Believe those who are seeking the truth. Doubt those who find it. Andre Gide


Tuesday, September 17, 2013

What's up with Japan? (G, evidently)

There is a very interesting monetary policy experiment happening in Japan these days. The outcome of the project will surely be discussed in future macro textbooks. While we are waiting for events to play out, I thought it might be of some interest to provide some context in terms of Japanese GDP data since 1980.

The first diagram reports the behavior of expenditure shares. C is private consumption (including imports), G is public consumption (including imports), I is both private and public investment (including imports), X is exports, and M is imports. By definition, the GDP can be decomposed into its expenditure components as follows: Y = C + I + G + X - M.

Recall that the great slowdown in growth occurred in 1990-91. Here is the picture:



Since the great slowdown, (C/Y) increased from 53% to 60% and (G/Y) increased from 13% to 20%. That's one heck of a consumption boom!

That consumption boom has been financed by a dwindling expenditure share accruing to domestic investment. In 1990, (I/Y) was about 32%, today, it is about 21%.

The next diagram plots real GDP, with its components C, I and G all normalized to 100 in 1980.


We see the great boom early on in the sample, fueled by domestic investment spending. Over that period of time, both private and public consumption grew at essentially the same rate as income (GDP).

Since the time of the great slowdown, the trajectories of these expenditure components have diverged significantly (so much for the "balanced growth" assumptions we frequently embed in our theories!).

What really stands out in this data, to my eye at least, is how G and I appear to have gone their separate ways.

It would be of interest to dig deeper into the data to find out what is going on. What is all that G being used for? Was it too low to begin with and is now just approaching its desired level? Is the increase in G crowding out investment I? Or are there other forces responsible for this pattern--and does the increase in G represent a desirable response to these other forces?

And, of course, the big question for monetary policy wonks: Is a massive asset-purchase program on the part of the Bank of Japan really what that economy needs? Or are policy interventions better directed elsewhere?

19 comments:

  1. Massive asset purchases are EXACTLY what the economy needs!

    With the Japanese economy stuck in on and off again deflation, and self-induced paralysis (much like the U.S.!) how can you even ask the question, David?

    ReplyDelete
    Replies
    1. It was easy to ask the question. :)

      So, you must believe that the modest deflation in Japan since their slowdown is related to the expansion in government consumption and the contraction in domestic investment?

      Delete
  2. Occam's razor, financial economist version: what are the frictions (probably imposed by the Japanese government) preventing businesses from investing? Japan has high tax rates, and the labor laws are quite constraining. There's also a lot of regulation & a slow court system that reduce foreign investment. Investors, domestic and international, would prefer other, freer countries that are still quite close geographically, and have better growth.

    ReplyDelete
    Replies
    1. Prof J: the problem with this explanation is that it relies on the (questionable?) proposition that all of these factors suddenly became worse in 1990. These same institutions that are frequently blamed for holding Japan back are likely the same institutions that facilitated its postwar growth miracle. If one wants to make a case that this is not so, then point to the changes in tax rates, labor laws, etc., that resulted in the great slowdown.

      Delete
    2. Right, I see your question now. Well, those factors have gotten worse over the years, but they certainly weren't sudden in 1990.

      According to Kindleberger & Aliber (2005), the Japanese asset bubble burst in 1990, with stock prices declining 60% over two years and not recovering until 2003. So I'm guessing the issue is in how the Japanese government responded to the bubble bursting.

      The evidence has it that the Japanese government followed the "too big to fail" advice that was followed by the U.S. gov't in 2007. This kept zombie banks alive for years. The bubble bursting also led other banks to curtail lending sharply as collateral value declined by orders of magnitude.

      Continued interference in markets hasn't helped at all, and I'm tremendously suspect of any suggestion that lower interest rates would have helped.

      Delete
  3. Japan has supply-side, structural problems of course, everyone does.

    And as a monetarist, I would like to see less of G, and more asset purchases.

    I see something amusing with the idea of defeatists who argue that Japan central bank cannot stop deflation, in essence they are arguing that the real balance effect is infinite.

    Think about it. Even IF Money printed and spent in LSAP's is hoarded as bank reserves and currency, The asset purchases of public and private debt still remove it from the financial books. No matter how much money banks receive, they'll hoard it? And gain a real return on rigid prices?

    ReplyDelete
  4. "What really stands out in this data, to my eye at least, is how G and I appear to have gone their separate ways."

    It depends on what you mean by G.

    Japan uses the System of National Accounts (SNA) so government investment is included in I.

    If you look at the sum of government final consumption and public sector gross investment (what we call GCE in the US) you'll notice there was one period during since the "great slowdown" that GCE as a percent of nominal GDP (NGDP) significantly fell. This was from 25.0% of GDP in 2002Q1 to 22.3% of GDP in 2007Q2. In fact 22.3% is the smallest share of GDP that GCE has been on seasonally adjusted quarterly NGDP records back to 1994Q1. On unseasonally adjusted records one has to go back to 1992Q3.

    This happened primarily because public sector gross investment declined from 6.5% to 4.3% of nominal GDP. You can see from your graph that the decline in I as a proportion of GDP leveled off during this time. An examination of the components of I reveals that residential investment as a percent of GDP was also level during this time. And since inventories are volatile let's exclude them as well.

    What we have left is private non-residential fixed investment (PNRFI). This rose from 12.7% of GDP in 2002Q2 to 15.3% of GDP in 2008Q2, which is easily the largest boom in this category since the great slowdown. In fact 15.3% is the largest share of GDP that PRNFI has been on seasonally adjusted quarterly NGDP records back to 1994Q1. On unseasonally adjusted records one has to go back to 1993Q1.

