Wednesday, January 11, 2012

Myths about Japan's Stagnation Revisited

Almost a year ago I posted two graphs and some comments on Real GDP Per Capita and Myths about Japan's Stagnation. I had realized that demographic factors and mild deflation played a different role than the conventional wisdom which held that they are each merely a factor in Japan's supposed "malaise". It turned out that demographics and the difference in real versus nominal growth entirely account for the illusion that Japan has suffered two lost decades. That is, real GDP per capita growth for the US and Japan since 1980 are remarkably similar, with one [longish] period of meaningful divergence. Here are the same graphs repeated from my post last year:

Annual Growth of Real GDP Per Capita in the US and Japan (1980-2009)

Annual Level (Indexed) of Real GDP Per Capita in the US and Japan (1980-2009)

Recently there has been some discussion among higher profile bloggers on this topic:
Noahpinion may have read my previous post because he or she used my first graph above without changing the file name I had chosen.

I won't discuss all these posts in detail and there are points in each I disagree with.

However, Paul Krugman's second post contains a graph of Japan's Log Real GDP per working-age resident. I had myself been interested in the per-working-age-resident data as compared to the per-capita data, since changes in the participation rate (partly due to demographics) could cause the two to differ, but hadn't gone to the trouble of getting the data. Unfortunately Paul Krugman starts at 1990 rather than 1980 and so I'm not sure his conclusion tells the full story:
"This picture suggests that the Japanese economy was indeed depressed for about 16 years, and deeply so after the slump of the late 1990s. But it may have returned to more or less potential output on the eve of the current crisis."
Look at my second graph (above) and you can see the extraordinary growth in the late 1980s in Japan. Was this growth "above potential"? Was Japan somehow borrowing from the future in the 1980s, and are the Austrians correct that slower growth in the 1990s to undo the "excesses" and get back to the trend line must be an inevitable outcome? (I do think the answer is closer to "no" than "yes", but don't have all the answers for how/why).

Is the correct framing Krugman's, i.e., that Japan stagnated from 1990 onward and didn't return to the potential growth trend line for a decade and a half? Or is the more accurate framing that Japan's growth accelerated above trend line in the 1980s and took a decade to revert to the trend line? (In a mostly smooth fashion, excepting the unfortunate 1997-ish austerity, rather than a crash and depression!)

I believe MMT shows how the "hangover" theory of economic growth is wrong (as opposed to asset prices, which MMTers generally agree must be allowed to adjust after bubbles pop). This is because it is always possible for the flow of national income to be sustained by government deficits or net exports even if the leakage to private sector savings increases. Perhaps there was a massive inventory effect from overbuilding of real estate in the 1980s, requiring less building in the 1990s. Should that or other dynamics have pushed unemployment so low as to cause accelerating inflation? Glancing at some graphs it appears Japan's late 1980s unemployment got down to around 2%, with inflation rising to around 4%. What would have happened if the economy hadn't slowed after 1990? Would inflation have accelerated upwards uncontrollably? It doesn't seem obvious that it would, but I don't know the answers.

If not an "overbuilding" dynamic or a labor force participation dynamic (not yet investigated), then perhaps the majority of the late 1980s surge can be explained by a huge above-trend rise in productivity. If so, is there any implication that productivity will inevitably grow below trend after such a surge? It doesn't seem like such a reversion should be inevitable, but perhaps there are dynamics specific to the types of productivity improvements in Japan in that time frame that would provide more answers.

Comments and insight are welcome.

I'll split a few more observations on demographics into a separate post to keep this from getting too long.


  1. Hbl,

    What do you think of the following?

    I think the phrase "lost decade" is quite right, only so long as it is interpreted to mean that Japan lost x years or 1-2 decade(s) or whatever because it lost that in reaching full employment.

    Keynes once said - wonderfully as always -

    "The difficulty lies, not in the new ideas, but in escaping from the old ones, which ramify, for those brought up as most of us have been, into every corner of our minds."

    Keynes of course was trying to confess that he had been wrong on many things, but what a brilliant way to say it !

    The reason I brought this up is that while I think the idea of "lost decade" is correct, the way most of us interpret it is completely wrong! (that's including me of course till very recently). We usually interpret it as the case that Japan has lost the race and is far behind other advanced nations. This is of course also the case of other confused commentators in the media over the years who have the same wrong thought in their mind and the myth has just spread.

    Do you know what I mean?

  2. Hi Ramanan,

    Good points -- yes I think I agree with you. In some sense there is a "lost decade" given that unemployment in Japan did rise after 1990 (rising from 2% to a bit under 6%, very roughly). So "full employment" was not maintained, but on the other hand in comparison to other advanced nations, Japan didn't really do any worse than "par"...

  3. I was curious about the per worker data as well. You can see my chart here.

  4. Thanks JP Koning. It does seem to follow about the same pattern as the per capita data, but is useful to compare, especially with other countries too.