Yes, raise interest rates. But how?

Steve Williamson has a very good post on the Phillips curve and the Fisher effect.  Because I have been encouraging macroeconomists to stop paying attention to inflation and start focusing on NGDP, I’m all for any and all attempts to mock inflation-oriented models like the Phillips curve.

Towards the end he suggests that if we want to avoid Japanese-style nominal stagnation we will have to raise interest rates.  Yes, but how?  The ECB tried that strategy in 2011 and the eurozone went right back into a double-dip recession. Then they had to cut rates to prevent the eurozone from collapsing.  So I think we can all agree that the ECB’s way of getting to higher interest rates is far from optimal.  And yet there is some truth to Williamson’s claim that a higher inflation equilibrium will involve higher steady-state interest rates than would a lower inflation equilibrium.  But how do we get there?

The explanation is a bit complicated, so first I have to make sure that all readers accept the proposition that the reason we have 2% inflation in recent decades is not luck, and not fiscal policy, but rather monetary policy.  Other long run inflation rates are feasible. We had zero inflation on average in the gold standard period. We had 8% inflation in the 1970s. In recent decades we have 2% trend inflation and not 0% or 6% or 16% because the central banks of the world decided 2% was the right number. God knows it wasn’t fiscal policy, Reagan greatly expanded the deficit in the early 1980s just as Volcker was bringing inflation down.  Can you imagine Congress trying to target inflation at 2%?  I believe the technical term is “ROFL.”

Why the emphasis on inflation?  Because I’m going to talk from here on out like central banks control inflation, and by implication NGDP growth, at least in the long run when interest rates are positive.  If that is true, then I define an easy money policy as a policy that will lead to higher expected price levels and NGDP ten years out, and a tight money policy as policies that will lead to lower expected price levels and NGDP ten years out, as compared to before the new policy is announced.  And this brings us back to the ECB policy failure of 2011.  The reason their policy failed is that they tried to raise interest rates with a tight money policy, whereas they should have tried to raise interest rates with an easy money policy.  How do we know it was a tight money policy? Two reasons:

1.  They said so.

2.  They adjusted their policy instruments in a way that modern central bankers and asset markets recognize as having “contractionary intent.”

You might think “contractionary intent” sounds a bit metaphysical, but it is actually pretty important, as Michael Woodford showed that what really matters is not the current setting of the policy instrument, but rather changes in the expected future path of that instrument.  Markets know that central banks determine the long run path of inflation or NGDP growth, and they care a lot about just where that path is set.

I still haven’t answered the question of how you raise rates the right way.  Clearly it makes more sense to raise rates via the expected inflation and income effects (expansionary) as compared to the liquidity effect (contractionary.)  Yes, but how do you do that, and how quickly can it be done?

Unfortunately the liquidity effect is quicker, indeed essentially instantaneous.  The other effects take longer, at least at the short term end of the yield curve.  I’ve often seen monetary shocks immediately affect long rates via the income and Fisher effects, but rarely short rates.  The shortest term I ever recall being dominated by the Fisher and income effects was the 3 month T-bill yield, which fell on a tighter than expected monetary policy announcement in December 2007.  But that was unusual.

As far as the how to raise interest rates question, that’s pretty easy.  You simply target the price of NGDP futures along the desired growth path, and let markets determine the monetary base and interest rates.  I believe interest rates would rise in a fairly short time with a 5% NGDP (level) target, perhaps a year or so. But it’s not really important how long it takes, because just as inflation doesn’t matter, nominal interest rates also don’t matter.  Only NGDP growth matters.

PS.  There is nothing unusual about the ECB’s failed attempt to raise rates in 2011. The exact same thing happened in Japan in 2000, and again in 2006.  The same thing happened in the US in 1937.  It’s what central bankers do; premature ejection from the zero rate bound.  They are as anxious to tighten as a 16 year old boy.  The markets need a lot of sweet talk first.  Some forward guidance.  Easing. You know what I mean.

PPS.  Tyler Cowen recently made the following comment about money neutrality:

Milton Friedman, some time ago, wrote that money was for the most part neutral, and that the new money rapidly mixes in with the old.  That made sense to me at the time, and it nudged me away from Austrian views, yet we have seen decidedly non-neutral effects from the various QEs and the periodic taper talk. –

This puzzles me.  Friedman and Schwartz’s famous Monetary History was devoted to showing that monetary policy has non-neutral effects on output and asset prices.  The Great Depression was one example they cited.  Friedman favored QE in Japan precisely because he believed its effects would be non-neutral.  So unless I have misunderstood Tyler, I think he misinterpreted Friedman.  Perhaps he was thinking of long run neutrality, which I think almost everyone accepts.


