Sunday, May 23, 2010

Japan's problem is supply, not demand (updated)

Paul Krugman wrote in the NYT that we are talking too much about Greece: "Despite a chorus of voices claiming otherwise, we aren’t Greece. We are, however, looking more and more like Japan."

According to Krugman the U.S risks ending up like Japan, because of "policy makers... doing too little".

First the cheap shot. A year ago, Krugman wrote "Well, I’m sure I’m not the only person to notice this: Japan doesn’t look so bad these days."

Does the U.S risk becoming a new Japan if we don't pursue even more Keynesian spending and borrowing policies?

Let's first look at the lost decade, 1991-2000. When the rest of the world was having rapid, IT-fueled growth, Japan was stagnating. Here are the growth rates in real GDP between 1991-2000:

For all the nice years Japan had 9.6% growth compared to 38.7% for the U.S and 22.7% for the EU.15. The U.S grew by an average of 3.7% per year, Japan only 1.0% per year.

But as most of you know Japan is undergoing a rapid demographic transition. The country was and is aging. Because the old and children cannot work, when we want to compare countries with very different demographic characteristics instead of calculating GDP per capita, it makes sense to calculate GDP per working age adult (people aged 15-65).

Whereas the number of potential workers in the U.S increased by 13% during Japans "lost decade" (1991-2000), and by 3% in for example France, the Japanese potential workforce actually shrank during these years. Adjusting for this, the growth in Japan was 9.8%, compared to 16.9% in Germany, 17.3% in France, 16.3% in Italy and 23.2% in the United States. The U.S grew by twice, not four times of Japan (remember that these were the best years of the U.S and the worst years of Japan).

The importance of the demographic transformation in Japan is even more clear if we include the entire 1990-2007 period.

In non-population adjusted figures, Japan's real GDP grew by 26% in total these years, the lowest in the OECD. In comparison the figures are 63% for the U.S and 44% for the EU.15.

But during this period the U.S saw it's potential labor force (the number of people between 15-65) increase by 23% and the EU.15 by 11%, while Japan had a decrease of 4%.

Between 1990-2007, GDP per working age adult increased by 31.8% in the United States, by 29.6% in EU.15 and by 31.0% in Japan. The figures are nearly identical!

Japan has simply not been growing slower than other advanced countries once we adjust for demographic change.

Also notice Italy (who does better than we think) and Ireland (who does worse, much of the growth was due to their young population).

Nor did productivity grow any slower in Japan than Europe.

Someone could ask why Japan did not outgrow the U.S, since they started at a lower level, or why the old Japanese don't work more to keep up income. But there has really been no dramatic change in institutions during this period, and thus little reason to expect Japan to catch up with the U.S. Japan is already as rich as Europe, so there is no catching up there. And at any case, there is nothing Keynesian deficit policies can do about a shrinking workforce.

Next, to Krugman's point that the problem is "policy makers... doing too little" (by which he means spending too little). Japan has been running Krugman-Obama sized deficits averaging about 5% of GDP for a decade and a half.

Here is their national debt as a share of GDP. Europe 4 are Germany, U.K, France and Italy.

Clearly, it did not work. Krugman is simply dogmatic when he claims that Japan's policy of massive deficits failed because the deficits were not large enough(!). What if someone wrote that the deregulation of the American financial markets did not work just because they did not go far enough? Would Krugman accept this line of reasoning?

Krugman is obsessed with demand, and ignores the (usually) far more important factor, which is supply.

Does the U.S. risk being the next Japan? Probably not, since the American workforce is growing. And at any case adjusting for population Japan has simply not been doing that badly in growth terms. Their problem now is their debt, which they have thanks to Keynesian policies.

Capacity utilization is high in Japan, including a low unemployment rate. Stimulating Demand just won't do it when the problem is supply. If Japan wants growth they have to go for supply factors, including hours worked.


Here are 4 pictures I hope will shed more light on the story.

First, this is the impression of Japan we have been having. This is just the size of the total economy. Notice the U.S and western Europe growing, while Japan is a flat line?

But now look at the picture when adjusted for the size of the potential labor force:

Japan is growing slower than comparable economies in the first half of the period, no doubt, especially since they started of at the top of the bubble. But not much slower than Europe, even though western Europe included initially poor, fast growing regions such as Ireland, Spain, and of course Greece.

