05 Apr 2013

The Japanese Experiment

Federal Reserve, Scott Sumner 31 Comments

“Tyler Durden” at ZeroHedge is as excited as a kid on Christmas Eve, thinking that the Bank of Japan has finally “lost control” because of the breakout of 10 YR Japanese government bond yields (doubling in 4 hours and triggering circuit breakers).

Let me make a falsifiable prediction of which I am certain: No matter what happens, Scott Sumner and other committed market monetarists will not admit that their plan failed and that the critics hitting them from an Austrian perspective were right.

For example, let’s say the BoJ’s new commitment to massive monetary expansions plays out in textbook Mises/Hayek fashion. There is a boom for a few years, then when CPI starts rising too rapidly the BoJ raises rates and Japan plunges back into an awful depression.

Clearly, Scott will blog at that time:

[HYPOTHETICAL SUMNER BLOG POST FROM 2016: “I cannot believe we’re going through this again. Finally the BoJ started taking my advice back in late 2012 / early 2013. We had several years of rising NGDP, which in turn didn’t destroy the yen as Peter Schiff said, but in fact led to rising real GDP–the best economy in decades. Then, because CPI broke 5% (which was mostly due to one-off events, as I’ve blogged about here and here), the BoJ tightened, leading to the biggest drop in NGDP since the 2008 crisis. Unemployment shot up to 9% after NGDP growth began slowing. How many historical examples do we need to see, that show awful recessions are caused by tight money? And yet the inflation-hawks continue to beat their drums.”

By the way, I’m not even saying the above is evidence that Sumner is a bad guy. Macroeconomics has so many moving parts, that it’s virtually impossible to look at a five-year period and conclude what “the objective lesson” is. You can tell just about any story you want, Austrians included.

31 Responses to “The Japanese Experiment”

  1. Brian Albrecht says:

    But I thought all we needed to do was “look at the facts” and we will have the answer through a positivist method…

  2. Rob Rawlings says:

    If the BOJ adopted NGDP targeting (and chose Sumner’s recommended rate for the US of 5%) then the scenario you envisage (real growth plus 5% inflation) could only happen if NGDP exceeded target for some reason (since NGDP is just RGDP + inflation). In that circumstance Sumner would likely advocate some monetary tightening to bring NGDP back to trend (which would have the effect of reducing inflation).

    If however the BOJ introduced polices that took NGDP below trend and led to 9% unemployment would he not be consistent in being as outraged by that policy as he was when the fed allowed that to happen in the US ?

    • Martin says:

      Rob, Scott’s plan depends crucially on expectations. The crucial question it seems is what makes sure that people will continue to believe that the monetary base is consistent with a certain level of expected NGDP as opposed to another one?

      Let’s say that the central bank doubles the monetary base every year and says “don’t worry NGDP won’t increase more than 5% p.a.”. Why is that credible? Now let’s say that the action of the central bank results in 6% in one year, why should the public continue to believe the central bank? What happens when the central bank loses credibility?

      • Rob Rawlings says:

        I think this is an incorrect view of the role of expectations

        Expectations matter for an NGDP targeting because in a time of depressed AD then a credible commitment to target NGDP by the central bank may encourage higher spending without the CB actually needing to increase the money supply. Credibly setting expectations does the work. However if expectations alone do not achieve the target then the CB will still have the option of really adjusting the money supply to hit the target. And I see no reason why expectations settings plus real adjustments will not allow the target to be hit.

        BTW: if the CB ‘doubles the monetary base every year and says “don’t worry NGDP won’t increase more than 5% p.a”‘this is highly unlikely to be credible.

    • Andrew Keen says:

      You’re assuming that RGDP can only go up.

    • Andrew Keen says:

      Actually, it’s worse than that. You’re assuming that RGDP will, on average, grow by 5% annually. As RGDP fails to grow by 5% annually for an extended period of time, the gulf between RGDP and NGDP will get wider and wider. That sounds like a recipe for a currency crisis to me, but it I’m sure Scott would argue rising NGDP will lead rising RGDP eventually and therefore I have nothing to worry about.

  3. Major_Freedom says:

    “You can tell just about any story you want, Austrians included.”

    Dude, you did not just go there.

    The only story Austrians can tell are stories constrained by praxeology. Not any story one wants. Certainly not Keynesian stories!

    You have some wiggle room in thymology, but still…

    We work at a higher standard. Love it.

  4. Tel says:

    Correct me where I’m wrong here, but if a government needs to borrow money on the bond market and they know that they can borrow unlimited amounts of short term money at 0% (or close to), then based on any sort of supply/demand curve you care to sketch up, the seller in this situation can choose whatever yield curve seems appropriate for the longer term debt.

    They only have to sell one single 1Y bond, in order for that to be the interest rate established. Then they only have to sell one single 2Y bond, etc.

    Is there anything to prevent such blatant manipulation?

  5. TravisV says:


    TravisV here. How can you tell if long-term interest rates are “artificially low”? How can you tell if they’re “artificially high”?

