineering over-target inflation to accommodate shocks.
I think I agree that the Fed tends to make too large and too infrequent changes in its policy instruments. Smaller more frequent (maybe less than monthly could be less surprising.)
Thomas L Hutcheson
Dec 28 2023 at 9:53pm
I think it would be very difficult to distinguish a Fed with a FAIT policy from a Fed with a FNGDPT policy. [Average is a forward looking average]
In a way COVID is the perfect example of the kind of situation the Fed has to set monetary policy instruments in response to. [Perfect because it was so clear and obvious] An economic event occurs that is going to require larger changes in relative prices and larger reallocation of resources than usual. Downwardly sticky prices of some goods mean that those relative price re-adjustments will require an increase in the average price level. The Fed has to figure out how much and engineer enough above target inflation (and hopefully not more than enough as it apparently did in 2021-22) to make it happen.
Replace “Inflation” with “Change in NGDP” and nothing much else changes.
Capt. J Parker
Dec 29 2023 at 9:44am
Bingo.
AND by the way, the ’21-23 policy glitch probably carried with it an IOR payment stream to banks that was larger than it needed to be AND increased costs to service the national debt that were larger than they needed to be had the Fed tightened sooner. So, it’s not taxpayers who have “won” either.
It’s bizarre, the Fed is directly responsible for both the 2008/9 recession and the 2021/3 inflation and (almost) everyone wants to say what a great job they are doing.
Thomas L Hutcheson
Dec 29 2023 at 11:07am
If the Fed were to do [and know to be doing] FAIT during that sort of inflation would probably need not average 4% before full employment were restored.
Any kind of targeting should prevent a deep slump. But shocks happen. Anything that requires a large re-adjustment in relative prices when some prices are downwardly sticky will require the Fed to engineer enough inflation to accommodate the changes/prevent a deep slump.
I do not see how NGDPLT gives conceptual guidance about what the Fed actually ought to be trying to achieve in response to sectoral shocks, whether positive, negative, demand or supply shocks: accommodate income maximizing changes in relative prices.
Scott Sumner
Dec 30 2023 at 11:54am
NGDPLT stabilizes wage inflation, which is the only sort of inflation that matters.
Thomas L Hutcheson
Dec 29 2023 at 11:22am
The people who think the Fed did a great job in 2008 onward [are there any?] are not the same people saying it did a great job 2021-23.
Although the Fed probably created a bit more over-target inflation than was minimally necessary to respond to the COVID shock, the speed of the recovery and the lack (so far) of a recession on bringing over-target back down to near target is pretty darn good.
David S
Dec 29 2023 at 11:30am
Let’s not forget that the Fed started hammering nails in the coffin for the housing market in 2004. Their misread of what was actually going on during that period is stunning—see Scott’s book and Kevin Erdmann for the complete story.
No congratulations are in order for any aspect of Fed policy from 2006 through 2010. I don’t want to repeat the inflation of 2021-2023, but the Great Recession was a lot worse for everyone. Not quite a Japan level “lost decade” but pretty close.
Scott Sumner
Dec 30 2023 at 11:52am
I think Fed policy in 2007 was fine. Housing wasn’t the issue, the problem was falling NGDP in 2008.
Comments are closed.