Here is the conventional story: Japan's GDP compared to EU/US. Bad years, followed by stagnation:

If only they could jump start demand somehow!

Here is the more nuanced, complex picture, when adjusting for 1 supply factor (potential workers).

Japan also illustrates the problem of over-zealous, imprudent Keynesianism. If they had not undertaken massive deficits in the 2000s (when there was no need, and perhaps not even opportunity, for policies aimed at stimulating demand rather than supply) they would have had dry powder now. There is no guarantee that 2 crisis cannot come within a couple of decades. Instead, the Japanese state is immobilized by their fiscal past.


  1. If the driving difference between Japan and the US is this changing age demographic or workforce, why haven't any economists pointed this out?

    I'd imagine the labor force variable is so well known, that someone must have performed an analysis controlling for it. Not to demean your observation here, but is this really just the answer?

  2. "why haven't any economists pointed this out?"

    The myth that Japan needs so much stimulus serves the interests of statists everywhere - in the USA, but most of all in Japan! Think of all the cash the politicians have been able to distribute for the last 20 years.

  3. @anonymous1 'statists everywhere' usually does not describe economists, at least in the states.

  4. I am obviously not the first person to realize this. googling for a second, here is some more.


    Your question is really "why don't they pay more attention to this? do they know something we don't"?.

    Beats me. There are plenty of important facts that do not get full attention in the public debate. Especially if people in the debate (such as Krugman) have a ideological agenda.

    Also, let me be proactive. Somebody wrote:

    I agree with most points in your article on Japan's "lost decade", particularly that Japan's problems are much more related to supply than demand (even in the 1990s, low productivity growth was a big factor
    holding Japan back, on top of the population

    However, I think the 1990-2007 comparison just goes to show that Japan really only had one bad decade (the 90s), while the U.S. had a quite bad decade in the 00s (allowing Japan to make up some of the lost ground). I think our economy's poor performance in the 2000s (flat stock market, for example) and Japan's good performance under Koizumi cannot be ignored.

    I answered (edited):

    1. It is not just the U.S. If it was just Japan and the U.S, fine (although the U.S is the technology frontier so I think our growth matters for what is possible for other countries
    that are already in the steady state).

    Japan did approximately as well as the average for 15 west European countries, Canada, New Zeeland (Australia did better). Beside these countries, who should we compare with?

    2. The point is not that Japan did not have a lost decade. The point is that the image of Japan stuck in zero growth is false. Japan seems to had already fully recovered from the lost decade by 2007, although this seems clowded due to their low total growth.

    3. The full period is important for evaluating
    Krugman-advocated Keynesianism.

    First, there is the question of substitution between periods. Here is a simple story. In the 1990s, While the IT-boom and other innovations was going on in the U.S, parts of Europe and
    Japan had their own problems. A few years later those 2 regions were growing fast by incorporating these technologies,
    while the U.S was growing slower because they had already incorporated most of them.

    Does it really matter that much if supply driven growth comes now or 7 years later? Is it worth experimenting with the economy that much? (and there is little evidence that Keynesian policies facilitate supply driven growth).

    What would Japans gains have been if they did even more Keynesianism, if it succeeded, they and caught up 1999 instead of a few years later? Pretty small.

    However the problems caused by Keynesianism, the 250% debt over GDP, are going to drag down Japan for years to come.

    4. Lastly, the eventual catch-up matters GREATLY for evaluating the risk of having a "lost" decade (which in Japans case was
    10% growth instead of 15-20% of other countries, not zero as the media portrays it).

    Japan, Finland and Sweden eventually caught up the ground they lost during the lost years. Me and you know this, business cycle driven problem do not effect long term growth. If you
    have empty capacity you eventually grow faster. But in the current debate the "lost decade" is portrayed and perceived as permanent lost ground, and permanent slow growth.

    Would people be so scared of "the lost decade" if they understood that (worst care) the U.S might grow less than average for 6-7 years, and then faster than average the next 6-7 years, so that we in 2023 will be exactly where we would
    have been without intervention?

