The Phillips Curve and interest rate targeting are dead. What next?

Larry Summers has an article in the FT where he admits that the Phillips curve hasn’t worked very well recently, and then advocates keeping interest rates near zero until inflation rises significantly.  Summers likes discretionary monetary policy, whereas I believe that’s how we got into this mess.  Stephen Williamson is also not a fan.  Here he comments on Summers:

If the Phillips curve doesn’t explain what’s going on, how do we get more inflation with continued ZIRP except through a Phillips curve mechanism? Further, Summers seems worried about the “next recession.” Presumably if the Fed still has ZIRP at that point, it’s powerless (except perhaps with unconventional tools) to do anything about it.

Next, we enter the realm of the bad analogy:

…a plane that accelerates too rapidly as it takes off may cause passengers discomfort while a plane that accelerates too slowly may crash at the end of the runway. Historical experience is that inflation accelerates only slowly so the costs of an overshoot on inflation are small and reversible with standard tightening policies. In contrast, aborting recovery and risking a further slowing of inflation is potentially catastrophic “” as Japan’s experience demonstrates. So in a world where economic forecasts are highly uncertain, prudence in avoiding the largest risks counsels in favour of Fed restraint in raising rates.

His assumption, again, is that continued ZIRP will make the inflation rate go up. But “as Japan’s experience demonstrates,” 20 years of ZIRP just serves to produce low inflation.

Of course I think that neo-Fisherians like Williamson get the causation exactly backwards—20 years of low inflation “serves to produce” the zero interest rates.  But Williamson is right to sense something is wrong with the standard Keynesian remedies.  Actually two things:

1.  The inflation targeting approach favored by new Keynesians doesn’t work well today because inflation is no longer closely correlated with output gaps.  In contrast, NGDP is closely correlated with output gaps, and hence central banks should target NGDP, not inflation.

2.  Interest rates are not a good policy instrument, as they can get stuck at zero for years, even decades.  We need a policy instrument with no zero bound (the base, forex rates, or my favorite, NGDP futures prices.)

The Fed thinks that monetary policy can go back to normal in the near future, and that they can return to something like a Taylor Rule approach.  Here’s who disagrees with the Fed:

1.  Me

2.  The 30-year bond market

3.  Stephen Williamson

4.  Paul Krugman and Larry Summers

I’d like to see NGDPLT, Krugman wants fiscal stimulus or a 4% inflation target, and Summers wants fiscal stimulus.

Perhaps it’s inevitable that big institutions like the Fed are reactive rather than creative.  It’s really hard to believe that they don’t see that interest rate targeting just won’t work in the future—that rates will go right back to zero in the next recession (assuming they rise above zero before it occurs.)  Or that IT is far inferior to NGDP targeting.

PS.  I have a piece in the Telegraph, and there’s also another Telegraph article that quotes Lars Christensen and me.

PPS.  Check out my new Econlog post.

PPPS.  Eggs are fine!

HT:  Michael Byrnes, TravisV



32 Responses to “The Phillips Curve and interest rate targeting are dead. What next?”

  1. Gravatar of TallDave TallDave
    11. February 2015 at 06:34

    It seems so obvious that investor/borrower expectations must drive nominal rates in the long run.

    As bad as the right has been on this issue, I can’t help wonder where we’d be if there weren’t a cadre of economists so dedicated to expanding the state irrespective of whether what we’re spending money on makes much sense outside of macro accounting identities.

  2. Gravatar of Kenneth Duda Kenneth Duda
    11. February 2015 at 07:32

    I totally agree that the neofisherites have causality backwards. It is so easy to see how low inflation expectations result in low nominal rates — it’s just the Fisher effect. A lender doesn’t have to charge as much if inflation devaluing future dollars less. They have no explanation for how low interest rates could cause low inflation except to argue that “the Fisher equation says so!” Sigh. I tried to argue this on Cochrane’s blog, where he has a similar post, but he didn’t respond.


    Kenneth Duda
    Menlo Park, CA

  3. Gravatar of foosion foosion
    11. February 2015 at 08:04

    Another advantage: higher NGDP seems much easier to sell to the populace than higher inflation.

  4. Gravatar of Cliff Cliff
    11. February 2015 at 08:09

    Nutrition can be a real rabbit hole, but eggs are not fine. First, while dietary cholesterol in general is fine, oxidized cholesterol is not and eggs contain loads of oxidized cholesterol. Second, egg whites are basically pure animal protein, and animal protein is consistently shown to increase cancer risk while providing no nutritional value (for people who already get enough protein, which is everyone in the U.S.). I sometimes eat a few egg yolks but it’s certainly not a health food.

