Credibility is not now and never has been a problem for fiat money central banks

Matt Rognlie made this comment on my John Cochrane post:

Woodford is more skeptical of the power of an NGDP target than you are; in particular, although he thinks it would help, he doesn’t think a level target completely eliminates the problem of the zero lower bound. I agree with him, and I suspect that we have similar reasons for this belief. In Woodford’s model, one can prove that the zero lower bound will still pose a barrier in some situations even when there is an NGDP level target, and the central bank will not immediately be able to move the economy to its target. There are various reasons why his model might not be a complete description of reality, but I think it is a pretty good benchmark, and the burden is on you NGDP zen masters to come up with a modification of the model (or alternative model) where the zero lower bound is no longer a barrier. : )

First of all I don’t think that Woodford and I actually disagree all that much.  I’m pretty sure we both agree that there is some level of asset purchases (short of buying up planet Earth) that allow you to hit a 5% expected NGDP growth target. Here’s where I think we differ:

1.  He thinks the probability that this might require “unconventional purchases” (i.e. non-Treasury debt) is higher than I believe it is.  At least for the US circa 2012.

2.  He regards Fed purchases of agency debt and German government debt and AAA IBM corporate bonds at fair market prices as non-monetary policy, and I regard it as monetary policy.

3.  I regard the Fed purchase and later sale of interest-earning assets at fair market prices with non-interest earning cash and reserves as being far superior to a policy of bridges to nowhere and high-speed rail from Tampa to Orlando.  I don’t know how Woodford views that comparison.  I’d also expect that sort of Fed policy to be highly profitable, on average.

Over the years I’ve batted down the “credibility” arguments so often that I’m starting to feel like Sisyphus.  So here we go again.

Elite macroeconomists have lots of wonderful models.  Unfortunately these models are partially built on stylized facts that are actually myths.  I’ll focus on three areas:  The myth that past attempts at monetary policy have been hamstrung by credibility problems.  The myth that an optimal policy at the zero bound requires the Fed to promise to be irresponsible in the future.  And the myth that adopting a NGDP target would lack credibility because it would mean abandoning a previous inflation target.

Many macroeconomists have the following ideas in the back of their minds:

1.  Volcker was appointed to the Fed in 1979, but inflation didn’t start falling until mid-1981.  That’s shows his policy initially lacked credibility. I demolished that myth in this post.  For those with limited time, consider:

a.  The highest inflation in my lifetime occurred in 1980, when CPI inflation peaked at 13%

b.  A sharp but brief recession in early 1980 hurt Carter’s re-election chances.

c.  Carter appointed Volcker.

d.  Volcker cut the fed funds rate by 800 basis points in the spring of 1980.

e.  In response to the sharp rate cuts, NGDP soared at a 19% annual rate in 1980:4 and 1981:1.

f.  Volcker tightened sharply in mid 1981, and monthly inflation rates almost immediately plummeted sharply.

2.  Then there’s the famous case of the valiant BOJ, fighting for inflation but failing to convince a skeptical Japanese public.  I’ve swatted that argument down many times.  On one occasion Krugman rebutted my claim with an argument so weak that even the normally ultra-polite Ryan Avent seemed slightly bemused:

And…I’m genuinely mystified. The only thing I can think of that would square this circle is if Mr Krugman and I are using different definitions of the word “prefer”. As best I can tell, he has conclusively shown that Mr Sumner is right, and Japan hasn’t been in a deflationary trap. It just needs to fire all of its central bankers.

Then there are these “time-inconsistency” models that were developed to explain why central banks couldn’t control inflation.  Immediately after the ink dried the models became obsolete, as central banks around the world did control inflation.  But economists love models, so now they turned them around and tried to show why the Japanese were (supposedly) unable to create inflation.  They claimed that you could only exit the liquidity trap by “promising to be irresponsible” in the future, i.e. by promising to run above target inflation after exiting the liquidity trap.

That “irresponsible” assumption is based on the myth that the inflation rate is a good indicator of the welfare costs of inflation.  But it isn’t, for all sorts of reasons that I have discussed elsewhere (and George Selgin discussed before me.)  What matters is NGDP.  The main cost of inflation is that the real tax on capital rises as inflation rises.  But in practice the rate of NGDP growth is more important, as high RGDP growth also pushes up nominal interest rates, i.e. nominal returns on capital.  Furthermore, when NGDP is below trend (like right now, or in the 1930s) nominal interest rates will be low, even if the growth rate of NGDP was fairly high.  So I could much more plausibly write down a model where stable NGDP growth, level targeting, was the policy that minimized the welfare costs of “inflation.”  In that model one doesn’t need to “promise to be irresponsible” to exit a liquidity trap, you just need to set policy in a Svenssonian fashion, so that future expected NGDP rises smoothly along a 5% trend line.

With NGDP level targeting there is no need to adopt policies that are temporarily suboptimal, in order to get the desired AD growth.  A policy of stable expected NGDP growth, level targeting, is optimal on both employment grounds, and “welfare cost of inflation” grounds, at least compared to any other policy with the save average NGDP growth rate.

Karl Smith recently did a post criticizing John Cochrane’s claim that adoption of NGDP targeting would be non-credible, as it would force the Fed to discard it’s previous policy (presumably inflation targeting.)  But there’s another weakness with Cochrane’s argument; the Fed was never really a pure inflation targeter. Rather they have a dual mandate, to keep inflation relatively low and stable, and to keep unemployment close to the natural rate.  For several decades the Fed kept NGDP growing at about 5% per year.  This policy seemed pretty successful, as the business cycle became less volatile, and inflation stayed relatively low and stable.  But not completely stable, as that would imply a single mandate.  The Fed always maintained that they allowed some fluctuation in inflation (as during oil price shocks), in order to smooth output.  Greenspan hinted on occasion that they were actually targeting NGDP.

Then everything changed in 2008, and the Fed missed badly on both its inflation and employment mandate.  In 2009 unemployment soared to 10%, and prices fell (deflation.)  Since July 2008, inflation has only averaged 1.1%.