  5. It might be worth pointing out that one reason (not the only reason, by any means) that Japan's population has grown so slowly is that immigration into Japan is much, much smaller than into almost any developed country. And that this is a consequence of policy actions. That is, one of the sources of polulation growth in most coountries is, as a matter of policy, not available to the Japanese economy.

    I am emphatically not taking a position on whether that policy is right or wrong.

  6. What about the point that in Japan's lost decade growth was not flat. Rather, they had several growth cycles get started but once they were in an expansion mode the government implemented higher taxes --tight fiscal policy -- or tight money that aborted the recoveries.

    Your analysis is ignoring too much in the way of policy mistakes that were major factors in the lost decade.

  7. @Spencer:
    'Rather, they had several growth cycles get started but once they were in an expansion mode the government implemented higher taxes --tight fiscal policy -- or tight money that aborted the recoveries.'

    This is exactly what Keynesian policy says we should do. Yet, it seems to cause horrible problems.

  8. Your numbers for Germany are wrong, there is no way the labor force increased by 23,3%. The only way for that to be true is if you count reunification as an increase in the labor force, in which case there is probably something wrong with your gdp numbers, (ie you're comparing the former West Germany in 1990 with all of Germany in 2007). I was wondering why Germany was such an outlier and that mistake would explain it

  9. @John - economists as a whole are slightly left of centre, with very few espousing libertarian beliefs. Those that do would share the view that Japan is an excellent case in point for the uses of keynesian economics to drive outcomes convenient for the political class.

  10. Aslak:

    It is not a mistake, it is deliberate. Adding 15 million Germans to Germany naturally increases its workforce.

    Since the GDP figure includes the GDP of the East Germans, the figure is the correct one (both Germanies).

    Why did East Germans reduce average German GDP? Who knows. However reunification is like immigration, you don’t get to exclude them from the national average.

  11. You are saying the 1990 GDP figure inclueds both Germanies? That doesn't seem to match up with GDP figures from the German federal statistics office.
    And with all due respect reunification is like immigration only in the sense that it increases the labor force, but there are so many other ways that reunification and immigration are not the same that any comparison is meaningless. It is pretty clear that the apparent underperformance of Germany is due to the reunification -if you take 1991 or 1992 as a starting point Germany given the (slight) decline in the size of the German labor force over that period.

  12. Never mind, I checked the OECD date. The growth for the combined Germanies is indeed 34,1%. But that means you can't add the East German labor force since East Germany is already included in the 1990 numbers -essentially you're adding the East Germans twice.

  13. I am not adding East Germany twice. The OECD already added it for me. I am just following their methodology.

    You are right that German under-performance is to a large extent due to east German poverty and the failure for them to integrate. Their culture and human capital were harmed by communism.

    West Germany as a region is still quite rich.

    The same, I suspect, is true for Northern Italy.

  14. I realize we're arguing over a small detail, which is a pity because overall I think this post is excellent, but nevertheless: Your GDP number for 1990 and for 2007 is for all of Germany including East Germany. The labor output of East Germany is therefore included in the 1990 number, and there is no increase in the labor force over the time period. At what point would the increase occur?

    Anyway, I think your last graph is key -it would be interesting to look at GDP levels per working-age adult across countries - I have a hunch it could really challenge some conventional wisdom. Demographic factors are generally overlooked in economics and I'm glad you're pointing out the importance of it.

  15. As I understand it the 1990 number for German workforce does not include East Germany.

    The German labor force goes from 44250.000 in 1990 to 55310 in 1991, which I think is reunification.

    I also believe that the 1990 number for GDP does not include East Germany, but that this (again) is added on 1991.

    So Germany becomes poorer during the period 1990-2007 (and especially 1990-1991), because poorer east Germany is added, and not yet absorbed.

    I think it is correct to accept this inclusion of unproductive East Germany, just as we accept the inclusion of unskilled Hispanic immigrants into American GDP.

  16. Not saying you are wrong here, but the change in GDP per worker, can be attributable to either productivity (the Solow residual) or the change in capital per worker (or, of course, a combination of both). If, for instance, the capital stock has remained largely constant, and with a declining workforce, we would expect to see an increase in the productivity of the remaining workers. Thus, ignoring the effect of capital per worker can clearly bias the numbers upwards.