  5. Gravatar of MT MT
    11. February 2015 at 08:14

    Scott, suppose the Fed adopted an NGDP or NGDPL target, as you suggest, and the NGDP futures market is consistently falling short of the Fed’s mandate, as the nominals-TIPS spread is doing today. What would you have the Fed do to reach its NGDP goal that is different from it has already tried at the zero lower bound?

  6. Gravatar of Kevin Davis Kevin Davis
    11. February 2015 at 08:20

    Lets not forget what Williamson would like to see…
    rising interest rates!
    I guess he may get his wish, although I don’t think that inflation (nor the Nobel committee) will come calling.

  7. Gravatar of ssumner ssumner
    11. February 2015 at 08:30

    Ken, I think the mistake here (and it’s not just neo-Fisherites), is to assume that the path of interest rates describes the stance of monetary policy.

    Cliff, Don’t meta-studies suggest that nutrition science is still pretty primitive? That findings might represent data mining, and are hard to replicate?

    I’ve seen so many claims made and then withdrawn over the years. . . .

    MT. The market would never fall short as I propose the Fed peg NGDP futures prices. They offer to buy or sell unlimited quantities of futures contracts at the target price. My Mercatus paper explains this more thoroughly.

  8. Gravatar of MT MT
    11. February 2015 at 08:43

    Scott, I’ll take a look at your paper, but I don’t see how this works in practice unless there is an arbitrage relationship between realized GDP and NGDP futures, like in the physical commodities market. That is, if the Fed decided to peg the April futures contract for oil at $60/bbl, investors would go buy oil today at sub-$50 and drive up the price to something close to $60. I don’t see a similar mechanism in place whereby realized NGDP could be made to converge to the futures price. Anyhow, will have a look.

  9. Gravatar of Derivs Derivs
    11. February 2015 at 08:56

    “as I propose the Fed peg NGDP futures prices”

    Scott, you confused me a bit with that too. I thought that was what you meant but the second half of the following sentence, written by you, made me unsure if that was the case.

    The Fed would peg the price of NGDP futures at $1.0365, “but only during the period where it was the target of monetary policy.”

    “I totally agree that the neofisherites have causality backwards.”

    I agree 100% with Williamson and Cochrane. The biggest error in all of economics. Sign is backwards. Inflation, without shortage of capacity, will follow risk-free rates.

  10. Gravatar of MT MT
    11. February 2015 at 09:16

    FWIW, I agree with Williamson and Cochrane as well. Consider a thought experiment: At the next FOMC meeting, the Fed announces it will pay 10% interest on excess reserves and cancels the rest of its meetings for the next 10 years. Where would the 10-year Treasury trade? Probably around 10%. Where would 10-year TIPS trade? I’d be surprised if they traded any higher than 4% real. Historically, 4% real risk-free is a fantastic rate, certainly higher than expected RGDP growth. Voila, 6+% 10-year inflation expectations overnight.

  11. Gravatar of Kevin Erdmann Kevin Erdmann
    11. February 2015 at 09:50

    Regulation of banks regarding mortgage lending will have much more of an effect on aggregate demand in the near term than the FOMC. I’m not sure if there is much difference between rates at 0.25% a year from now and rates at 1.5%. But the difference between 0 growth in real estate credit and 10% growth would be huge. And much of the difference would be real, not inflationary, because right now it’s a limit on supply.
    Growth would put us back to 2004, so it would probably just be a matter of time before the bubble busters insist on a crash. Maybe, given that reality, it would be better to limp along at the zlb.

  12. Gravatar of Kevin Erdmann Kevin Erdmann
    11. February 2015 at 09:55

    Mt, you’d get massive supply based inflation and demand based deflation. The worst of both worlds.

  13. Gravatar of Doug M Doug M
    11. February 2015 at 10:26

    The Phillips curve? As soon as policy makers said that that we can trade unemployment at the cost of inflation and started using the Phillips curve to drive policy, the linkage fell apart.

    Is this the Lucas critique in action?

    The Phillips curve was busted about as soon as it was discovered, and should be struck from the literature.

    And, the literature today still says that the long-run Phillips curve is vertical. Wrong! In a world of inflation targeting the Phillips curve is horizontal.

    end rant

    new rant

    eggs… My grandmother (an MD well before it was fashionable to be a female MD) thought that dietary cholesterol was the biggest pile of BS. If you are young and have high blood level cholesterol (LDL), it is a risk factor for heart disease. If you are middle age, and had always had low LDL but now it is rising, you are not particularly at risk.

