Tuesday, February 10, 2015

WIlliamson on Fisher, Phillips and Fjords

Steve Williamson has an excellent blog post "Pining for the Fjords" Point 1, the Phillips curve is dead, UK version. (And, too many are cheering, "Long live the Phillips curve!"). Point 2, Steve seems to have signed on to the new-Fisher view that a zero rate fixed for a long time, with apparently credible fiscal policy, will drag inflation slowly down, not up.

Point 1: The Phillips curve in the UK.

Source: Steve Williamson
 ...from peak unemployment during the recession, the unemployment rate drops about 2 1/2 points, while the inflation rate drops about 3 points... Presumably utilization has been rising in the U.K., but inflation is dropping like a rock.
See his post for early and late samples, core inflation, etc.
The Phillips curve is not resting, sleeping, or pining for the fjords. It is dead, deceased, passed away. It has bought the farm. Rest in peace.
 Or, borrowing another picture from Steve,

Steve continues
The Bank Rate has been set at 0.5% since March 2009. Here's the latest inflation projection from the Bank: (Inflation returns to 2% now that unemployment has decreased.) So, like Simon, the Bank seems not to have learned that the parrot is dead. In spite of a long period in which inflation is falling while the economy is recovering, they're projecting that inflation will come back to the 2% target.
The best part, which you might miss at the bottom of his post
...20 years of zero-lower-bound experience in Japan and recent experience around the world tell us that sticking at the zero lower bound does not eventually produce more inflation - it just produces low inflation.


  1. I agree, I really liked Steve's post, and the Phillips curve is looking quite deceased.

    I have continued to be unable to understand why low interest rates would result in low inflation. I can see how low inflation expectations lead to load interest rates --- you can lend money at a lower rate if you are less concerned about inflation eroding your real return. So John, my question to you and the neo-Fisherites is: Why do you think causality runs from low rates to low inflation, when it is so easy to see why it would run the other way, from low inflation (expectations) to low rates?


    Kenneth Duda
    Menlo Park, CA

    P.S. As for where the low inflation expectations come from... that's easy... look at the Fed, can't wait to tighten money even when inflation is below target and has been for 6+ years straight... no one believes that the Fed cares about hitting its supposed inflation target:



    1. Ken,

      "So John, my question to you and the neo-Fisherites is: Why do you think causality runs from low rates to low inflation, when it is so easy to see why it would run the other way, from low inflation (expectations) to low rates?"

      You are only looking at the lender's expectations. What about the borrower's?
      The causality runs from interest rate to inflation because when a debt is agreed upon, both parties are fully aware of what the agreed upon interest rate is. Neither is likely aware of what the other's inflation expectations are.

    2. Ken,

      Start an economy with no money. The first dollar is borrowed into existence at an interest rate. The price level (and inflation rate) are established after the loan is agreed to.

      The effect of borrowed money occurs after the interest rate is agreed to and the loan is made.

    3. Sorry Frank, I'm totally confused. Are you making an argument that low interest rates lead to low inflation? If so, could you please spell it out in more detail, because the fact that future inflation isn't known at the time the loan is made doesn't (in any way that I can see) cause a low rate of interest to result in low inflation in the future.

      I hope you recognize that Cochrane, Williamson, and the Neofisherites are claiming something that stands conventional thinking on its head, so I think their viewpoint warrants more explanation than pointing out the continuing coincidence of low rates and low inflation and saying, "See? I told you!"


    4. Ken,

      "...so I think their viewpoint warrants more explanation than pointing out the continuing coincidence of low rates and low inflation and saying, See? I told you!"

      Low rates are not a coincidence because of active monetary policy. And so what warrants explanation is why low rates instituted by the central bank have had little effect on inflation and economic growth.

      I don't purport that causality between nominal interest rates and inflation is established either way - too many other factors at play.

    5. Ken,

      "Why do you think causality runs from low rates to low inflation, when it is so easy to see why it would run the other way, from low inflation (expectations) to low rates?"

      Low inflation and low inflation expectations are not the same thing. So what if Neo-Fisherites believe that low interest rates will lead to low inflation and the conventional thinking is that low interest rates will lead to increased inflation expectations. Conventional thinkers / Neo-Fisherite thinkers are not contradicting each other - more like talking past each other.

