A Japanese recovery?

Japan’s new GDP numbers look a bit puzzling:

Japan‘s economy expanded the most in a year last quarter as consumer spending and export gains outweighed the weakest business investment since the wake of the March 2011 earthquake and tsunami.

Gross domestic product rose an annualized 3.5 percent, a Cabinet Office release showed in Tokyo. Private consumption, making up 60 percent of GDP, contributed 2.3 percentage points to the jump. The Bank of Japan may upgrade its assessment of the economy after a May 22 policy meeting, according to people familiar with the central bank’s discussions.

. . .

Annualized real growth exceeded all but two of 36 estimates in a Bloomberg News survey. Nominal GDP, which is unadjusted for changes in prices, rose 1.5 percent, also the most in a year. The nominal gain was 0.4 percent from the previous three months, less than the median forecast for a 0.5 percent increase. 

On the plus side, it certainly supports my claim that faster NGDP growth in Japan would not merely lead to equally higher inflation.  But as Matt Yglesias correctly points out, it’s almost too good for the demand-side framework.  Faster nominal growth should boost both RGDP and inflation.

Still, I’d caution readers not to overreact to this data.  Recall that in the 15 years before the 2008 recession Japanese NGDP was basically flat (one percent RGDP and minus one percent deflator).  The Japanese labor market had mostly adjusted to low NGDP growth, although I think money illusion near the zero nominal wage increase point had still modestly depressed Japanese employment and output. 

If we use a standard expectations-based natural rate model, then a 1.5% NGDP growth rate in Japan is sort of like a 6.5% NGDP growth rate in the US., i.e. about 1.5% higher than “normal.”  If America suddenly got a 6.5% NGDP growth rate for a single quarter, and it broke down as 1% inflation (i.e. 1% less than the normal 2%) and 5.5% real GDP growth, then I think most people would see that as a sign of robust AD, even though they’d be mildly surprised by the low inflation number.  In other words:

1.  Japanese data must be seen from the Japanese perspective, not the American perspective.

2.  Quarterly results bounce around quite a bit.

I think this is modest evidence that Abenomics is beginning to work, but I also believe that Japanese RGDP growth cannot be maintained at 3.5% for any extended period of time.  Still, faster NGDP growth also makes the debt situation in Japan slightly less catastrophic.

I’ve been quite busy recently, and hence haven’t had time to blog.  I will eventually get to the old comments.



33 Responses to “A Japanese recovery?”

  1. Gravatar of TallDave TallDave
    16. May 2013 at 08:34

    I think by the end of the year the question will no longer be “did it work?” but rather why it worked.

  2. Gravatar of Geoff Geoff
    16. May 2013 at 08:50

    GDP data, even real GDP data, are notoriously over-estimated in inflationary environments. It’s the way they’re calculated. Adding up total money spent, and then “corrected” using typically sloppy price deflators.

    If the US had 100% price inflation starting this year, then the estimate for real GDP would be much higher than it really is. We might see 5% or 10% real GDP estimates, even if there are negligable changes in actual output.

    I thought it was conventional wisdom that GDP related statistics were terrible.

    Ah well, people have to have jobs and economic statistical analysis is a whole industry.

  3. Gravatar of Edward Edward
    16. May 2013 at 09:17

    NGDP is more accurate than price indexes. And it doesn’t follow that estimates for rgdp would be higher in an inflationary environment.

    And NGDI is even more accurate than NGDP

    So there


  4. Gravatar of Geoff Geoff
    16. May 2013 at 09:28

    “And it doesn’t follow that estimates for rgdp would be higher in an inflationary environment.”

    Sure it does. RGDP is estimated by observing money expenditures, not the tons of steel, the iPods, cars, computers, food, and so on, all of which constantly change and are incommensurable, not only cross sectionally, but inter-temporally as well.


  5. Gravatar of Randomize Randomize
    16. May 2013 at 10:46


    RGDP estimates should only be inflated if, as you say, price deflators are off base. As long as they’re using a chained CPI (like the PCE?) that takes into account actual changes in volume, RGDP estimates should be pretty darned close.

  6. Gravatar of Scott Freeland Scott Freeland
    16. May 2013 at 11:19


    I don’t see why it doesn’t follow that monetary stimulus can increase RGDP with low inflation in economies that are depressed. Working through the expectations channel, why wouldn’t supply keep up with demand more or less until getting at or near GDP growth potential?

