Now that’s price stability!

The Economist just published a piece that looks at the performance of quantitative easing (QE) in Japan from 2001 to 2006.  The article suggests that the policy failed, and provides extensive testimony from The Bank of Japan’s Governor, and one of its top economists.

THE Bank of Japan (BoJ) pioneered the process known as quantitative easing (QE) in 2001-06, when it massively boosted the reserves that commercial banks held at the central bank. Its verdict on how well QE worked then ought to interest policymakers today. It will also discomfort them. For all that it propped up Japan’s creaking banking system, QE did not really improve the economy nor end the country’s deflationary mindset (see chart).

Yes, by all means “see chart.”  In fact, take a very close look at the chart.  I don’t know about you, but to me that looks like perhaps the most successful monetary policy in all of world history.  Can you think of a central bank that did better?  So why is The Economist so pessimistic?  And for that matter why does Paul Krugman keep citing Japan as an example of why QE doesn’t work?  There are two reasons:

1.  Many people don’t realize that the Bank of Japan is not composed of a bunch of wimps.  When they say “price stability” is their goal, they mean price stability, not 2% inflation.  So when we evaluate the performance of the BOJ we have to judge them by the criteria they set.  How well did they achieve their goals?  At first glance that chart looks pretty bad, just as The Economist says.  The line zigzags all over.  But take a closer look.  QE didn’t start until about March 2001.  Prices had already been falling for quite some time, and QE did not immediately halt the deflation.  But this is the CPI we are talking about, which means it includes all sorts of sticky prices that adjust slowly to a weak economy.  And the Japanese economy was very weak in 2001.  But also notice that by the end of 2001 the CPI reached 98.  And then what?  As I read the graph the CPI stayed at 98, plus or minus about 0.7%, for the next six years.  I’d say that’s very stable prices.  The BOJ got what it wanted.  The chart shows QE being abandoned in 2006.  But isn’t it more accurate to say that QE wasn’t needed anymore?  The monetary base required for price stability fell sharply in 2006, and the BOJ responded by sharply reducing the monetary base.  QE is not just about increasing the base; it is also about decreasing the base when appropriate.  Even today, in late 2009, the CPI is still 98.  And way back in 1993?  It was 98.  Yes there was a two percent bubble in 2008, presumably due to oil prices, but is that really so bad?  Most modern inflation targeting proposals allow for some “flexibility” during supply shocks.  Most would say the flexibility in 2008 was exactly appropriate.  So tell me, why isn’t the BOJ performance since 2001 considered the best central bank performance in all of world history?  Can you find me a better 8 year stretch?  If price stability is the goal, I highly doubt it.  So why is this outcome viewed as being so disappointing?  One reason is that many in the West favor a positive inflation rate.  In fact I have been arguing that our current 1% inflation is too low.  But that’s not the point.  If one is going to judge the effectiveness of QE from a technical perspective, then it must be judged in terms of the BOJ’s goals, not the Fed or ECB’s goals.  And by their own standards they were brilliantly effective.

2.  Is that all there is to it?  No.  If you read the piece in The Economist they discuss all sorts of other problems in Japan.  But guess what?  All of the other problems discussed have absolutely nothing to do with the long run consequences of monetary policy.  In the long run monetary policy affects only nominal variables like inflation and NGDP, not real variables.  Notice how The Economist alludes to the slow growth in Japan during recent decades.  But is that really so surprising?   Their enviable export model maxed out in 1990.  Their population is falling.  Their service and agricultural sectors are hobbled by counterproductive regulations.  In short, their supply-side sucks.  So why wouldn’t you expect a fairly low trend rate of RGDP growth?  Banking problems?  What does that have to do with the long run trend rate of inflation?  If Japan had 3% inflation in the long run, how does that solve its structural problems?

You might be wondering whether I am serious, or I am being sarcastic.  A bit of both.  I am being completely serious that the Japanese case provides zero evidence that QE doesn’t work.  The BOJ got exactly what they aimed for.  Krugman has picked the worst possible example to criticize the effectiveness of QE.  I am a bit less certain about the wisdom of their policy goal.  Perhaps two percent inflation is a bit better if there is money illusion in the labor markets.  And unexpectedly low inflation rates might cause the debt overhang to persist for quite some time.  I recall that real estate prices in Japan have been falling for more than 15 years.  That is the “long run” by almost any definition.  So I suppose that they might have been better off with 1% or 2% CPI inflation, if it was up to me.  But it’s not my call.  The BOJ decided they wanted price stability, and they got it.  Oh boy did they ever!