NGDP targeting would bring the Fed back to the pre-2008 policy, which was so successful. Indeed one famous NGDP proponent (Christy Romer) called for a 4.5% target, an obvious nod at the recent estimates of slightly lower trend RGDP growth.  Although I don’t think the NGDP target needs to be adjusted for changes in trend RGDP growth, I’d have no objection to a once a decade nudge to reflect changing estimates of trend RGDP.  That’s a side issue.  The key point is that Romer’s proposal is fully consistent with the Fed’s mandate, and with actual Fed policy pre-2008.  There should be no loss of credibility in returning to that successful policy.

On theoretical, practical, and historical grounds, policy credibility is simply not a problem for fiat money central banks.

PS.  I couldn’t find the link to the Ryan Avent post.

 

PPS.  Matt Yglesias presents a different argument for why credibility is a phony issue.


Tags:

 
 
 

52 Responses to “Credibility is not now and never has been a problem for fiat money central banks”

  1. Gravatar of OneEyedMan OneEyedMan
    6. September 2012 at 18:08

    Interesting post and I appreciate the effort to seriously address the credibility question.

    Yglesias’s argument doesn’t persuade me at all. With most relationships in life you have exit. That is, you can quit, fire , divorce, move, transfer, etc. In many that you can’t exit, you have voice to vote the bums out. By design, the Fed is protected from both exit and voice. As such, they are exactly the sort of institution where you’d worry that they would go back on their word when it is attractive to do so. After all, how can you punish them? Worse, they can dress up any going back on their word in a bunch of verbiage about how changing economic conditions required the change, and so would obscure that they really had gone back on their word.

    I think that your reputation argument actually conflates two positions. One argument is that the central banks cannot achieve inflation. The other argument is that the Fed today binds the Fed tomorrow. I thought Cochran’s credibility argument was the later where as your argument is mostly about the former.

    In many ways I think congress and the Fed are in similar situations. Congress can do a lot today with the government’s balance sheet by growing or shirking spending and revenue today. They can also change the law that controls benefits and other spending tomorrow. Unfortunately, whatever they make the another congress can unmake. You want to ruin the government balance sheet? No problem, indeed it is trivial, just like you can debase the currency with ease. Want to promise that some future congress will damage the balance sheet at 5% a year? That is no longer a trivial matter. That’s why I worry that credibility issue is not just a silly hobgoblin of the uninformed.

  2. Gravatar of Which of these pictures of central bank behavior looks more realistic? « Increasing Marginal Utility Which of these pictures of central bank behavior looks more realistic? « Increasing Marginal Utility
    6. September 2012 at 18:43

    […] Scott Sumner: Then there are these “time-inconsistency” models that were developed to explain why central […]

  3. Gravatar of K K
    6. September 2012 at 20:40

    “I’m pretty sure we both agree that there is some level of asset purchases (short of buying up planet Earth) that allow you to hit a 5% expected NGDP growth target.”

    Maybe. Where’s your evidence for this? I read his Jackson Hole paper and all I saw was a strong case for Wallace irrelevance.

    “He thinks the probability that this might require “unconventional purchases” (i.e. non-Treasury debt) is higher than I believe it is.”

    Again, where do you get this from? He argues that purchases of both treasury and non-treasury assets are irrelevant. But in models in which it makes a difference, non-treasury assets are only helpful to the extent that they are positive beta. So AAA debt doesn’t help at all. Here’s Woodford:

    “But it is important to note that such “portfolio-balance effects” do not exist in a modern, general-equilibrium theory of asset prices “” in which assets are assumed to be valued for their state-contingent payoffs in different states of the world, and investors are assumed to correctly anticipate the consequences of their portfolio choices for their wealth in different future states “” at least to the extent that financial markets are modeled as frictionless. It is clearly inconsistent with a representativehousehold asset pricing theory”

    He must spend at least 20 pages just trashing QE, on principle.

    “The myth that an optimal policy at the zero bound requires the Fed to promise to be irresponsible in the future.”

    Try *not* to get this result in a multiperiod model with ratex agents and potential for AD deficiency resulting from sticky prices. The world awaits a Lucas Critique compliant model with demand deficiency that can achieve an optimal solution independent of the ZLB.

  4. Gravatar of Bob Murphy Bob Murphy
    6. September 2012 at 21:05

    Scott, when you title this post: “Credibility is not now and never has been a problem for fiat money central bankers,” I think you are saying something that sounds shocking but only because you don’t mean what most readers will think you mean.

    As I skim your post, it looks like you are NOT denying what the average economist thinks about “central banker credibility.”

    I think the average economist would say this has to do with the public’s expectations of (price) inflation, and how the Phillips Curve tradeoff is worse when they expect high inflation. So, in order to get a desired (low) inflation target, the central bank has to engineer larger unemployment to convince the public it’s serious, than would be the case if the public already expected low inflation.

    Are you denying any of this?

  5. Gravatar of Saturos Saturos
    6. September 2012 at 21:23

    “Elite macroeconomists have lots of wonderful models. Unfortunately these models are partially built on stylized facts that are actually myths.”

    I think the broader point here is that in a social science with a dearth of controlled experiments, economists are always suffering from some sort of Lucasian bias which leads them to protect the status quo (as well as default statism/Hayekian social-engineering bias). For example, even though the vintage Keynesian Philips curve is dead, belief in the sacrifice ratio still persists, whereas we know that “the markets know the mind of the Fed better than it does itself”.

    Marcus Nunes’ latest is relevant, though also for a different reason than the one in the title: https://thefaintofheart.wordpress.com/2012/09/06/when-money-illusion-was-a-policy-assumption/

    “Although I don’t think the NGDP target needs to be adjusted for changes in trend RGDP growth, I’d have no objection to a once a decade nudge to reflect changing estimates of trend RGDP.”

    Aha! Now we see the true face of the NGDP targeters! You pretend to hate discretionary instability, but in reality you are all in favor of accelerating money inflation conducted by the central planners! Well you never fooled me!

    (Maybe if I say it first then MF will stay away…)

  6. Gravatar of Matt Rognlie Matt Rognlie
    6. September 2012 at 21:40

    Thanks for the reply. I think we agree on most of the discussion about credibility. The disagreement is really about the possible need for “unconventional” asset purchases, and the effect of those purchases if they are needed.