Tags:

 
 
 

29 Responses to “Yes, raise interest rates. But how?”

  1. Gravatar of Andrew C Andrew C
    17. December 2013 at 07:40

    I think Tyler meant that money was non-distortionary, and while it might raise AD or such it doesn’t tend to favor one sector of the economy over another?

  2. Gravatar of Kevin Donoghue Kevin Donoghue
    17. December 2013 at 08:08

    A very good post, Scott? If it is, your approach to economics is totally wrong. Williamson believes that raising nominal interest rates will raise inflation, because that’s what his (rather strange) notion of equilibrium implies.

  3. Gravatar of StatsGuy StatsGuy
    17. December 2013 at 08:20

    Scott, before you post stuff like this:

    “They are as anxious to tighten as a 16 year old boy. The markets need a lot of sweet talk first. Some forward guidance. Easing. You know what I mean.”

    You might ask yourself WWSSS –

    “What would Sheryl Sandberg Say?”

    Otherwise, great post. I’m glad Tyler Cowen is considering the evidence for non-neutrality in the short term – earlier, you had been dismissive of it, but I think with TC raising comments, you might ask “if it WERE real, what MIGHT be causing it?”

  4. Gravatar of ssumner ssumner
    17. December 2013 at 08:21

    Kevin, Yes, I think he’s wrong about that, but I also think some of his critics go too far in the other direction. The Japanese low interest rate policy is NOT a good policy. We do need higher rates. The problem is that Williamson seems (and I emphasize seems) to think this can be done via a policy that the markets will view as contractionary. My claim is that the higher rates must come via an action that the markets view as expansionary.

    Monetary policy works on so many levels that miscommunication is frequent. I’m still not sure I completely understand what he is getting at.

    I meant the bulk of the post was excellent, until that one sentence that I saw as being misleading, which triggered this post.

  5. Gravatar of Kevin Donoghue Kevin Donoghue
    17. December 2013 at 08:24

    It’s worth looking at Stephen Williamson’s responses to comments to see how very wide the gulf is between him and Scott Sumner.

    Causality isn’t really an important part of the story. I’m just talking about relationships that have to hold in equilibrium. But, I think it’s clear that, if the central bank can control anything, it can control the nominal overnight interest rate. Thus, it’s useful to think of policy in terms of the overnight interest rate as a policy instrument. In general we want to think about alternative policy rules – how the overnight rate should respond to various endogenous variables – when we evaluate policy. In this case, though, my starting point is a very simple question: What happens if the central bank pegs the nominal interest rate at zero for a long time? And the answer is: no one should be surprised if this produces low inflation.

    In response to the question “are you suggesting that higher public debt by the U.S. government would now be stimulative?” he writes:

    Yes, I think part of our problem is insufficient government debt. There of course many ways to get more government debt. You can have a temporary tax cut. That’s straightforward.

    With views as divergent as this, he and Scott could share a Nobel prize, like Myrdal and Hayek, or Shiller and Fama.

  6. Gravatar of ssumner ssumner
    17. December 2013 at 08:26

    Saturos, Without knowing who Sheryl Sandberg is, it’s hard for me to consider what she would say.

    You said:

    “Otherwise, great post. I’m glad Tyler Cowen is considering the evidence for non-neutrality in the short term – earlier, you had been dismissive of it, but I think with TC raising comments, you might ask “if it WERE real, what MIGHT be causing it?””

    I presume that’s a typo, I’ve always believed money was very non-neutral in the short run. Did you mean to say “earlier he had been dismissive of it”?

  7. Gravatar of Kevin Donoghue Kevin Donoghue
    17. December 2013 at 08:26

    Sorry Scott, I didn’t see your reply before posting my latest.

  8. Gravatar of Brian Donohue Brian Donohue
    17. December 2013 at 08:27

    Scott, the yield on 30-year TIPs is 1.59%. This does not strike me as artificially low so much as the New Normal. We should get used to it.

    I don’t see rates being too low, particularly long-term rates, unless we really are gonna get clobbered by inflation at some point.

  9. Gravatar of ssumner ssumner
    17. December 2013 at 08:30

    Kevin, Good catch. Yes, he seems to be relying on the liquidity effect, which is what I’ve assumed all along. He doesn’t seem to get that you need to raise rates via the Fisher effect, but that you can’t generate higher inflation through a move the markets see as contractionary.

    It’s not clear to me why he rejects the conclusion that low rates now can (sometimes, not always) lead to higher inflation later. Markets understand this concept. If the Fed raised rates tomorrow I say TIPS spreads fall. What does Williamson say?

  10. Gravatar of ssumner ssumner
    17. December 2013 at 08:32

    Brian, I made the same point a few days ago. Trend RGDP growth has clearly slowed in America and Europe. In Japan it was already low.

