I agree with you that moving to a higher inflation target would be a bad market signal right now, but NGDP targeting isn't necessarily inflationary. It would have been relatively inflationary in 2006 when Scott and others bough the idea to public awareness, but it would have been deflationary during the post-COVID inflation relative to the 2% norm. The other possibility - more superficially hawkish but still a level target - is George Selgin's productivity norm, in which the monetary authority would aim to have prices fall in line with productivity. Much more logistically difficult to pull off, but its the only deflationary proposal I've ever come across that can be defended against the accusation that it will just be a return to regular banking crises.
In retrospect, yes, but targeting ngdp has the advantage that it’s self-adjusting. A period of high inflation will automatically lead to tighter policy and low inflation to looser policy without any change in the target.
As a “rule” it works directionally well for negative supply shocks although it may need flexible over-target NGDP growth for a large shock. A positive supply shock would be problematic in that it woud mean less that tren or even negative inflation whicsh is never good (if the target was well chosen. But if applied flexibly with over target nominal growth, it is OK
I think that depends on why negative inflation is so bad and how that’s transmitted through the economy.
If you look at the problem as primarily price stickiness then yes, negative inflation by itself definitely will cause an economic contraction. But if you look at it as primarily expectations, then if people expect negative inflation they will adjust and nominal prices will not be (as) sticky. At least in principle it’s easier to set and hold NGDP expectations than inflation expectations.
Good point. Stickiness itself can be endogenous and may not be invariant to the inflation target.
Another good point is that my kind of explanation of sectoral shocks and sticky prices in principle lets you derive an inflationtarget, a one good, ont input one price modle cannot excetp with arbitrarily assumed expectations.
Admirable post, although I don't come down in the same place on the takeaways.
For one thing, you assume that the neutral interest rate has been coming down and will keep doing so. I agree that it was low or negative in the 2010s, because of demographics and productivity slowdown, and that created a sluggish ZLB economy for years.
But interest rates are higher now, and I don't think it's transient. It's fundamental. AI is changing the productivity trend and putting it on a higher path.
A 4% inflation Target would have been useful 10 years ago. I don't think we particularly need it anymore.
In my money and banking classes in the mid 80s the dominant opinion was that the Fed should be opaque in order to prevent the economy from adapting before a policy change and neutralizing the impact.
Inflation targeting in a liquidity trap / deflationary environment (like 2008-2009) might have a valid use, in order to stimulate consumption and investment, but otherwise I don’t see the point.
This is a change in thinking based mainly on the reflexivity revolution. In the Keynesian era, the idea was that unexpected monetary expansion can create real growth.
But per rational expectations, an economic theory that depends on the actors not understanding the theory for its predictions is void. So this kind of thinking is gone.
The idea now is to create credible policy so that the public’s expectations match and reinforce the policy direction. Inflation targeting is a reasonable way to do that. Not the best way, but a way.
Right. Some target and a credible method for reaching that target is better than no target. But a nominal GDP target or even a price level target is better than an inflation rate target. The problem with a rate target is expectations accumulate - if your target is 2% and you hit 3%, you should really be targetting just over 1% the next year. This is why people feel inflation long after its gone - their price level expectations on which their wage and debt agreements were based are upset and only slowly correct over time. A level target would make up for this. Of course, the converse is true - arguably in 2006 the Fed unintentionally tightened monetary policy by coming in below its inflation target and that was one of the sources of the pressure on the banks that led to the GFC.
Flexible NGDPLT is. given identical views about what maximum real growth is, identical to Flexible PLT. In both the central bank needs to judge how much over-target NGDP or PL growth is needed to allow relative prices to adjust given that some prices are downwardly sticky. And both need to know what the minimum level of NGDP or PL growth maximizes real growth.
As a matter of fact, I'm not persuded that we need an AIT greater than 2% PCE. We were almost there before Libertaion Day shock. And the problem in in 2008-2009 was not a to-low target, but that the Fed failed to use all its tools (including committing to higer infation) to achieve suffientl inflation.
Notwithstanding the spposed change in rationalle, Fed behavior has remains pretty consistent with FAIT. There is no need to change, although greater trnparancy about what it is actually dong would help.
The 2% target is arbitrary. New Zealand improvised it in 1990. The Fed didn't adopt it until 2012. The number was never derived from first principles.
But whether the target is 2% or 4% or replaced with NGDP targeting, the order of arrival doesn't change. New money reaches asset prices before it reaches wages. That sequence held under Greenspan's instinct, Bernanke's explicit target and Powell's flexible average. The divergence between what workers produce and what they're paid widened through all of it.
Why do the gains keep arriving in the same order regardless of the target?