  14. Gravatar of Matt McOsker Matt McOsker
    11. February 2015 at 11:09

    Ken, the jury is still out for me on the causality of interest rates and inflation. The one mechanical example that is interesting is when rates are zero (exclude storage costs for the moment), the spot price of a commodity is the same as the future price. But if rates were 8%, then using a forward price formula the future price would be 8% annualized higher.

  15. Gravatar of TravisV TravisV
    11. February 2015 at 11:14

    Dear Commenters,

    Any theory for why the 10-Year Treasury yield has been soaring lately???

  16. Gravatar of Njnnja Njnnja
    11. February 2015 at 11:28

    2. The 30-year bond market

    It’s even worse than that. The fed says that in 2017, ff will (on average) be greater than 3.5%. And not even the most dovish governor thinks the ff rate will be less than 2%.

    Compare that to OIS. Implied forward FF rates in that market are 1.5-2%. Clearly they can’t both be right. And if the market is correct, then there is still some time before anything like the Taylor rule for short term rates will be functional again.

    Good labor market news, Grexit, and technicals (supply coming out this week). Long overdue backup.

  17. Gravatar of Derivs Derivs
    11. February 2015 at 11:39



    Fu(t,T) = PV(t) * (1+r)^{(T-t)}

    With exception of shortage pricing and seasonality. Just throwing that in, in case someone checks your math against a real market.

  18. Gravatar of Matt McOsker Matt McOsker
    11. February 2015 at 11:53

    Yes Derivs – there are other variables that come into play. Then you can look at how businesses might adjust pricing to reflect short term financing costs. But like I said, I can see some mechanics at play, but the jury is still out for me.

  19. Gravatar of TravisV TravisV
    11. February 2015 at 13:58

    Noah Smith has a new column that is getting attention:

  20. Gravatar of benjamin cole benjamin cole
    11. February 2015 at 15:37

    Excellent blogging.
    I have posted a couple of times over at Marcus Nunes blog graphs (swiped from Tim Duy) that show the Phillips Curve has become the “Philips Phlat line.” This is great news, actually. I believe there is plenty of room for stimulus without much inflation. The trade-off has radically improved since the 1970s. It is a cause for celebration, but some people find another reason in a better trade off why we need an asphixiating monetary policy.

  21. Gravatar of Donald Pretari Donald Pretari
    11. February 2015 at 17:30

    Scott, I will tell you what a financial crisis caused by moral hazard looked like, because I lived it. I began seriously reading econ blogs and posting comments on them in August of 08 based upon reading the following post:

    I couldn’t understand this. My view was that Implicitly Guaranteed agencies like Fannie/Freddie were actually Explicitly Guaranteed by our govt. Why, then, was China exiting Agencies?

    “But foreign central banks have dumped nearly $11 billion from their record holdings of this debt in four weeks, to $975 billion, and won’t return in force before it’s clear if “” and how “” the government will back Fannie and Freddie”

    There’s no doubt that foreign central banks had been buying Agencies with the understanding that they were explicitly guaranteed, and China said as much at the time.

    Here’s what William Gross was advising in July of 08 to future President Obama:

    “Its potency regarding inflation will not be felt fully during the peak deficit period. Rather, inflation will accelerate during the subsequent recovery as the government bonds acquired during the recession are transformed once again into risk bearing assets and high levels of investment. That suggests that intermediate and long-term yields on government bonds have already bottomed and will gradually rise throughout your first, and perhaps second Administration.”

    I mention China and Pimco because they invest real money. Both China and Pimco
    presumed explicit guarantees, although China, I guess, sensed something funny about our level of commitment. Fannie and Freddie was saved, but, by that time, confidence of bailouts was eroding.

    On the day preceding that fall of Lehman Brothers, a special trading session was held during which investors could buy Lehman assets on the cheap. No sales.
    Zero. Instead, the talk was all about how fast Merrill-Lynch would fail if Lehman failed. On that Monday, investors with serious money began churning assets in a flight to safety. They obviously wanted explicitly guaranteed investments. And that’s what happened, although it took until Citibank was bailed out in November for the policy of everything is guaranteed to be understood, and the Vix, for example, finally began to settle.

    Prior to Lehman’s collapse, there was already in place the beginnings of a recession and deleveraging of the housing market. But the panic only came about when govt guarantees of important US financial assets seemed unreal, and it took the statement of complete guarantees to restore order. So, the crisis was caused by not correctly understanding the moral hazard underlying the financial industry.

    Absent this crisis, it’s possible that the recession and deleveraging of the housing market could have been accomplished without a panic, meaning that there was an unnecessary loss of wealth and well-being due to this mistake.