    We all agree that an NGDP level target is to some extent self-enforcing. Commitment to the target path ensures that there will always be a certain amount of nominal growth going forward, and in most models this makes the zero lower bound less of a constraint.

    But there is still a possibility, very likely in my view, that large shocks a la 2008 will push the economy away from the NGDP target for a sustained period of time, and that mere commitment to the target will not be enough to get the nominal economy on track right away. This doesn’t mean that an NGDP target is bad policy. To the contrary, I think we would be far better off in a hypothetical world where the Fed implemented an NGDP target path in 2007, even if it “failed” to hit the target for a few years. I just don’t want to oversell the policy!

    Why would an NGDP target fail to be self-enforcing? One possibility, not so far from reality in my view, is that underlying inflation at the current low level is quite persistent, and won’t change at an appreciable pace even when catch-up inflation is expected. (I think that the main nominal rigidity in the economy is wage rigidity, and the wage structure has an incredible amount of inertia in the current environment.) If this is true, then in the medium term an NGDP level target is very nearly just a target for RGDP. And we know that targets for RGDP aren’t necessarily enough to conquer a liquidity trap. In fact, the monetary rule in practically every liquidity trap paper in the New Keynesian literature is identical to an RGDP target (due to “divine coincidence” and the limited space of shocks in these models); and these papers all feature a meaningful liquidity trap! Indeed, in most of the optimal policy exercises I’ve seen, especially Werning (2012), it is necessary to substantially overshoot the trend level of output and prices. Merely returning to trend isn’t quite forceful enough. (But again, I’m not going to make the perfect the enemy of the good; I would be ecstatic if in the current economy the Fed simply promised to return to trend NGDP.)

    What do we do when the Fed can’t immediately hit the NGDP target, even if it lowers short rates to zero? This is where we enter the world of “unconventional” monetary policy. You believe that as long as the Fed commits to a NGDP target, a moderate level of asset purchases should be enough to hit it. (I think I’ve seen you write here that a program on the order of QE would be enough, if only it was combined with level targeting.) I don’t see any reason to be so confident. I do think that if the Fed set out to buy every asset in the world until it achieved its NGDP target, at some point it would probably succeed. But at this point, its balance sheet might be more like $10 trillion (larger than all mutual funds in the US put together!) or $15 trillion (larger than the entire US banking sector!), not $2.5 trillion. (Of course, I think monetary stimulus is so important that I wouldn’t mind seeing the Fed go this far. But there are a lot of reasons why it would be difficult.)

    Why do I think such a large intervention would be necessary? QE has mainly been effective as a signal of the Fed’s future intentions, and to some extent as a commitment device. But if you already have an NGDP target, then the “signaling” channel is no longer meaningful; future policy is pinned down by your commitment to the target path. The only remaining channel for asset purchases is the “portfolio balance” channel—the notion that if you buy a bunch of assets, you’ll raise their price and lower their yield. And this channel may be far, far weaker than is generally recognized.

    Even at long maturities, there is a highly effective arbitrage between Treasury rates and OIS forward rates. Those forward rates represent the market’s risk-weighted expectation of future policy. If the Fed can’t change the expectation of future policy (which, again, is pinned down by the NGDP target), all it can do is change the risk weighting, by redistributing interest rate risk from the private sector to its own balance sheet. As Woodford points out, in a representative household model a simple Ricardian argument shows that this achieves nothing; if the “Fed” rather than the “private sector” experiences losses, the private sector will ultimately pay anyway through higher taxes. No doubt the world is a little more complicated than that. But it is still hard for me to see how redistributing risk through a few trillion dollars of QE will accomplish very much, simply because the asset markets subject to interest rate risk are so big. In the US there are

    — $16 trillion in nonfinancial equities.
    — $13 trillion in mortgages.
    — $11 trillion in Treasuries.
    — $5 trillion in nonfinancial corporate debt.
    — $4 trillion in municipal securities.
    — $3 trillion in other bank loans.
    — $2.5 trillion in consumer credit.

    That’s over $50 trillion total. (I include “mortgages”, “consumer credit” etc. directly rather than MBS or bank deposits because they’re the claims on which the other assets ultimately are based.) And remember, the Fed can’t directly affect the expected cash flow on these assets — it can only shuffle around risk (in a manner subject to a Ricardian critique) and hope that the risk premium falls a little. Is there any reason to think that a few trillion in purchases will accomplish much? I don’t see any.

    Buying “agency debt” or “AAA IBM corporate bonds” might do a little more, but there isn’t much of a spread between these assets and Treasuries. And only a fraction of the spread that does exist is an honest-to-god “risk premium”, as opposed to a simple difference in the expected cash flow — if agency debt has a higher chance of default than Treasury debt, the Fed can’t do anything about it, no matter how aggressive its asset purchase program.

    Ultimately, I think that the Fed can have some traction through asset purchases—but those purchases have to be far more aggressive and “unconventional” to move nominal output by a substantial amount. This is true even when the Fed is fully committed to an NGDP target. Don’t get me wrong; I’d be happy to see $1.5 trillion more plowed into MBS and T-notes. But my best guess is that this would be a mild positive, no more valuable than slightly improved forward guidance or a slightly more aggressive NGDP target.

  7. Gravatar of Saturos Saturos
    6. September 2012 at 22:15

    Bob Murphy, yes that’s exactly what he’s denying, re. Volcker.

  8. Gravatar of Saturos Saturos
    6. September 2012 at 22:20

    Scott, any thoughts on Clinton’s speech?

  9. Gravatar of Ritwik Ritwik
    7. September 2012 at 01:58

    On 7th Septmeber 2012, the Chinese government announced 1 trillion RMB in fiscal stimulus. As a result, the equity markets rallied the most they have in 3 years. The PBoC didn’t mutter a word.

    http://ftalphaville.ft.com/blog/2012/09/07/1149801/china-stimulates-sort-of/?utm_source=dlvr.it&utm_medium=twitter

    With this one anecdote, I have conclusively shown that market fiscalism works and that the Sumner critique is invalid.