  6. Yosef says:

    Bob, you have a bunch of posts on Sumner in a row, so I just want to make sure I have this straight in my head.

    Sumner had said for a while that in response to monetary expansion, long term interest rates should rise. Then, in praising the latest BoJ action, he mentions that long term rates have actually declined. You point out that such a decline is in contrast to his usual prediction on interest rate response to monetary expansion. And Sumner responds, well, that’s only sometimes the case, doesn’t always have to be that way, I never said always.

    But then, long term Japanese interest rates do in fact rise. So am I correct in saying that had Sumner kept quiet in the first place, he could now point out that “Aha! Everything is proceeding as I have predicted!”

  7. TravisV says:


    Is there ever going to be a written 2,500-word debate between you and Scott Sumner?

    • Bob Murphy says:

      Travis, when I didn’t answer the question addressed to “Bob” did you switch to “Dr. Murphy” thinking that was the reason? 🙂

      I have been super busy with work deadlines. It’s 100% my fault Scott and I haven’t debated yet.

      Of course nominal interest rates, especially long-term ones, can be high because of expected price inflation. But I still think the Fed is helping MBS owners by buying $40 billion or whatever per month, which is what Sheldon and I were arguing when Scott took us out to the woodshed.

      • TravisV says:

        Dr. Murphy,

        I don’t see how MBS purchases could push long-term treasury rates to “artificially low” levels.

        I still think the 1970’s is a major problem for your story. Back then, the Fed bought tons and tons of treasuries and long-term interest rates soared. Today, the Fed is buying lots of treasuries and long-term interest rates are ultra-low.

        Prof. Sumner’s explanation for this discrepancy? Expectations. Your explanation? I’m not sure if you even have one.

        • Bob Murphy says:

          I don’t see how MBS purchases could push long-term treasury rates to “artificially low” levels.

          I didn’t say they could. Sheldon and I were talking about one thing, and (in my opinion) Scott chimed in with a non sequitur.

          • TravisV says:

            What about T-bills and Treasuries? Are the Fed’s huge purchases of government debt pushing long-term treasury rates to “artificially low” levels?

            If so, then why did those same purchases result in sky-high long-term treasury rates during the 1970’s?

            • Bob Murphy says:

              Travis I’m not sure what you are trying to achieve here. Of course expectations play a big role in this. I think we will eventually turn into a 1970s scenario. I think Treasuries are in a bubble right now.

              My big beef with Scott vis-a-vis Japan is that he has said (as I’ve documented) that expansionary money ==> higher long-term yields as the default, normal case. But then when I zinged him on it with respect to Japan’s recent announcement, he acted like he was saying it was a crap shoot.

              • TravisV says:

                What mechanism do you think could trigger the hyperinflation? Lots of money creation? The Fed has been doing that for 3.5 years and it hasn’t even come close to happening.

              • Bob Murphy says:

                I’m glad we at least agree the Fed has created lots of money over the last 3.5 years. Scott recently praised the fact-checking of an article that denied this.

  8. TravisV says:

    Dr. Murphy,

    If you believe that “easy money” results in low long-term interest rates, then how do you explain the the sky-high long-term interest rates of the 1970’s?

  9. von Pepe says:

    This is an awesome post.

    I tried over-and-over to give Sumner a chance but he says “never reason from a price change”. Yet he continually takes victory laps when either the stock market or interest rates (price changes) go his way. Otherwise, don’t reason from a price change on the days it doesn’t correlate…

  10. TravisV says:

    Dr. Murphy,

    I read your “Hypothetical Sumner blog post from 2016.” It’s rather silly.

    To see why, imagine a scenario where a huge supply shock sent oil prices to $250 per barrel. When the central bank sees that, should they aggressively tighten monetary policy to bring oil prices down to, say, $150 per barrel?

    No no no no no, they shouldn’t do that! Surely you understand that they shouldn’t. Right?

    You may disagree with the popular central bank policy of targeting inflation at 2%. But you at least ought to admit that targeting NGDP is far far far superior to targeting inflation.

    • Major_Freedom says:

      Suppose you conflated an oil supply shock with an inflationary expectations shock, or an actual inflationary shock.

  11. TravisV says:


    First, a supply shock (OPEC gets hostile?) is an “actual inflationary shock.”

    Second, an “inflationary expectations shock” is a demand shock where NGDP skyrockets through the roof. Maybe foreigners suddenly decide they want to buy tons and tons of American stuff. In a case like that, the central bank should tighten monetary policy until NGDP comes back down to the old trend line.

    • Major_Freedom says:

      1. A supply shock is not an inflationary shock. An inflationary shock is a shock to the money supply.

      2. An inflationary expectations shock is a shock to cash preferences / spending.

  12. TravisV says:


    You could posit some scenario where real GDP falls 5% and inflation increases 10%. But that’s an implausible scenario. If real GDP falls 5%, there should be extremely little upward pressure on prices.

    • Major_Freedom says:

      Implausible….should be….

      We’re not lab rats.

      • Razer says:

        Not if Travis gets his way.

Leave a Reply