  22. Gravatar of Sina Motamedi Sina Motamedi
    11. February 2015 at 23:19

    That econlog post is great.

  23. Gravatar of ssumner ssumner
    12. February 2015 at 11:15

    Derivs and MT, We all know how inflation can cause high rates, what is the causal mechanism for high short term rates to cause inflation?

    BTW, I very much doubt whether Cochrane would agree with that claim. Also, why do TIPS spreads rise when the Fed cuts its fed funds target by more than expected?

    Donald, Good point about moral hazard. We handled that really poorly.

    Thanks Sina.

  24. Gravatar of Derivs Derivs
    12. February 2015 at 17:59

    “We all know how inflation can cause high rates, what is the causal mechanism for high short term rates to cause inflation?”

    I think it is a chicken-egg problem. I had never thought about it until one day when I heard a central banker mention raising rates to defend a currency and lower inflation. 99.99% of the press seems to agree, when fact is that it actually devalues a currency in term. Therefore rate increase pressure is actually inflationary. A currency devaluation in itself is a form of inflation. A commodity producing country tied to the dollar would be in deep recession by now. It is the devalution that appears to be acting as a hidden reflator for your NGDP, if I understand you correctly?

    As for your question about a causal mechanism for short term interest rates driving inflation. Let’s go to the shortest term of 1 day, let’s assume we are at 0 inflation, and let’s, for whatever reason, blow out available overnight rates.

    I am in a transaction to purchase a car from someone for $100,000, this is all the money I have in the world. As I am about to write the check, the news announces that overnight rates went to 100%. I ask to sign the contract, but sneaky me informs the dealer that I can not find my checkbook, so I ask if I can make payment tomorrow morning. The dealer doesn’t realize what is happening and says OK. I am ecstatic because leaving that $100,000 in the bank will double my money to $200,000, essentially allowing me to have the car for free. Unfortunately for me, there was another customer who was looking at the car and overhears the conversation, he is also aware of the news about the overnight rate. He informs the dealer that he wanted the car, but being polite did not want to interfere with the transaction. Giving me a dirty look, this ethical gentleman offers the dealer $150,000 for the car, payable tomorrow. Regaining my memory, I find my checkbook and indignant at this man interfering with my transaction, I bid $190,000, payable tomorrow. Up to $200,000 I go… Price is being set by rate, never is an inflation expectation relevant in this pricing issue.
    But even at lower rates, this is what is happening. In Brazil I will receive almost 2% this month on my TIPS. My spending can/will increase accordingly. If I was getting 0, my spending would have to remain the same. Interest rates appear to become the vehicle for pulling prices up and the need for money to be pulled into the economy.

    You would know a lot better than I would, but I believe all hyperinflations occurred with very high short term rates (obviously, just not high enough to make inflation turn). Which is another thing, my basic understanding is that raising rates is the cure for inflation. That would suggest an inverse relationship. Yet clearly the relationship is positively correlated, it would take some Herculean micromanagement to make an inverse relationship appear positively correlated. So I am totally confused there, too.

    The TIPS question, I am not prepared to answer, but if you have examples readily available I would appreciate you identifying which moves you are discussing. I never marry my opinions and actually seek to destroy them at every available opportunity. It is only what is left after I take it behind the woodshed and beat it that I remain stuck with. Honestly, I doubt I would have a good answer regardless. Aren’t you now asking me to reason from a price change, a point with which I never disagreed with you on. Possibly it’s a knee jerk reaction? Misinterpreting of news? Much as I mentioned above about 99.99% of people erroneously believing raising rates strengthens your currency. Heck original theory with QE, if I remember, was keep long rates high because low short term rates were going to cause ‘booku’ inflation later.

    Well I do hope I made some sense. I am suffering the opposite of you. 120 heat index today. Makes the brain like mush.

  25. Gravatar of Major.Freedom Major.Freedom
    12. February 2015 at 19:08


    Noah Smith’s posts always contain fallacies. Every single time. And they always seem to somehow lead to government as solution.

    “…there’s a strong possibility that recessions don’t naturally heal.”

    “If Farmer is right, then no matter how rational it looks, our economy is driven by the vagaries of shifting human expectations. It is not a self-correcting system like Milton Friedman envisaged, but a fragile thing. And that opens up the possibility that maybe we need government to knock us out of the bad equilibrium — somehow.”

    That is a non sequitur. If it is true that “vagaries of shifting human expectations” are what drives people, then it would necessarily drive government agents. There is no other human source of corrections than the existing human population.