  7. Gravatar of EconPerspective EconPerspective
    16. May 2013 at 12:35

    This post of Historinhas “Jobless claims jump in warning sign for labor market” is very good: http://thefaintofheart.wordpress.com/2013/05/16/another-data-point/

  8. Gravatar of J J
    16. May 2013 at 13:15

    Geoff is probably right that, ceteris paribus, higher inflation leads to higher RGDP estimates. But, that doesn’t mean that RGDP numbers are over-estimated in inflationary environments. RGDP might always be underestimated (how do we take quality improvements into account anyway?) Also, I wouldn’t call Japan’s economic environment inflationary…

  9. Gravatar of J J
    16. May 2013 at 13:19

    Professor Sumner,

    Perhaps I’m confused, but I thought the BoJ was successfully hitting 1-2% inflation targets. What happened to that? Also, your example for the US would seem reasonable now because we have below average (2% being average) inflation already; 1% inflation would not be less than what we have now. In Japan, it seems that faster NGDP growth is associated with slower inflation. And why is business investment so low given that expected consumption is increasing rapidly? These numbers seem weird indeed.

  10. Gravatar of Ashok Rao Ashok Rao
    16. May 2013 at 15:16

    “Geoff is probably right that, ceteris paribus, higher inflation leads to higher RGDP estimates.”

    Actually, “other things equal”, there’s no reason to believe this is true. Deflators aren’t *theoretically* any less worthwhile at higher inflation levels.

    Maybe the *measurement* becomes harder. Remember, inflation is a very ephemeral concept, in some ways, that we can only measure with limited accuracy (which I think is high, but limited nonetheless).

    So, ceteris paribus, I don’t think anything. As for Geoff’s point, I don’t know what the evidence is to that effect, or if can even be falsified at all. Because to do that you need to know “real” inflation, which is by definition not possible.

    And if you have a “better” measure to check CPI, why not just use that??

  11. Gravatar of J J
    16. May 2013 at 15:24

    Ashok Rao,

    I don’t know much about how official estimates are produced, so this is all speculation. I believe that if the price of a computer goes up by $100, we assume that part of this increase reflects a quality increase, so we attribute part of the $100 to inflation and part to RGDP.

    It may be that RGDP measures are always too low because they don’t account for quality increases (I believe Scott has previously posted that he believes this to be the case). In that case, more inflation would lead to MORE accurate RGDP estimates. Nonetheless, I think that price increases are partially interpreted as RGDP growth. I suppose things get complicated if inflation and RGDP move together (which is often the case) because then the percentage of price increases that are interpreted as RGDP may be too small or too large. In the case that a price increase purely reflects inflation, I think that such a price increase will result in an unwarranted higher RGDP estimate.

  12. Gravatar of Patrick R. Sullivan Patrick R. Sullivan
    16. May 2013 at 18:56

    Poor example. As the prices of computers have been coming down steadily, their prices have been declining spectacularly.

  13. Gravatar of Ben J Ben J
    16. May 2013 at 19:39

    Geoff, you’re reduced to rejecting GDP figures out of hand? You can do better than that.

  14. Gravatar of Geoff Geoff
    16. May 2013 at 19:56

    Ben J:

    No, not “out of hand.” There is a reason for it.

  15. Gravatar of Geoff Geoff
    16. May 2013 at 19:56

    And I don’t know what you mean by “reduced.”

  16. Gravatar of Benjamin Cole Benjamin Cole
    16. May 2013 at 19:59

    Great post, but I am always perplexed when economists predict that real GDP growth at some robust rate—in this case of Japan, 3.5 percent—“cannot be maintained.”

    Sheesh, look at real growth in the USA in the 1960s, or the 1990s. There were some great years in there.

    Or China.

    When there is money to be made, people will work. Invest. People unretire. Spouses enter the workforce. Unused warehouses become converted into working space.

    Lots of slack everywhere in the Western world now.

    3.5 percent? Fuggetaboutit. We can do 5, and that’s no jive.

  17. Gravatar of Daniel Daniel
    17. May 2013 at 02:35

    Ben J

    No, he can’t “do better than that”. That much should be clear by now.

  18. Gravatar of Saturos Saturos
    17. May 2013 at 04:09

    Pretty funny: https://twitter.com/DLin71/status/335121336747835392

  19. Gravatar of Geoff Geoff
    17. May 2013 at 04:49


    I always keep in mind that even my most ignorant of opponents are still capable of learning. It is those who are unsure of their own convictions who tend to assume what you are assuming. It’s usually a result of feelings of helplessness and ineffectuality that are projected onto others.

  20. Gravatar of Britmouse Britmouse
    17. May 2013 at 04:54

    I don’t fully trust my ability to read the Japanese stats, but unless I am reading this wrong:

    The domestic demand deflator was flat (0.0%) qoq from 12Q4 so the negative GDP deflator must be an effect of relative shifts in export and import prices, right?

  21. Gravatar of Negation of Ideology Negation of Ideology
    17. May 2013 at 05:26

    A related article in the Times:


  22. Gravatar of W. Peden W. Peden
    17. May 2013 at 07:46

    Britmouse posted a great 1933 public-info film on his blog-


    – that’s what I call expectations management!

  23. Gravatar of Charles A Charles A
    17. May 2013 at 08:53

    Looks like not everyone’s a believer….


    At least he only makes 20x what do, to come up with this crap.