    Just kidding, Scott.:)

    But then, that’s the kind of argument you have been making for the past week or so. 🙂

  10. Gravatar of Major_Freedom Major_Freedom
    7. September 2012 at 04:39

    ssumner:

    He regards Fed purchases of agency debt and German government debt and AAA IBM corporate bonds at fair market prices as non-monetary policy, and I regard it as monetary policy.

    If the non-market Fed announces and follows through on purchasing these securities, the securities will no longer be priced at the going market rates. The prices would rise above the going market rates. This is because of the law of supply and demand. The Fed represents an additional demand to previously existing market demand, and since the Fed does not increase the supply of securities, we know that demand rises ceteris paribus, and hence prices rise.

  11. Gravatar of Major_Freedom Major_Freedom
    7. September 2012 at 04:45

    ssumner:

    The main cost of inflation is that the real tax on capital rises as inflation rises.

    A major cost of inflation is capital misallocation.

  12. Gravatar of Bob Murphy Bob Murphy
    7. September 2012 at 04:47

    Saturos, OK, if that’s what he’s saying, then he has to show us that the recession in the early 1980s wasn’t much worse than other recessions, right? But no, he can’t, because that recession was awful. If I wanted to bolster my claim that “it’s important for central bankers to have credibility” I would point to this example. I don’t see how Scott has refuted it in any way. Nobody was denying (to my knowledge) that you could quickly get inflation under control, they were saying that if you started out at 13% inflation and wanted to get it under control, it would really suck compared to starting out at 4%. How does the Volcker example not demonstrate this with flying colors?

  13. Gravatar of Major_Freedom Major_Freedom
    7. September 2012 at 04:56

    Matt Roglie:

    I think that the main nominal rigidity in the economy is wage rigidity, and the wage structure has an incredible amount of inertia in the current environment

    In your sphere of knowledge, does there exist the idea that inflation itself exacerbates wage rigidity? That wages do not fall as quickly as prevailing monetary conditions would have them fall, because both wage earners and wage payers have come to expect positive inflation due to the fact that they were born into and have worked in an inflationary world?

    The reason why I ask this is because I find those who call for more inflation as an indirect means to getting around the problem of wage rigidity are more or less chasing their own tails.

  14. Gravatar of Saturos Saturos
    7. September 2012 at 05:36

    MF, I think you’re right that we could cure money illusion by instituting a regime of persistent deflation, large enough to offset productivity gains to nominal wages, so that workers expected their nominal wages to be lower each year, and were forced to start thinking intelligently and subtracting the expected price level (eventually, as they realized that real growth was still continuing). But the problem here is that it’s difficult to get equilibrium deflation in excess of the real interest rate, unless you’re willing to tolerate sharply curtailed aggregate investment. And in that case we would also have to have continuing money inflation.

  15. Gravatar of John Lynch John Lynch
    7. September 2012 at 05:37

    Scott, a little Googling has returned me this link, which I believe is the full Ryan Avent post in question: http://www.economist.com/blogs/freeexchange/2010/07/monetary_policy_5

  16. Gravatar of Major_Freedom Major_Freedom
    7. September 2012 at 05:42

    Who gets to decide what “unconventional” securities purchases the Fed will engage in if they buy securities other than treasuries?

    What will be the decision criteria? Who contributes the most $$$ to board members (in the form of lucrative jobs, etc) before or after their tenure at the Fed?

    Now that “market” monetarism has gone into corporate-state alliance (i.e. fascist) economics, I am curious as to which citizens will have greater “equal” rights than others. Why IMB is chosen as having a backstop buyer of equity/debt that is the Fed, but independent entrepreneurs in their garages do not.

    Let’s convert the category “Too big to fail” from including financial institutions only, to including institutions such as those which manufacture military machinery and domestic spying equipment! Maybe GE and Caterpillar? No longer will these poor souls have to rely on financing through the Treasury only, what with the limitations of taxation and borrowing. Now they can get free financing directly from the Fed!

    Oh wait, that already is happening…

    Just like Keynes arrived on the scene in the 1930s to intellectually justify what the government was already doing at the time, so too it seems “market” monetarists are arriving on the scene just in time to provide intellectual justification for what the government is already doing!

    Regular as clockwork.

    I can’t wait for Sumner’s textbook that explains why the student and professor readers of his book don’t deserve Fed money through “unconventional” asset purchases. Why only those who provide resources to the state deserve it. Hey kids! Forget about providing civilian goods for the consumers. Become a cog in the military industrial complex and you’ll receive free Fed financing for your equity/debt.

    50 years from now reactionary statists will look at this time and say the same thing socialists are saying about Marx: “They didn’t intend for THAT to happen!”

  17. Gravatar of Major_Freedom Major_Freedom
    7. September 2012 at 05:54

    Saturos:

    MF, I think you’re right that we could cure money illusion by instituting a regime of persistent deflation, large enough to offset productivity gains to nominal wages, so that workers expected their nominal wages to be lower each year, and were forced to start thinking intelligently and subtracting the expected price level (eventually, as they realized that real growth was still continuing). But the problem here is that it’s difficult to get equilibrium deflation in excess of the real interest rate, unless you’re willing to tolerate sharply curtailed aggregate investment. And in that case we would also have to have continuing money inflation.

    Gradual deflation does not need to be in excess of real interest rates. The key is falling prices qua falling prices. Aggregate investment can remain fixed, or on a gradually increasing growth path, alongside falling prices of capital, as new capital comes into production by virtue of physical increases independent of nominal spending increases.

    Real interest rates can be whatever they happen to be.

  18. Gravatar of John Thacker John Thacker
    7. September 2012 at 06:40

    The BOJ example is clearly a case of the central bank wanting deflation even when the politicians disagree. See this article about the Japanese situation that I sent to Timothy B. Lee:

    http://www.bloomberg.com/news/2012-04-03/rejection-of-bank-of-japan-nominee-by-ldp-adds-pressure-to-ease.html

    All the opposition parties, across the spectrum, plus a group of rebels in the ruling DPJ, rejected a BOJ nominee because he was too skeptical of QE and favored the deflationary policy.

    The BOJ is clearly a case of central bankers failing to try.