    Your second graph seems to show an anomaly that coincides with this decline in GCE and increase in PNRFI. This is the surge in real GDP (RGDP) growth from 2002Q1 to 2008Q1. During this time RGDP grew at an average annual rate of 1.9%. This compares to an average annual rate of 0.9% from 1992Q1 through 2002Q1.

    This period of resurgent growth is often referred to as the "Koizumi Boom" in Japan. Between 2002Q1 and 2008Q1 real (adjusted by the GDP implicit price deflator) GCE grew at only a 0.2% average annual rate. In contrast real PFNRI grew at an average annual rate of 4.7%. Real private consumption grew at the same rate as RGDP.

    A closer examination of Japan's GDP reveals that both exports and imports surged during this time as well. Exports rose from 10.2% of nominal GDP in 2001Q4 to 19.3% of GDP in 2008Q3. Imports rose from 9.4% of GDP in 2001Q4 to 19.5% of GDP in 2008Q3. From 2002Q1 to 2008Q3 real (adjusted by the GDP implicit price deflator) grew at an average annual rate of 11.0%. Real imports grew at an average annual rate of 12.1%.

    "And, of course, the big question for monetary policy wonks: Is a massive asset-purchase program on the part of the Bank of Japan really what that economy needs?"

    Is there anything that can explain why Japanese RGDP (as well as real PNRFI, real exports and real imports) grew so fast despite the fact real GCE was stagnant during the Koizumi Boom? Well, there is the fact that Japan’s original ryōteki kin’yū kanwa (QE) was officially announced in March 2001 and concluded in March 2006.

    Just a coincidence?

    ReplyDelete
    Replies
    1. "From 2002Q1 to 2008Q3 real (adjusted by the GDP implicit price deflator) grew at an average annual rate of 11.0%. Real imports grew at an average annual rate of 12.1%."

      should read

      "From 2002Q1 to 2008Q1 real (adjusted by the GDP implicit price deflator) exports grew at an average annual rate of 11.0%. Real imports grew at an average annual rate of 12.1%."

      Delete
    2. Mark: Wow, you know Japanese data very well. Not many non-Japanese (or even Japanese) do.

      FWIW, I wrote the following post on this matter a while back.

      http://himaginary.blogspot.jp/2009/05/paul-krugmans-elementary-mistakes-on.html

      Delete
    3. himaginary,
      I actually read your post when I was researching for the following two posts:

      http://thefaintofheart.wordpress.com/2013/06/10/richard-koo-also-misinterprets-japans-lost-decades-a-guest-post-by-mark-sadowski/

      http://thefaintofheart.wordpress.com/2013/06/11/richard-koo-also-misinterprets-japans-lost-decades-part-ii-a-guest-post-by-mark-sadowski/

      In fact you inspired a section in the second post (my take on the very same issue).

      I've spent hours and hours learning the Japanese national accounts and flow of funds. So yes, I'd like to think I'm well versed in their idiosyncrasies.

      Delete
    4. Mark: I am out of town right now and will decompose those expenditure components when I return.

      But in the meantime, I have no idea what point you are trying to make. I measure G as government consumption. I include government investment in I. The point is that G is rising, C is rising and I is falling.

      I don't see how anything you say gets around this basic fact.

      Delete
    5. David,
      You said, "It would be of interest to dig deeper into the data to find out what is going on." But you suggested the story might lie in disaggregating government final consumption, when it really lies in disaggregating investment.

      You concluded your post by asking if a large scale asset program was what the Japanese economy really needs. Public investment is fundamentally different from private investment, in that it is not a perfect substitute for private investment, and in that it is largely exogenous to monetary policy. In my opinion the only way one could possibly answer your concluding question is by disaggregating investment.

      Delete
    6. Mark:

      The facts are that C/Y and G/Y (private and government consumption) are rising. This is independent of how one breaks down investment.

      I still have no idea how monetary policy is supposed to affect this secular trend.

      Delete
    7. David,
      Private and government consumption were 57.74% and 18.49% of nominal GDP respectively in 2002Q1. Private and government consumption were 57.79% and 18.26% of nominal GDP respectively in 2008Q1. The Koizumi Boom is the only period since the great slowdown when neither form of consumption increased significantly as a percent of GDP.

      Delete
  5. Replies
    1. Dave, the shares are nominal.

      I'm very interested to know where you want to go with this question. I know for the US at least, that C/Y real and C/Y nominal look very different, indicating a secular relative price change.

      How would you want to see the data above reported and why?

      Delete
  6. In government consumption, "Social benefits in kind etc" and "Consumption of fixed capital" are increasing. The former is increasing because people are getting older, and the latter is increasing because infrastructure is getting older ;-).

    ReplyDelete
  7. "the big question for monetary policy wonks: Is a massive asset-purchase program on the part of the Bank of Japan really what that economy needs? Or are policy interventions better directed elsewhere?"

    The central bank should target broader money as it is more relevant to the broader economy than narrow money. Broader money rates are nowhere near the zero bound.

    ReplyDelete
  8. David, this sectoral accounting may give the impression that Japan has brought some balance to its economy. However, I'm going to guess the majority of the gains to the "consumption boom" is really just the increased cost of raw material imports, which Japan has long sourced internationally to manufacture its goods. The last decade of course saw the repricing of the planet, so to speak. And Japan's long and clever post-war arbitrage, in which it sourced cheap energy and other inputs, is now over. Indeed, Japan's current monthly energy imports cost 40% of the total value of its monthly exports.

    Here's a chart of that metric, using the latest data from the MOF:
    http://www.peakfish.com/japan-total-energy-import-costs-as-a-percent-of-total-export-value-2010-2013/

    Thanks for your post. It helpfully breaks down components, and strengthens my view there is little to nothing Japan can do now to escape its predicament.

    G

    ReplyDelete