  24. Gravatar of Edward Edward
    17. May 2013 at 09:21

    Scott Freeland,

    I think you misunderstood what I am trying to say. I am saying that monetary stimulus CAN increase real growth will low inflation, provided the economy is in a depressed state.

  25. Gravatar of Bill Ellis Bill Ellis
    17. May 2013 at 09:22

    Some highlights from Fed governor Sarah Bloom Raskin’s speech at the Society of Government Economists and the National Economists Club. May 16, 2013.

    “The increase in economic activity and the decline in the unemployment rate are, of course, welcome, but we still have a long way to go to reach what feels like a healthy economy. In fact, the pace of recovery has been slower than most had expected…”

    “…Specifically, we said that we anticipate that an exceptionally low funds rate is likely to be warranted at least as long as the unemployment rate remains above 6-1/2 percent, inflation over the period between one and two years ahead is projected to be no more than 1/2 percentage point above 2 percent, and longer-term inflation expectations remain well anchored. These thresholds are intended to make monetary policy more transparent and predictable to the public by making more explicit our intention to maintain policy accommodation as long as needed to promote a stronger recovery in the context of price stability….”

    “… In my view, the large and increasing amount of inequality in income and wealth, which has been an ongoing development for decades, may have exacerbated the crisis and I think more research is required to determine whether it may also pose a significant headwind to the recovery from the crisis for years to come. So, while I am hopeful that pressures will ease further as home prices continue to rebound, I also believe that some of the restraints on the recovery may be quite long-lasting….”


  26. Gravatar of Britmouse Britmouse
    17. May 2013 at 10:50

    W. Peden, I’ve watched that three times now, it makes me tear up a little. “Old kid Depression had us groggy for a spell but we’ve socked him right on the button with inflation.”

  27. Gravatar of Daniel Daniel
    17. May 2013 at 11:39

    the large and increasing amount of inequality in income and wealth […] may have exacerbated the crisis and

    Extraordinary claims require extraordinary proof.

    Now I know humans have a strong egalitarian streak (thanks to our hunter-gatherer nature), but the conclusion simply does not follow from the premises.

  28. Gravatar of W. Peden W. Peden
    17. May 2013 at 11:50


    There are too many great moments to mention them all, e.g. they found a “Professor of Economics”, to do the extremely exacting work of drawing graphs. In other words, “This is really complicated stuff, folks, and gee- you understand it!”

  29. Gravatar of W. Peden W. Peden
    17. May 2013 at 12:19

    The UK Labour party aim for 2% inflation and full employment, including no regional unemployment-


  30. Gravatar of dtoh dtoh
    18. May 2013 at 20:01

    I think the mistake you make is that you assume a closer more direct link between the base and inflation. In my experience, it is less direct link and is dependent primarily on a) the expected balance between short-term supply and demand, and b) a hysteresis effect depending on past inflation.

    IMHO, you will need fairly high RGDP growth for an extended period of time before you get anywhere close to the BOJ’s inflation target.

    Many (you included I think) will argue that Japan needs structural reforms (Abe’s 3rd arrow) in order to maintain higher RGDP growth. However, there is a lot of slack in productivity because of the difficulty of terminating employees and quite of lot of potential RGDP growth that can be had with no structural reform just by getting people to work more efficiently.

    My view of the mechanism is

    Higher real prices of financials assets and/or higher NGDP expectations > increased exchange of financial asset for real goods and services (equals higher NGDP) > higher RGDP > eventually leads to higher prices.

    It’s pretty clear to me that the expectation of higher NGDP growth is playing a key role.

    I believe we will see continued 3%+ RGDP growth for a couple of years with inflation well below the BOJ target. The only thing that might change this calculus is political pressure for higher wages in the spring public and private union wage negotiations. If this happens it will have been caused by the misguided assumption that the problem in Japan is deflation and the misguided targeting of inflation rather than NGDP. Deflation is a symptom not a cause and inflation is a result not a solution.

    The problem is tight money and the solution is simply more money (or to be accurate an increase exchange of money for financial assets).

  31. Gravatar of ssumner ssumner
    23. May 2013 at 18:15

    J, Remember that correlation doesn’t prove causation. I don’t believe there is any inflation in Japan right now.

    dtoh, I don’t think fast RGDP growth causes inflation, I think fast NGDP growth causes inflation.

  32. Gravatar of dtoh dtoh
    29. May 2013 at 03:29

    You may be right. I haven’t had enough time to think about this. In a way though, your formulation follows from mine. If you have fast NGDP growth and no inflation then definitionally you have fast RGDP growth, which will cause inflation.

    In my view, there are two major things that impact inflation. Actual or expected imbalances in real supply and demand and the path of past inflation.

  33. Gravatar of Sumner, the Japanese Recovery, and NGDP vs. RGDP | Last Men and OverMen Sumner, the Japanese Recovery, and NGDP vs. RGDP | Last Men and OverMen
    21. February 2017 at 03:58

    […]     http://www.themoneyillusion.com/?p=21206 […]

Leave a Reply