  19. Gravatar of Adam Adam
    7. September 2012 at 07:20

    I literally do not even understand Rognlie’s line of reasoning. Don’t we need some reason to believe that whatever model he expects you to refute reflects reality before there is any obligation to refute it?

    But we know that the model, like all models, is not reality. So, again, absent any significant cost of doing so, why don’t we try the policy in reality, which may tell us whether the model is wrong?

  20. Gravatar of Floccina Floccina
    7. September 2012 at 07:26

    This brings to mind the problem that the central banks cannot buy assets that appreciate with inflation.

  21. Gravatar of Major_Freedom Major_Freedom
    7. September 2012 at 07:36

    Adam:

    So, again, absent any significant cost of doing so, why don’t we try the policy in reality, which may tell us whether the model is wrong?

    How do you know the policy does not have any “significant cost”, unless you know from experience? Or are you saying a priori the model incurs no significant cost?

    Then, you have to explain WHO incurs such costs. Costs are subjective, meaning, they relate to individuals, not “society”. There are no gains and costs to “society” apart from identifiable gains and costs to individuals. Since inflation of course does not affect everyone’s income or bank balances equally, then the model necessarily will incur costs on some individuals on net, and gains on other individuals on net.

    In other words, some people have some rights, and other people have a different set of rights. Different rights for different people. Hmmm, there is a name for such a policy, and it rhymes with schmascism.

  22. Gravatar of Major_Freedom Major_Freedom
    7. September 2012 at 07:38

    Floccina:

    This brings to mind the problem that the central banks cannot buy assets that appreciate with inflation.

    So…they cannot buy any assets?

  23. Gravatar of Saturos Saturos
    7. September 2012 at 07:55

    MF, bonds depreciate with inflation. And the mild deflation you propose would not be sufficient to overcome money illusion, but rather worsen its effects.

  24. Gravatar of Floccina Floccina
    7. September 2012 at 07:56

    @Major_Freedom
    They buy debt which tends to depreciate when we get inflation. We do not trust them to buy most other things.

  25. Gravatar of Patrick R. Sullivan Patrick R. Sullivan
    7. September 2012 at 08:01

    Speaking of Cochrane, he and Russ Roberts have a terrific video that ‘Mad Dog’ Mike Sax will not enjoy;

    http://www.youtube.com/watch?v=1eCYq2vD5GY

    but, don’t mention monetary policy. Maybe in part 2.

  26. Gravatar of Major_Freedom Major_Freedom
    7. September 2012 at 09:26

    Saturos:

    MF, bonds depreciate with inflation.

    Yes, in theory, higher inflation is supposed to raise bond yields by lowering their market prices.

    But I was assuming the context was assets purchased by the Fed. If the Fed buys bonds, then bond prices will rise, thus lowering their yields.

    And the mild deflation you propose would not be sufficient to overcome money illusion, but rather worsen its effects.

    Quite the opposite. If gradual deflation sees prices fall, then the downward rigidity of wages and other prices would be reduced, not worsened. This is because wage earners and wage payers would come to expect gradually falling wage rates as a matter of course. Thus, when prevailing monetary conditions call for lower wage rates, the pressure not to reduce them would be reversed.

  27. Gravatar of Major_Freedom Major_Freedom
    7. September 2012 at 09:31

    Floccina:

    They buy debt which tends to depreciate when we get inflation. We do not trust them to buy most other things.

    But if the Fed purchases debt, debt prices will tend to increase, not decrease.

    You’re talking about the pedagogical model that assumes higher inflation necessarily increases yields via lower bond prices. But I am saying the Fed increases bond prices by purchasing bonds. That lowers yields.

    For example, do you think treasury bond prices would be as high as they are now, or, equivalently, do you think treasury bond yields would be as low as they are now, if the Fed wasn’t engaged in buying treasury bonds with money created from nothing? In other words, suppose tomorrow the Fed stopped inflating and stopped buying treasuries, period. Do you think the prices would rise or fall? If you say they would probably fall, then you see my point. Inflation going to zero would make treasury bond prices fall and yields rise.

  28. Gravatar of Major_Freedom Major_Freedom
    7. September 2012 at 09:39

    ssumner:

    NGDP targeting would bring the Fed back to the pre-2008 policy, which was so successful.

    De facto NGDP targeting is not the same thing as de jure NGDP targeting. You are ignoring the fact that humans learn of the very policy in question. This fact is expressed in Goodhart’s Law, Campbell’s Law, and the Lucas Critique (specifically Goodhart’s Law since this is a policy issue).

  29. Gravatar of Scott Sumner Scott Sumner
    7. September 2012 at 09:41

    K, You said;

    “”I’m pretty sure we both agree that there is some level of asset purchases (short of buying up planet Earth) that allow you to hit a 5% expected NGDP growth target.”

    Maybe. Where’s your evidence for this?”

    Umm, maybe I assumed that because he doesn’t seem like a complete moron.

    And in the worst case where I’m wrong, then we in the USA have all the marbles.

    Ritwik, That’s not “fiscal policy” in China. I’ll try to address a few other comments when I have time.

  30. Gravatar of Morgan Warstler Morgan Warstler
    7. September 2012 at 10:07

    Jesus Christ.

    If tomorrow the Fed adopted 3% NDGPLT, and said BAM! We’re raising rates!

    Immediately, the credibility issue is solved.

    Sure, DeKrugman and company would shit a brick, but right then, EVERYONE would take the Fed very seriously.

    Then what happens?

    1. Well if Scott / DeKrugman is to be believed, if they are right,t he economy goes into a tailspin… stock prices crater.

    BUT, if they are right, once we are back below 3% NGDPLT, the machine kicks in and rates get lowered.

    And we are OFFICIALLY ON NGDPLT, and that’s the MOST IMPORTANT THING, right????

    Yes, yes, Unemployment is still high, but save that for layer.

    2. Maybe Scott / DeKrugman is wrong. Maybe the Fed is CREDIBLE, and as Ron Paul / Paul Ryan etc run around cheering and the prices of homes starts to fall, and even more Sovereign Funds pile into the Dollar…

    BUT, something else super great happens…

    US Debt borrowing rates have gone up, people have focused on % tax revenues going of Debt service.