      Why do conventional thinkers believe that all expectations will be fulfilled when in an economic system there is a trade off between the expectations of one group versus another (for instance the expectations of borrowers and lenders)? Because borrowers and lenders don't agree to terms on the real cost of debt and don't pre announce their inflation expectations, one of them is likely to be disappointed.

  2. Do you think it's possible to even get reliable estimates of "potential" GDP (and hence the output gap)? Personally, I don't.

    Removing market frictions would seem to be a better use of time than trying to game people's expectations.

    1. This is a huge problem for the Fed. They don't know what the natural rate of interest is, they don't know what the natural rate of unemployment is, and they don't know what potential output is, *and* they believe the effects of their policy decisions take 18 months to affect inflation and unemployment.

      I'm not sure how to remove market frictions, but I agree with you about gaming expectations being a waste of time. The best alternative I know of is targeting the level-path of NGDP using a prediction market to guide OMO's. This would be such an improvement. Rather than groping around a maze of reals with a lag, the Fed could target a single nominal and stabilize the macro-economy far more effectively than current policy.

  3. "the honesty and courage to use your real name" Folk have lots of reasons to have an internet identity that are neither dishonest nor cowardly. We're not all tenured professors in societies with First Amendment protections.

  4. Declaring the Phillips Curve dead after a couple of years of observations in the UK may be pre-mature...but the Phillips Curve has become the "Phillips Phlat-Line" in the US since the mid-1990s.

    But this cuts both ways: Since much lower rates of unemployment do not result in higher inflation, why not guns the presses and shoot for much lower rates of unemployment? You say the Phillips Curve is dead. Good!

    BTW, my eyeball of US graphs of inflation/unemployment in the USA since the mid-late 1990s suggests that you can buy a five percent decrease in unemployment rate for a one percent increase in inflation.

    In other words, inflation up one percent, an unemployment down five percent. The "Phillips Phlat-Line."

    Who would not take that trade?

    Your central bankers. Hello ECB, and Fed.

    The BoJ? They are learning....

    1. This is a nice summary of the alternative interpretation. Phillips is not gone, phillips is incredibly strong! The spread of points around the line however suggests one of those correlations that evaporates if you push on it, rather than a tight, exploitable and very strong link from a little inflation to a lot of unemployment.

    2. Great,

      How many more economists should the central bank hire?

      The Phillips curve is a relationship between inflation and employment, not money supply and inflation.

    3. Unfortunately, I cannot reproduce the "Phillips Phlat-Line" chart here, but see


      There are a lot of classic economic reasons to support the idea that the US economy is far less inflation-prone than the 1970s.

      The fraction of private-sector workers unionized is now 6.6%, down from one-third in the 1970s.

      The minimum wage is well down, adjusted for inflation (yes, the Donks are doing bad...).

      Foreign trade, unimportant in the 1970s, is now a huge fraction of GDP 9I think 1/3). No more can a Big 3 (auto) Big Steel, Big Aluminum, or Big Labor make price actions stick. Who has pricing power today (aside from Apple)?

      The Internet makes information, including price information, just about free. Craigslist has made small retailing possible--buy a great used computer for half-price.

      Retailing has gone from Sears, the Wal-Mart, to Internet to $1 stores...egads...

      Whole industries that used to be rate-regulated in th 1970s, including trucking, railroads, airlines and phones, are now rate-deregulated, along with much of finance (remember hefty stockbroker commissions and Reg Q?).

      The real question: Could we ever see demand-pull inflation in the US again? Or, since the USA sources globally (even labor, think call centers, or architects trading blueprints by PDF, even immigration), would more demand in the USA just create more and more supply?

      John Cochrane: You are hunting the wrong bear. You should be hunting large increases in demand and output, not price stability. The latter is here, an not that important anyway. We had moderate inflation in the 1960s, and 1982-2007, and we boomed.

    4. I think your examples make exactly the opposite point. Hello Uber, goodbye "sticky" prices. Without sticky prices, "demand" falls apart, inflation has no effect on unemployment, all variation in unemployment is variation in the "natural rate" and Ed Prescott triumphs. All of these are cited as reasons that recessions up to 1921 were short and swift, but not so thereafter. Even I'm not that extreme, but that is the implication of your stories.

    5. Sticky prices and wages of course.

    6. Ben,

      "You should be hunting large increases in demand and output, not price stability."

      The two are not mutually exclusive.

    7. Wow. Benjamin, it seems to me John is exactly right. Every example you gave made is an example of reduced price stickiness. If price stickiness went away completely, monetary policy would stop mattering and the Philips curve would go away completely.