HT:  Marcus.



10 Responses to “Now that’s price stability!”

  1. Gravatar of marcus nunes marcus nunes
    16. October 2009 at 18:46

    I knew you would comment on it with gusto! In any case, that´s human nature: you always judge others based on your own “values”.

  2. Gravatar of ssumner ssumner
    17. October 2009 at 05:00

    Thanks again for the tip Marcus. Even if my “gusto” is predictable, I hope the specific approach I took caught a few of my readers off guard.

  3. Gravatar of Mike Sandifer Mike Sandifer
    17. October 2009 at 07:17

    Dr. Sumner,

    Your analysis seems so clear,as do your perspectives on monetary policy in general.

    Why is it that so many economists that receive media attention seem to often offer sloppy, ill-considered interpretations of data? Krugman seems to do this all too often, even though I agree with his politics.

  4. Gravatar of Bill Woolsey Bill Woolsey
    17. October 2009 at 10:05

    What is the growth rate of nominal expenditure in Japan?

    The supply side situation is poor? That means that their productive capacity is low? Growing slowly? What?

    Anyway, I don’t agree that 2% CPI inflation is wise.

    As for real estate deflation, so what? If nominal expenditure is growing with productive capacity, falling land prices shouldn’t cause problems. Or, rather, what problems do you see?

    Did Japan ever settle into a 2% inflationary equilibrium? Isn’t a better explanation that prices were stable, There was deflation. Then the price level stabilzed at a new, lower level. Maybe that didn’t work out so well for them. I believe and hope that, evenutally, any adverse effects of the deflation would disappear.

    But perhaps this points to the value of nominal expenditure targeting. (Maybe it is just me. Why is everything looking like nominal expenditure targeting is the least bad option?)

  5. Gravatar of ssumner ssumner
    17. October 2009 at 13:32

    Mike, Thanks for the compliment. A lot of famous people are very busy. They glance at the data and as soon as it fits in with their preconceived ideas they move on. I like to really look closely at macro data.

    Bill. First of all I believe NGDP is at about the same level as 1993. This is because the GDP deflator has been falling at almost 1% a year, even while the CPI has been fairly steady. RGDP has rising around 1% a year. So Japan may have been a bit too deflationary even by your standards.

    I would distinguish between what they should have done in the 1990s, and today. In the 1990s I think it was foolsih to suddenly downshift to stable prices and even a bit of deflation in some years, at the same time their banks and property markets were imploding. Now that they have reached a steady-state you may be right. But there is a second problem. Remember I argue that modern central banks don’t know how to operate when nominal rates hit zero. Well the BOJ is a good example, as NGDP fell sharply this year in Japan, which I think was undesirable. Their long run steady trend of NGDP covers up growth around 2004-07, and then sharp decline recently.

  6. Gravatar of Doc Merlin Doc Merlin
    18. October 2009 at 00:41

    “In the long run monetary policy affects only nominal variables like inflation and NGDP, not real variables”

    I can’t agree with this. It might be true if the money was widely distributed throughout society, but the QE money would go to particular groups causing large market distortions.

  7. Gravatar of jean jean
    18. October 2009 at 07:11

    European version for this post:

  8. Gravatar of ssumner ssumner
    18. October 2009 at 09:55

    Doc, Two points; the effect has nothing to do with who the money goes to. For instance, more money does not create more loanable funds, contrary to what people think. Credit and money are unrelated. Most of the money that is created goes to tax evaders. Cash demand is mostly motivated by a desire for secrecy.

    In addition, the cyclical effects you describe would be short run, not long run.

    jean, Buiter’s right, the euro is too strong. Indeed even the dollar is too strong is the only sense that matters–in terms of the amount of goods and services it can buy.

  9. Gravatar of The Ambrosini Critique » Blog Archive » Japan data The Ambrosini Critique » Blog Archive » Japan data
    18. December 2009 at 10:59

    […] Real per capita GDP has risen 1.4% a year and yet prices have been constant. […]

  10. Gravatar of スコット・サムナーのマネー観、マクロ観 by Scott Sumner – 道草 スコット・サムナーのマネー観、マクロ観 by Scott Sumner – 道草
    7. May 2011 at 23:48

    […] e.  日本銀行はQEを採用した後、とんど完璧にゼロ%のCPI目標を達成した。任務完了(皆さんが読んできた日本についての論考の99%と反対に)。政策は”有効”だった。   […]

Leave a Reply