    AND we get TONS AND TONS of pressure to REDUCE public employee compensation… ending pensions IMMEDIATELY gets rates lower WITHOUT having the rest of the real economy suffer.

    —–

    I don’t think people are think through logically the best easiest way to get NGDPLT done.

    Someone explain why I’m wrong.

  31. Gravatar of Michael Michael
    7. September 2012 at 10:47

    “Someone explain why I’m wrong.”

    Because you are asking the Fed to pursue contractionary policy starting today along with a guarantee of no catch-up growth. That’s not politically sustainable.

    Even someone like Woolsey, who favors a 3% level target, wants catch up growth first (I think).

  32. Gravatar of Major_Freedom Major_Freedom
    7. September 2012 at 12:14

    Michael:

    That’s not politically sustainable.

    Is this an economics blog or is this a political strategy blog?

  33. Gravatar of Scott Sumner Scott Sumner
    7. September 2012 at 12:50

    Matt, You said;

    “What do we do when the Fed can’t immediately hit the NGDP target, even if it lowers short rates to zero? This is where we enter the world of “unconventional” monetary policy.”

    I think you are missing my argument here. With a reasonable NGDP target nominal rates will probably never hit zero. But if a central bank wants to set a low target path (as in Japan), then unconventional purchases may be a consequence of that low target. But there’s no reason for a central bank, or the broader society, to cry about that. It’s a choice!

    It reminds me of all these right wingers who insist that the Fed needs to do a tight money policy, which drives rates to zero, and then cry because seniors can’t earn anything on their saving. It’s decision variable. Societies get to choose the NGDP target, which means they get to choose the ratio of base money to NGDP. If they choose a high base to GDP ratio, and the ratio of public debt to GDP is low, then you are forced to do unconventional monetary policy. That’s life. But there are no alternatives, once you have made that choice. Well actually I suppose you could push the public debt ratio up to 200% of GDP, like Japan, but that’s not much of an alternative.

    You said;

    “You believe that as long as the Fed commits to a NGDP target, a moderate level of asset purchases should be enough to hit it. (I think I’ve seen you write here that a program on the order of QE would be enough, if only it was combined with level targeting.) I don’t see any reason to be so confident.”

    Are there any examples in all of world history of countries with high trend NGDP growth rates, and high base to GDP ratios? (By high, I mean more than 50%, with no IOR.) Even in Japan it’s only 23% of GDP, and NGDP growth is negative!

    You said;

    “Ultimately, I think that the Fed can have some traction through asset purchases””but those purchases have to be far more aggressive and “unconventional” to move nominal output by a substantial amount.”

    If the Fed were truly serious about adopting something like Romer’s NGDP target plan, they’d have to sharply REDUCE the monetary base, or else face hyperinflation. There’s no way you can shoot for 10% NGDP growth over the next few years and expect the public to want to hold 18% of GDP in zero rate cash. It won’t work. So there’s no point in talking about how much more the Fed would have to buy, they’d have to SELL assets fast.

    I need to do another post.

    John, Thanks, I added the link.

    Saturos, Clinton’s a great speaker, but I can’t comment, because I don’t watch TV, except sports. And I hate politics, so I almost never watch speeches.

    Bob, You may have misunderstood me, I agree it’s much harder to reduce inflation from 13%, in the sense that you get more unemployment. But in a technical sense it’s easy, if you are serious about tight money.

  34. Gravatar of Major_Freedom Major_Freedom
    7. September 2012 at 13:22

    ssumner:

    It reminds me of all these right wingers who insist that the Fed needs to do a tight money policy, which drives rates to zero

    Tighter money would raise rates, at least initially, for the rates would reflect all the previous inflation that has increased profits and has put upward pressure on interest rates, but have been in turn kept down by the Fed’s counter-pressure of inflation.

    The way the Fed is holding rates down at the present time is by increased rates of inflation into the banking system. They don’t hold rates down by promising to do nothing, i.e. not inflate.

    The Friedmanite notion of “low interest rates signal money has been tight” is a consequence of a misguided inference from empirical data. Sort of like how one can infer from a chart of global temperature and pirate attacks that increased global warming signals pirate attacks have recently been on the rise.

    Connect NGDP to everything, even interest rates? Economic science tells us that low interest rates are brought about by inflation, at least initially, then once the inflation is reduced, interest rates rise in line with additional profits generated by prior inflation.

    The Fed lowered 10 and 30 year treasury rates via Operation Twist by inflation, not deflation.

  35. Gravatar of Morgan Warstler Morgan Warstler
    7. September 2012 at 13:46

    Because you are asking the Fed to pursue contractionary policy starting today along with a guarantee of no catch-up growth. That’s not politically sustainable.

    THIS IS NOT AN ANSWER. I am not asking about what is “optimal” – and again, there’s an ARMY of hegemonic interests that want rates raised TODAY.

    But that’s not my point…

    Even past MF’s point, I’m after something DEEPER.

    Scott, EXPLAIN THIS!

    There’s catch up AND there is a future full of NGDPLT.

    Nothing I’ve seen in 3 years of daily Sumner 101 says that the really important thing is the catch up to past trend.

    EVERYTHING I’ve read here says that there is real growth advantages to a stable price level knowable as far into infinity as possible.

    So let’s say we have two choices:

    Option 1: we dick around for 10 years waiting for get NGDPLT with 5 pts of make up.

    Option 2: we go 4% NGDPLT with no make-up

    WHICH IS PREFERABLE???

    OK, now that we have chosen Option 2…

    Let’s discuss shortening the time horizon on Option 1.

    If it is 5 year of waiting? Is option 1 better than Option 2?

    If it is 3 year of waiting? Is option 1 better than Option 2?

    If it is 2 years of waiting? Is option 1 better than Option 2?

    Ok what if Option 2 take 1 year and Option 1 takes 4 years?

    ——-

    Look, you all get judged by this kind of question, this is the real shit, this is how real operational considerations happen.

    We have been waiting a LONG TIME to get NGDPLT considered.

    What is the gut reasoning behind continuously trying to make NGDPLT sellable to the LOSERS?

    Why not go tot he WINNERS and turn NGDPLT into a super hard money King Dollar thing?