    Brilliant post, otherwise! And brilliant blog too.

  17. Yes for productivity what you say (and many other similar biases) are a huge problem. You make a very good point.

    For GDP, well, I would argue investments should adjust beforehand for population trends. The firms know which market and workforce is growing and which is not.

  18. This is an interesting analysis but your conclusion veers off at the end to attack keynesian economics.

    "Japan also illustrates the problem of over-zealous, imprudent Keynesianism. If they had not undertaken massive deficits in the 2000s (when there was no need, and perhaps not even opportunity, for policies aimed at stimulating demand rather than supply) they would have had dry powder now. There is no guarantee that 2 crisis cannot come within a couple of decades. Instead, the Japanese state is immobilized by their fiscal past."

    Um, so you're saying that Japanese policymakers in the 90s should not have enacted fiscal stimulus do deal with their current crisis, to keep some "dry powder" for a crisis 20 years down the road?

    And yes, the 90s were a period of crisis for Japan. You say there was "no need" for stimulus in the 90s given real gdp per worker turned out ok. But that is partly because there was stimulus!

    I think the fundamental flaw in your analysis is that it misses the balance sheet issues. The government *had* to run up debt to offset the deleveraging in the corporate sector. If not the downturn would have been far worse. Fiscal policy was not the best tool (I would have favoured monetary policy, specifically a credible inflation target and actual quantitative easing) but coming off a massive asset bubble policymakers were scared of unorthodox methods. So fiscal policy was a necessary but second best method to prevent the bubble collapse from doing even more damage, similar to the 1930s US.

    So I see the data you see and draw the precise opposite conclusion: Japan managed to do fairly well because of Keynesian measures. The chief drawback of Keynesian measures is the resultant debt, but in Japan's base this is a feature, not a bug.

  19. 1.

    "Um, so you're saying that Japanese policymakers in the 90s should not have enacted fiscal stimulus"

    Don't use straw-men if you have real arguments. I said they did TOO MUCH Keynesianism. Never did I say that they should not have done *any* Keynesianism at all. I certainly don't believe that, I believe in sophisticated, moderate Keynesianism (both theoretically and as a policy took).

    And if you read my post more carefully, you will realize it is a critique of Keynesian theory, not Japanese policy.


    "to deal with their current crisis, to keep some "dry powder" for a crisis 20 years down the road?"

    Yes. If running current debt you risk risking a deep crisis 20 years from now, you should borrow less. And more importantly you say pay back, as soon as possible. The crisis was in 1991, Japan had a deficit as a share of GDP of 5% on average 1995-2005!


    "And yes, the 90s were a period of crisis for Japan. You say there was "no need" for stimulus in the 90s given real gdp per worker turned out ok."

    What are you talking about? You quote me saying "no need". But I specifically said that for "the 2000s", not "the 90s".

    4. "But that is partly because there was stimulus!"

    The Japanese economy would have gone into deeper debt without the 1990s stimulus. But I never disputed this, or criticize those policies, which I have no reason to disagree with. That is in your head. I criticize policy for running debt later, in the 2000s, in want trying to keep of growth for a country that was shrinking due to demographic factors, when they should have been PAYING BACK some of the loan so they would not be in crisis now.

    Also, there is exactly 0 effect of Japanese gdp in 2007 (my end period) from fiscal stimulus in 1991. Zero.

  20. I'll concede that I misread you about "no need" for stimulus in the 2000s rather than the 90s. Apologies.

    I understand this post is focused on Keynesian policy but I'll just add one more thing re Japan.. part of the reason I misread your post is that IIRC the Japanese govt was rebuilding its finances for most of the 2000s:,3343,en_2649_34113_40372839_1_1_1_1,00.html

    Yes debt/GDP continued rising in the 2000s as primary deficits were still being run. But the budgets were moving in the right direction, and debt may have stabilised absent the 2008 recession. Hence it is not clear to me that Japan was attempting to offset bad demography with debt.

  21. thanks for the link ravage:
    the deficit figures are:

    2002: -8%
    2003: -7.9%
    2004: -6.2%
    2005: -6.4%
    2006: -2.9%
    2007: -3.4%

    These are worse than Bush figures (Bush had a deficit of -1.2% of GDP in 2007). And Bush had a much lower debt as a share of GDP to start with, and a much faster growing population (which reduces the need to pay down the loan, you can just outgrow it). No one accepts the argument that Bush was a fiscal conservative.