  11. Gravatar of ssumner ssumner
    17. December 2013 at 08:40

    Andrew, The term “non-neutral” means “having no real effects.” There is no well established definition of “distortionary.” The real effects of monetary policy cause output to grow faster in some sectors than others. Is that a distortion? Who’s to say?

  12. Gravatar of ssumner ssumner
    17. December 2013 at 08:42

    Andrew, I would add that there is very little evidence that easy money is causing high asset prices, mostly because money is not easy.

  13. Gravatar of Kevin Donoghue Kevin Donoghue
    17. December 2013 at 08:42

    I won’t pretend to understand Stephen Williamson. As an example of where he loses me, consider again his “very simple question”: What happens if the central bank pegs the nominal interest rate at zero for a long time?

    He says the answer is: “no one should be surprised if this produces low inflation.” I say no one should be surprised if it produces hyperinflation either. Textbook theory (usually citing Wicksell) says that if you peg the nominal interest rate, anything can happen. To me it seems like lunacy to simply point to the Fisher relation and say: i=0 therefore we can solve for the rate of inflation taking the real interest rate as given. But that’s the way these freshwater guys seem to think.

  14. Gravatar of Michael Byrnes Michael Byrnes
    17. December 2013 at 09:40

    Nick Rowe has a good post on this, with a useful analogy:

    http://worthwhile.typepad.com/worthwhile_canadian_initi/2013/05/if-you-dont-like-low-interest-rates-you-want-the-bank-of-canada-to-loosen-monetary-policy-now.html

    Rowe: “The Bank of Canada is like a man balancing a pole upright in the palm of his hand. If he wants the pole to move north, he must first move his hand south, so the pole begins to lean north, so he can then move his hand north to prevent the pole falling over. Except the pole has expectations.”

  15. Gravatar of Philippe Philippe
    17. December 2013 at 09:41

    Zimbabwe showed the world how to get high inflation. Run a government budget deficit equal to 78% of GDP and raise interest on excess reserves to 900%.

  16. Gravatar of TravisV TravisV
    17. December 2013 at 10:02

    Prof. Sumner,

    Please see the new post by Yglesias below:

    http://slate.me/1fjIRrx

    Does Yglesias believe tax rates on capital should be higher? Maybe I’ve found a big disagreement between you two!

  17. Gravatar of jknarr jknarr
    17. December 2013 at 10:03

    I continue to read Williamson that he is right for the wrong reasons.

    “Liquidity premium associated with interest-bearing safe assets” really derails what is otherwise a potentially useful framework.

    The positive causal theory (instead of a posited-and-observed real liquidity premium) is overleverage — a shortage of unleveraged (low-credit-risk) fixed income, and a large supply of default-prone debt.

    As long as we are talking Fischer, take it all the way to debt deflation! There can reasonably be a debt-deflation premium — i.e. very low real interest rates — that is a function of overleverage in the economy. This is, again, a positive theory: interest rates are only the product of supply- and demand- for debt – not some magical outcome of inflation.

    In this context, inflation is monetary de-leveraging (and the major advantage to NGDPLT is that it would guarantee a 5% deleveraging per annum, regardless of the mix between real- and nominal- effects). High inflation takes down leverage (“creates more safe assets”) and so reduces the duration of the debt-deflation risk, which ought to result in higher real rates.

    Reserve-creating QE only provides more capital to keep the existing state of overleverage: it capitalizes banks and keeps leverage in place. To add to this, reserves are more temporary than currency: at any time, the Fed can reach in and withdraw $2 trillion of reserves — not so possible with physical currency.

    That is, again Williamson, reserves can be disinflationary due to debt-deflation expectations; but currency can be inflationary — it is effectively a permanent increase in the monetary base, and does not capitalize debt formation.

    The biggest nonlinearity might just be the difficulty of deleveraging a highly-leveraged economy: a titanic amount of default- or a titanic amount of inflation is necessary to clear the market and discharge the debt.

    Jacking up the policy rate will ration debt and boost defaults, and thereby clear the decks for inflationary expansion: i.e. deleverage.

  18. Gravatar of Dan W. Dan W.
    17. December 2013 at 12:04

    jknarr, I agree.

    The higher the interest rate the greater the opportunity cost for holding non-performing assets. If the central bank’s goal was to encourage economic activity it could do no better than to motivate banks to recapitalize and to allow the market to bid its price for these assets.

  19. Gravatar of TravisV TravisV
    17. December 2013 at 12:08

    Russell Redenbaugh, a macro-investor, is offering his newsletters for CHEAP:

    http://scottgrannis.blogspot.com/2013/12/reading-world.html

    2-week trial for $1.

    Anyone know whether he grasps MM concepts (like Scott Grannis does)?

  20. Gravatar of TravisV TravisV
    17. December 2013 at 12:16

    Krugman bubble talk!!!