      However, as a practical matter, I do not see price stickiness going away any time soon. It is 100 times easier for me to give raises than to cut salaries for my engineering staff. My long-term supply contracts with component suppliers are negotiated and fixed in nominal terms over many years. I borrow long term at fixed nominal rates. I see nominal rigidity everywhere I look (except Uber), and every reason to expect a negative aggregate demand shock to result in lost output rather than falling prices, with or without Uber, CraigsList, etc.


      Kenneth Duda
      Menlo Park, CA

    8. Ken-
      See my answer below. My main idea is that our economy is much less inflation prone than 40 years ago, ergo a lot more stimulus will not result in inflation. IMHO!

  5. John Cochrane:

    Thanks for your response.

    Well...well...I am not presenting the "get to inflation and you cut real wages and boost demand for employment" argument....

    I think we have a problem with real aggregate demand being weak. So, how do we get it stronger? Well, deflation will do it, if nominal demand stays flat.

    But, there is still stickiness, just not upward stickiness like in the 1970s. Still today, employers do not cut wages---social convention. There are other serious negative side effects to deflation; see Japan 1992-2012. I just do not see the real estate dudes doing much in a deflationary environment.

    So how to boost aggregate demand in the USA?

    Print a lot more money. It is not inflation I seek, but a honking boost in real aggregate demand. This is what Milton Friedman recommended for Japan. See:


    This was well after Friedman did his "zero interest rates are best, and with minor deflation" paper in the early 1960s.

    It is hard to reconcile his 1998 paper with that 1960s paper. I would say his views evolved and matured, but that is biased.

    Anyway, when was the last time the USA had a problem with inflation? The last time with recession and unemployment?

    If the USA sources globally today (and it does) can we even get to a bad demand-pull inflation? That would take a honking lot of demand--and growth along the way!

    Really, I think the economics profession needs to stop talking about microscopic rates of inflation, or zero inflation, or minor deflation, and start talking about getting to booming demand.

    How can we set the house on fire?

    The easiest way to go to QE hard and heavy for a real long time. Print boatloads of money.

    As you can see, I like to speak English, not economics-ese! As does the new BoJ governor Yutaka Harada. He says things like, "We need to print a lot more money."

    That is what I say.

    No, inflation, per se, is not the answer. Demand is the answer, A lot more money in circulation will do it, and likely we will get only moderate inflation.

    BTW, of course, lower taxes and regulations, and cut structural impediments. Fine. Thank goodness, we have actually done that in the USA since the 1970s (remember 90% top MTRs?).

    1. If it is more money printing you seek, you are barking up the wrong tree.


      What does QE have to do with printing boatloads of money?

    2. Or rather than printing more money, why not let go of the amounts already printed? Tell the Fed to stop referring to ultra accommodative policy, while also releasing balance sheet reduction strategies in the context of multiple years missing 2% PCE inflation. It sends the message that the target is not a target, but a ceiling, which forces the market to examine Fed reactions to determine the real zone of acceptable inflation.

      Of course, all of that can be true even if the Neo Fisherites are correct. The Fed's public actions are not internally consistent with their stated underlying logic. If you hand people money and promise to take it away given certain outcomes, people will work really hard to discover your reaction functions.

  6. "I hope you recognize that Cochrane, Williamson, and the Neofisherites are claiming something that stands conventional thinking on its head, so I think their viewpoint warrants more explanation than pointing out the continuing coincidence of low rates and low inflation and saying, "See? I told you!""

    No, it is far from a revolutionary concept. It just requires one to understand things like interest rates and how a curve goes to a flat contango, at low rates, in a commodity market. Then it makes complete sense.

  7. The Phillips curve was dead from the moment it was discovered.

    Call it the Lucas critique... A soon as policy makers started using the Phillips curve as guidance the relationship between unemployment and inflation broke down.

    When policy makers thought that they could target unemployment, the Phillips curve became a vertical line, and economists invented NAIRU to explain the failure. Now that we live in a world of inflation targets, I expect the Phillips curve to be basically horizontal.


Comments are welcome. Keep it short, polite, and on topic.

Thanks to a few abusers I am now moderating comments. I welcome thoughtful disagreement. I will block comments with insulting or abusive language. I'm also blocking totally inane comments. Try to make some sense. I am much more likely to allow critical comments if you have the honesty and courage to use your real name.