    When a country is consolidating its budget, during good times there should be a surplus, preferably a large one. Just reducing the deficit a little is not enough.

    This is what Sweden and Finland did after they had crisis in 1991, comparable to the one Japan had (in relative terms). They had surpluses for many years, to pay down the debt.

    And we have not even taken into account cheating with the figures, which many suspect is going on.

    The Japanese government has a large responsibility for their current situation.

  22. Interesting post, but there are at least four problems: (1) Japan's growth of 26% from 1990-2007 reflects growth of about 14% from 1990-2, and a touch less than 1% per year since. (2) Cultural norms in Japan are such that many more people over the age of 65 work, (3) the correct metric is the "difference-in-difference" -- japan was converging rapidly to western living standards up until 1992, then stopped converging, so all else equal, japan should have continued to grow faster to catch up, not grow at the same rate, and (4) since older workers receive higher salaries and are more productive, if japan's workforce aged more than the US's, they should have grown faster. This implies that GDP per person might actually be a perfectly fine metric to look at. And by that metric, things haven't gone well in Japan. Or look at the increase in suicides -- again, not a pretty picture.

    Lastly, "Keynesians" like myself and Krugman actually believe that the best way out of the trap for Japan is simply to print money, so it's a bit strange you're arguing that we've misinterpreted Japan b/c we just always want to deficit spend when we're recommending seignorage.

  23. honestly, i dont see us picking up 23% workforce from 1990-2007 in germany. our population is also aging and shrinking - at least according to politicians and news channels etc. ;) ... must be that the number includes picking up full eastern germany economy ... which really has a significant impact and makes those numbers definitly useless for germany imo (simply because eastern germany had zero productivity when they joined) ... might explain the smallish 8.8% GDP growth per working age adult that is far off from whate every other country had.

    - asac

  24. This comment has been removed by the author.

  25. Your blogs are totally worth gift quantify and liveliness.
    unsecured loans online

  26. I touch really precise version these articles I stingy there are writers that can create moral stuff.
    quick low rate loans

  27. You guys make it really easy for all the folks out there.
    online lending

  28. I am sure you have a great fan following out there.
    online payday advance

  29. Always so interesting to visit your site.What a great info, thank you for sharing. this will help me so much in my learning.

  30. offers high page rank do follow instant approval blog sites list Instant Approval Blogs

  31. I am really impress with your article and youhave done good work on your blog i feel i have to do work like this if i want to succes my blog too. Low Calorie Recipes

  32. That’s a great site you folks have been carrying out power for homes

  33. After eight a few several weeks of preparing, I just adopted on the water. I adopted some advantage from your page actually. home improvement

  34. Hi, This blog is very useful and I got to learn certain facts regarding finance and business. I like to follow your post regularly to get more info like that in future. Thanks

    Japanese yen

  35. Info is out of this world, I would love to read

  36. The gorgeous post learned a great deal Thanks greatly!
    payday loans online no credit check

  37. Having such type of incredible blogs provoking the people to come and visit this site and comment. Thanks for all
    cash advance payday loan

  38. This blog is fully fabulous in all aspects.
    car accident claims

  39. Really nice and informative post. Thanks a lot for sharing with us.
    Horror movies

  40. Thanks for the post. Japan can surely get by. Unemployment there is not really an issue in my opinion but there are still few of them who are out of work. I always wondered what is income protection insurance and what is it about. It is only recently when economies started to collapse when I realized it is about being secured in case of unemployment.

  41. I want to appreciate your excellent work. Nodoubt your post is praiseworthy. This information is quite unique for me because I have very little information about this.
    Funny Pictures

  42. Stupendous blog you guys have provided there, I will absolutely valuate your effort.

  43. I’m definitely coming again to see these articles and reviews

  44. I have got the superb information from these blogs payday loan lenders only

  45. Thanks for posting this informative article thanks for sharing your great post,wish you have a nice day,happy every day.

    Commercial Solar Panels & Solar Energy


Google Analytics Alternative