    Check out this old paper of his from 1986: http://www.nber.org/papers/w1644

    “part of the dollar’s strength can be viewed as a speculative bubble. At some point this bubble will burst, leading to a sharp fall in the dollar’s value.”

    Is Krugman more or less respectful of the Efficient Markets Hypothesis than he was back then (1986)?

  21. Gravatar of Jim Glass Jim Glass
    17. December 2013 at 13:09

    It’s worth looking at Stephen Williamson’s responses to comments to see how very wide the gulf is between him and Scott Sumner.

    “Causality isn’t really an important part of the story. I’m just talking about relationships that have to hold in equilibrium…. , my starting point is a very simple question: What happens if the central bank pegs the nominal interest rate at zero for a long time? And the answer is: no one should be surprised if this produces low inflation.”

    Indeed. Equally, what if the central bank increases base money by 100x and then holds its amount there rock steady unchanged for a very long time. No one should be surprised if this produces a rock-steady, near-zero inflation, or even mildly deflationary gold standard-like, final equilibrium price level. The causality of how that final equilibrium is arrived at — the intervening 10,000% inflation — isn’t really an important part of the story.

    Is he really saying this?

  22. Gravatar of ssumner ssumner
    17. December 2013 at 13:14

    Kevin, Good point, but I’d be shocked if any other freshwater economists think that way. These guys also believe in efficient markets, and the markets certainly don’t agree with Williamson.

    Michael. Nick always has good posts on that topic.

    Travis, I did a post on that this morning, I’ll put it up soon.

  23. Gravatar of jknarr jknarr
    17. December 2013 at 13:27

    This discussion keeps rolling on.

    http://uneasymoney.com/2013/12/13/does-macroeconomics-need-financial-foundations/

  24. Gravatar of Martin Martin
    17. December 2013 at 15:02

    Scott,

    what Tyler Cowen is getting at are the Cantillon-effects or it matters who gets the money first. According to Friedman, and you based on your previous posts on this topic, it does not matter who gets the money first or what the central bank buys, Tyler is now saying that QE is offering evidence that it does matter.

  25. Gravatar of benjamin cole benjamin cole
    17. December 2013 at 17:30

    Excellent blogging.

  26. Gravatar of Steve Williamson Steve Williamson
    17. December 2013 at 19:09

    “Towards the end he suggests that if we want to avoid Japanese-style nominal stagnation we will have to raise interest rates. Yes, but how?”

    Easy. Raise the interest rate on reserves.

  27. Gravatar of ssumner ssumner
    17. December 2013 at 19:46

    Thanks jknarr.

    Martin, Tyler does not mention Cantillon effects, nor does he provide any evidence that QE had Cantillon effects. Yes, QE affects the price of some assets more than others, but that’s 100% consistent with the views of Friedman and myself. Money is non-neutral, that’s why some asset prices change more than others.

    I’ve argued that the Cantillon effects are trivial during “normal times” and small during the recent QE. If there is any evidence they are not small I’d love to see it. When the Fed shocked the markets in September by not doing QE 10 year bond yields fell by less than 0.2%. I’d say that’s a pretty small effect, if you are trying to explain the level of bond yields. But in any case, nothing I’ve said in this blog rules out a much larger change. My point is that you’d get roughly the same change in interest rates whether you bought US bonds or German bonds.

    If Tyler meant Cantillon effects I presume he’d use that term, non-neutral has a completely different meaning. But obviously you might be right, let’s see if he clarifies.

    Steve, I claim that action would lower TIPS spreads. The best way to raise inflation rates is to raise the NGDP target path, and do level targeting. The expectations of faster NGDP growth (if sufficiently large) will tend to raise TIPS spreads and probably long term rates.

  28. Gravatar of Geoff Geoff
    17. December 2013 at 21:11

    “Because I’m going to talk from here on out like central banks control inflation, and by implication NGDP growth, at least in the long run when interest rates are positive. If that is true, then I define an easy money policy as a policy that will lead to higher expected price levels and NGDP ten years out, and a tight money policy as policies that will lead to lower expected price levels and NGDP ten years out, as compared to before the new policy is announced.”

    That doesn’t follow. For one could also say:

    “Because I’m going to talk from here on out like central banks control inflation, and by implication the aggregate money supply, at least in the long run when interest rates are positive. If that is true, then I define an easy money policy as a policy that will lead to higher expected price levels and money supply ten years out, and a tight money policy as policies that will lead to lower expected price levels and money supply ten years out, as compared to before the new policy is announced.”

  29. Gravatar of Stephen Williamson a Man With Something for Everyone to Like | Last Men and OverMen Stephen Williamson a Man With Something for Everyone to Like | Last Men and OverMen
    17. February 2017 at 05:52

    […]      https://www.themoneyillusion.com/?p=25417 […]

Leave a Reply