    Using Smith’s unorthodox logic, if governments agents are driven by “vagaries of shifting human expectations”, then that would “open up the possibility that maybe we need markets to knock us out of the bad equilibrium.” But that contradicts Noah’s stated belief. His conclusion does not actually follow. It is made up. Asserted with the very “vagaries of shifting human expectations” that he claims to be able to transcend by likening his mind with government.

    All he is saying is that violations of property rights has better outcomes than respect for them.

  26. Gravatar of Nick Nick
    13. February 2015 at 03:20

    sorry if someone’s already posted about this but it made me angry. SF Fed president John Williams to FT:

    ‘However Mr Williams dismissed such calls, warning of the risk that the Fed gets behind the curve on inflation and that it could end up being forced to hike rates “much more dramatically” to rein in inflation, provoking market turmoil. Given the trails with which monetary policy operates it was better to start raising interest rates “gradually, thoughtfully”, he said.’

    So he’s saying loose money right now would lead to … Higher rates later? And that by raising rates earlier we will lower the rate we eventually end up raising it to? Whoa–breakthrough stuff! But wait, the conclusion is that we need to raise rates earlier?

  27. Gravatar of ssumner ssumner
    13. February 2015 at 06:25

    derivs, Your mistake is to talk about changes in interest rates as a monetary policy, not the effect of a monetary policy. We know that central banks can raise rates with either tight or easy money policies. So never reason from a price change, first ask why rates rose. The liquidity effect or the Fisher effect? And we have a mountain of evidence that when the central bank raises rates with tight money it causes the currency to appreciate, and the inflation rates (TIPS spreads) to fall. That point is not even disputed. And when they raise rates with an easy money policy inflation will rise and the currency will depreciate.

    I would suggest checking out the dramatic market reaction to the unexpectedly expansionary January 2001 Fed rate cut. Short term rates fell due to the liquidity effect, and inflation expectations rose. Long term interest rates rose due to the Fisher effect, and inflation expectations rose. That pretty much covers the bases.

    Nick, Thanks, I’ll do a post.

  28. Gravatar of Derivs Derivs
    13. February 2015 at 09:08

    “And we have a mountain of evidence that when the central bank raises rates with tight money it causes the currency to appreciate…”

    You asked me to demonstrate how interest rates can move price. I replied, one input-one output. You are now moving the goal posts by introducing a second variable into the equation. A tight monetary policy should cause a currency to appreciate, a loose one – depreciate. I agree 100%. So of course you can offset what I said by now introducing that variable.

    “first ask why rates rose. The liquidity effect or the Fisher effect?”

    I have found that those who look at markets from a non-participatory perspective are always going to have a considerable amount more belief in fundamental factors explaining market moves than I will. Personally I always attempted to remove subjective interpretation from models. I don’t believe there is anyone in this world who knows the fair value of a traded asset. The only way you can know anything about value is how the market tries to find equilibrium. The market is just a giant feedback loop always searching for equilibrium. If you wish to assume the search for equilibrium is predominately fundamental and not predominately mathematical, you are not looking at how things are actually being traded. The overwhelming majority of volume that trades daily, the genesis of that volume is mathematical.

  29. Gravatar of ThomasH ThomasH
    13. February 2015 at 14:03

    @Tall Dave,

    “I can’t help wonder where we’d be if there weren’t a cadre of economists so dedicated to expanding the state irrespective of whether what we’re spending money on makes much sense outside of macro accounting identities.”

    I guess we’d be where we are since there is no such “cadre” and even if there were, who would pay attention. Governments act as if they are borrowing-constrained consumers.

  30. Gravatar of ThomasH ThomasH
    13. February 2015 at 14:30

    While I agree that it looks like NGDP targeting could be better, we ought to recognize that IT has not really had a fair test. Monetary authorities have never acted as vigorously when inflation was below target as they have when it is above. They have treated it as a inflation CEILING target. And if they did the same with NGDP how do we know things would be better?
    If the Fed were committed to buying/selling NGDP futures that would certainly stabilize NGDP but If the Fed were committed to buying selling whatever it takes to keep inflation on target that would also stabilize NGDP (maybe not as well but relative to post 2008, better than missing their target). It seems that the issue is the unwillingness to buy/sell without limit to meet the target more than the target itself.

  31. Gravatar of TallDave TallDave
    13. February 2015 at 20:21

    ThomasH — Oh, if only that were so. Instead, one of the most influential economists claims preparing for a nonexistent alien invasion would boost the economy.

    Governments act as if they were people spending money seized from other people for other people, because they are.

  32. Gravatar of ssumner ssumner
    14. February 2015 at 08:36

    derivs, Individual trades are often random, the market is rational.

Leave a Reply