I agree with you that moving to a higher inflation target would be a bad market signal right now, but NGDP targeting isn't necessarily inflationary. It would have been relatively inflationary in 2006 when Scott and others bough the idea to public awareness, but it would have been deflationary during the post-COVID inflation relative to the 2% norm. The other possibility - more superficially hawkish but still a level target - is George Selgin's productivity norm, in which the monetary authority would aim to have prices fall in line with productivity. Much more logistically difficult to pull off, but its the only deflationary proposal I've ever come across that can be defended against the accusation that it will just be a return to regular banking crises.
NGDP is as "inflationary" as its implicit inflation target is.
In retrospect, yes, but targeting ngdp has the advantage that it’s self-adjusting. A period of high inflation will automatically lead to tighter policy and low inflation to looser policy without any change in the target.
As a “rule” it works directionally well for negative supply shocks although it may need flexible over-target NGDP growth for a large shock. A positive supply shock would be problematic in that it woud mean less that tren or even negative inflation whicsh is never good (if the target was well chosen. But if applied flexibly with over target nominal growth, it is OK
I think that depends on why negative inflation is so bad and how that’s transmitted through the economy.
If you look at the problem as primarily price stickiness then yes, negative inflation by itself definitely will cause an economic contraction. But if you look at it as primarily expectations, then if people expect negative inflation they will adjust and nominal prices will not be (as) sticky. At least in principle it’s easier to set and hold NGDP expectations than inflation expectations.
Good point. Stickiness itself can be endogenous and may not be invariant to the inflation target.
Another good point is that my kind of explanation of sectoral shocks and sticky prices in principle lets you derive an inflationtarget, a one good, ont input one price modle cannot excetp with arbitrarily assumed expectations.
Admirable post, although I don't come down in the same place on the takeaways.
For one thing, you assume that the neutral interest rate has been coming down and will keep doing so. I agree that it was low or negative in the 2010s, because of demographics and productivity slowdown, and that created a sluggish ZLB economy for years.
But interest rates are higher now, and I don't think it's transient. It's fundamental. AI is changing the productivity trend and putting it on a higher path.
A 4% inflation Target would have been useful 10 years ago. I don't think we particularly need it anymore.
In my money and banking classes in the mid 80s the dominant opinion was that the Fed should be opaque in order to prevent the economy from adapting before a policy change and neutralizing the impact.
Inflation targeting in a liquidity trap / deflationary environment (like 2008-2009) might have a valid use, in order to stimulate consumption and investment, but otherwise I don’t see the point.
This is a change in thinking based mainly on the reflexivity revolution. In the Keynesian era, the idea was that unexpected monetary expansion can create real growth.
But per rational expectations, an economic theory that depends on the actors not understanding the theory for its predictions is void. So this kind of thinking is gone.
The idea now is to create credible policy so that the public’s expectations match and reinforce the policy direction. Inflation targeting is a reasonable way to do that. Not the best way, but a way.
Fair point and I did read Soros’ “The Alchemy of Finance” back in the day. I see no issue with markets effectively doing the Fed’s work for it.
But I think there are more robust ways than a rigid inflation target for the Fed to establish and maintain credibility.
Right. Some target and a credible method for reaching that target is better than no target. But a nominal GDP target or even a price level target is better than an inflation rate target. The problem with a rate target is expectations accumulate - if your target is 2% and you hit 3%, you should really be targetting just over 1% the next year. This is why people feel inflation long after its gone - their price level expectations on which their wage and debt agreements were based are upset and only slowly correct over time. A level target would make up for this. Of course, the converse is true - arguably in 2006 the Fed unintentionally tightened monetary policy by coming in below its inflation target and that was one of the sources of the pressure on the banks that led to the GFC.
Agreed - “Fedspeak” serves the need opaqueness well, but explicit inflation targeting makes little sense in this respect (and in most respects)
Flexible NGDPLT is. given identical views about what maximum real growth is, identical to Flexible PLT. In both the central bank needs to judge how much over-target NGDP or PL growth is needed to allow relative prices to adjust given that some prices are downwardly sticky. And both need to know what the minimum level of NGDP or PL growth maximizes real growth.
As a matter of fact, I'm not persuded that we need an AIT greater than 2% PCE. We were almost there before Libertaion Day shock. And the problem in in 2008-2009 was not a to-low target, but that the Fed failed to use all its tools (including committing to higer infation) to achieve suffientl inflation.
Notwithstanding the spposed change in rationalle, Fed behavior has remains pretty consistent with FAIT. There is no need to change, although greater trnparancy about what it is actually dong would help.
The 2% target is arbitrary. New Zealand improvised it in 1990. The Fed didn't adopt it until 2012. The number was never derived from first principles.
But whether the target is 2% or 4% or replaced with NGDP targeting, the order of arrival doesn't change. New money reaches asset prices before it reaches wages. That sequence held under Greenspan's instinct, Bernanke's explicit target and Powell's flexible average. The divergence between what workers produce and what they're paid widened through all of it.
Why do the gains keep arriving in the same order regardless of the target?