    Because ONCE we have it in place, that where the real magic happens right?

    ——

    Look, DeKrugman says outloud he’s just willing to try Sumner’s thing because it is a backdoor to some short term inflation.

    This IMMEDIATELY raises the issue: well then, what if we get NGDPLT and DeKrugman gets NOTHING he wants short term?

    I do not believe Sumner thinks the catch-up is actually necessary in the mid term IF it gets NGDPLT done sooner.

    So c’mon boyos, if I’m wrong WHY am I wrong? What is the assumed logic about the likelihood of getting this thing to happen?

  36. Gravatar of CA CA
    7. September 2012 at 14:50

    Simon Wren-Lewis comments on Woodford’s paper:

    “To put it another way, fiscal policy would still be a vital stabilisation tool at the ZLB even if the central bank targeted nominal GDP (NGDP). It is reluctance to accept this last point which is a particular characteristic of ZLB denial.”

    http://mainlymacro.blogspot.com/2012/09/zero-lower-bound-denial.html#comment-form

  37. Gravatar of Morgan Warstler Morgan Warstler
    7. September 2012 at 15:25

    This is Wren-Lewis, who seems not to understand a level target:

    Now suppose we have a large negative demand shock so that we hit the ZLB. We will hit the ZLB whether we have an inflation target or NGDP target. In both cases output falls below the natural rate and current inflation is too low (below its desired level). We will miss today’s NGDP target. What the NGDP target tomorrow does is reduce the impact of the shock on current output and inflation, because inflation that is too low today means the central bank will aim for inflation and output above their normally desired levels tomorrow (to hit the NGDP target tomorrow), which supports output today through expectations effects.[1]

    And then he footnotes:

    [1] Having a NGDP target tomorrow only becomes useful (and different from an inflation target) if we miss the NGDP target today. We could eliminate the ZLB constraint today by having a much more expansionary monetary policy tomorrow (which exceeded any NGDP target), but that has greater costs tomorrow (and is also less credible today).

    ——-

    1. I swear he’s talking in a circle.

    2. If there’s a level target, we’re less likely to miss, because the market as a whole doen’t pay attention to the swings, it just keeps focused on the LT 4.5% YOY, QOQ, MOM.

    3. Lets call a spade a spade here, Monetary is preferable because it is far more of a flat tax, than Fiscal.

    As such, it stymies the goals of liberals, but using Monetary to make their dreams nearly impossible.

    I think this is why Simon talks in circles.

  38. Gravatar of Bob Murphy Bob Murphy
    7. September 2012 at 17:03

    Scott,

    OK thanks for the clarification, but then, I think your post title is a bit misleading. I understand that in the context of your NGDP proposal, some economists object that it won’t work because of credibility issues, but I think your post title makes it sound like you are saying something far stronger. Like, it never makes sense for a central banker to worry about credibility, or that it is nonsense when current central bankers are very cautious because they don’t want to squander they credibility they’ve built up since the 1980s.

  39. Gravatar of Eric Johnson Eric Johnson
    7. September 2012 at 18:35

    “He thinks the probability that this might require “unconventional purchases” (i.e. non-Treasury debt) is higher than I believe it is.”

    I was under the impression that the Fed’s charter limits the types of assets that it can purchase. It can only purchase treasury and govt agency debt. Is that true?

  40. Gravatar of Saturos Saturos
    7. September 2012 at 21:20

    MF, you seem to have a habit of saying dumb things and then pretending that that wasn’t what you said. Normally in these situations I don’t reply, letting my last post speak for itself, but this time I’m going to press:

    When Floccina wrote that central banks couldn’t buy assets that appreciated with inflation (ie anything other than bonds), you contradicted him, saying, “So… they cannot buy any assets?”. We reminded you that bond prices fall when intelligent investors begin to expect inflation. You then switched to arguing that bond prices appreciate when the Fed buys them, because, you know, supply and demand. Normally I would let this sort of deliberate weaseling slide, but the thing is that this is basically how all of the arguments with you go, which is why there has been a recent spate of commenters “giving up” on these forums on the possibility of having meaningful discussions with you.

    The same goes for the deflation thing. My comment said, “…we could cure money illusion by instituting a regime of persistent deflation, large enough to offset productivity gains to nominal wages, so that workers expected their nominal wages to be lower each year…”. Now, I doubt you failed high-school reading comprehension, so you must have got that point. And yet you had to get in one last dig at your interlocutor:

    If gradual deflation sees prices fall, then the downward rigidity of wages and other prices would be reduced, not worsened. This is because wage earners and wage payers would come to expect gradually falling wage rates as a matter of course.

    Is it any wonder that people are sick of your attempts to argue the points on this blog, when you clearly lack the ability to engage in meaningful argument? For a start, you have to be able to accept when you are clearly wrong about something. This is something that I have to do all the time in real life, and even a couple of times on this blog. MF, it would do you a world of good if you learnt to do the same.

    And that is all I have to say about that.

  41. Gravatar of Max Max
    8. September 2012 at 02:29

    Scott, what you calling base demand is simply the byproduct of the Fed’s purchases. Note that money held by the public hasn’t exploded. It’s all bank reserves.

    The reserves are how the Fed has chosen to fund its purchases. In principle it could have issued bonds instead. There’s no reason to believe the macro effect would be any different.

  42. Gravatar of RebelEconomist RebelEconomist
    8. September 2012 at 07:10

    I don’t see why credibility or even expectations should matter that much, whether at the ZLB or not. If the central bank supplies sufficient base money to satisfy the demand for money as a safe asset, I would expect further money to start to bid up the price of many items other than safe assets, including new production (ie to raise NGDP), whether or not the central bank makes a credible conditional commitment to re-sell assets if prices rise by more than a certain amount. Why would I retain no or low interest money just because I know that the central bank may make a good offer to get money back later? I do not know when that will be, how much money the central bank will be aiming to buy, and how good an offer other holders of money will need to be offered to persuade them to part with theirs. I would expect money to be a fairly hot potato regardless of credibility or expectations.

  43. Gravatar of Negation of Ideology Negation of Ideology
    8. September 2012 at 10:11

    It has always facinated me that when you have a real problem right now, i.e. the slowest NGDP growth since Hoover and the associated high unemployment and poverty, people wring their hands over some hypothetical “problem” that exist largely in their imaginations.

    Why spend so much time worrying about the Fed making “unconventional” purchases when there are $11 Trillion in Treasuries? If we run out of national debt, I suggest the following in order:

    1. Dance a celebratory jig because our net interest cost has dropped to zero and will soon be positive.
    2. Buy up any government backed debt in existence, for pure risk free profit.
    3. Buy up state and local debt secured by future federal aid, for more pure risk free profit.

    The current monetary base is about $3 Trillion. Before we get to step 1, it would have to be about $11 Trillion. After step 3, it would be about $16 Trillion. If we get there and we need more we can start worrying about what unconventional assets to buy, and how to mitigate default risk.

    Until then, can we please end this enitrely unecessary man made depression?

  44. Gravatar of Bill Ellis Bill Ellis
    8. September 2012 at 14:48

    Scoot, I was waiting for this post because I wanted to see if you would address how political realities surrounding the Fed could effect Credibility of NGDP Targeting as a solution to our economic downturn.

    You did not.

    NGDP targeting is just a tool. It has no more or less inherent credibility than any other tool. Tho some tools are better suited to certain jobs, Credibility lies in the hands of those who operate the tools.

    NGDP targeting is only as credible as the commitment to the target is.

    The FED in not the monetary dictator of the United States. Does anyone really believe that Paul Volker could have brutally wrung inflation out of our economy with out bipartisan support ?

    Does anyone believe that Ben Bernanke can pursue an effective Money-debasing-inflationary-NGDP target with out the repubs turning it into an election issue in 2014 ? Who believes Romney would have Ben’s back ?

    For any FED policy to be credible it has to be politically viable.

    Markets Don’t ignore politics ‘cuz politics matter.

  45. Gravatar of Bill Ellis Bill Ellis
    8. September 2012 at 14:58

    Scott says…”I hate politics, so I almost never watch speeches.”

    You can hate politics, but you can’t ignore them. The market doesn’t.

  46. Gravatar of flow5 flow5
    8. September 2012 at 20:18

    “Volcker cut the fed funds rate by 800 basis points in the spring of 1980”

    “In response to the sharp rate cuts, NGDP soared at a 19% annual rate in 1980:4 and 1981:1”

    “Volcker tightened sharply in mid 1981, and monthly inflation rates almost immediately plummeted sharply”

    Not so. Volcker didn’t cut rates (control other short-term rates). The FFR followed the market. NgNp soared because of the introduction of the DIDMCA. Total reserves increased at a 17% annual rate of change, & the money supply exploded at a 20% annual rate – until year-end. Then came the “time bomb”, the widespread introduction of ATS, NOW, & MMMF accounts –which vastly accelerated the transactions velocity of money. This in turn propelled nominal NgNp to 19.2% in the 1st qtr 1981, the FFR to 22%, & 15.49% AAA Corporates. Volcker didn’t then tighten monetary policy. The one-time transformation of savings investment accounts to interest bearing transaction accounts (& owners taking advantage of this new liquidity), simply expired.

  47. Gravatar of Scott Sumner Scott Sumner
    9. September 2012 at 05:50

    Bill, Not watching TV isn’t ignoring politics, it’s observing it more clearly, without that emotional filter.

    And 5% NGDP growth was not at all politcally controversial for decades, and would not be controversial today. The Fed’s current tools are controversial (QE), but those are the wrong tools.

  48. Gravatar of Bill Ellis Bill Ellis
    9. September 2012 at 13:13

    Scott, 5% NGPD was not politically controversial. With the rise of the hard money Teapublicans it has become so.
    They would certainly depict getting back to NGPD trend as printing money, debasing the dollar and raising the cost of living.

    All I am saying is any arguments over credibility have to include political considerations. And I would argue that the political consideration are not in favor of credibility… And that the repubs are the impediment to credibility in this case.

    Romney Says Monetary Easing Won’t Help
    By Matthew Yglesias .

    http://www.slate.com/blogs/moneybox/2012/09/09/romney_on_monetary_policy_i_don_t_think_that_easing_monetary_policy_is_going_to_make_a_significant_difference_in_the_job_market_.html

  49. Gravatar of ssumner ssumner
    11. September 2012 at 05:43

    Bill, If it’s politically impossible to stimuluate right now, we’re screwed. And obviously fiscal stimulus wouldn’t work either. There’s not much more to say. I think 5% NGDP targeting would be politically acceptable in the long run, even to the GOP. Coming back to the trend line is a tougher sell, even I don’t favor that any longer.

  50. Gravatar of Morgan Warstler Morgan Warstler
    11. September 2012 at 07:05

    DAMMIT SUMNER!!!!!!!!!!!!!!!!!!!!!

    This is AS CLOSE AS you have come to saying:

    IF we get no make up, but got a 5% LT – this is STILL a big improvment and here is why…

    This is POST WORTHY.

    DeKrugman et al, are saying just the opposite, to them they don’t really buy it, but hey its some immediate inflation.

    They don’t get how it benefits the economy, just getting the NGDPLT switch turned on… why is the change in policy so much more important than the immediate make-up???

    Tell everyone! As a hypothetical, this is the clear line in the sand, it will move the discussion forward.

  51. Gravatar of Adam Adam
    11. September 2012 at 07:55

    Bill – You raise a good point about political constraints on the Fed, although I think both (1) Romney would indeed get Bernanke’s back because it would be more politically advantageous to him to preside over a strong economy than the play to the (apparently growing) fringe of hard money nuts, and (2) Volcker might have been able to do what he did, given the lag in the political branches’ ability to replace him, but it would take a lot of… um… courage.

    Of course, Romney is unlikely to get elected at this point.

  52. Gravatar of UK vs NGDPLT, More Debate « uneconomical UK vs NGDPLT, More Debate « uneconomical
    30. January 2013 at 15:33

    […] the last two points, Scott Sumner addressed the argument about credibility and time-inconsistency last year; to me it seems utterly bizarre to have this discussion in 2013, doubting whether the […]

Leave a Reply