tag:blogger.com,1999:blog-17232051.post7642746999019156608..comments2024-03-18T22:32:52.802-04:00Comments on Noahpinion: The Koizumi years: A macroeconomic puzzleNoah Smithhttp://www.blogger.com/profile/09093917601641588575noreply@blogger.comBlogger52125tag:blogger.com,1999:blog-17232051.post-46424913665593182232013-03-09T22:19:57.001-05:002013-03-09T22:19:57.001-05:00The raw fact is the change of a ratio and you don&...The raw fact is the change of a ratio and you don't discuss US growth at all. What if the question were why was US growth slower post 2000 ? This would not be hard to answer (to put it mildly). I'd put up a graph of Japanese GDP per working age adult. I'm sure it would show a similar puzzle (I haven't looked at one but you have).<br /><br />My guess about the explanation is roughly the plucked string. It is hard to keep a good economy down (or flat). Roberthttps://www.blogger.com/profile/14455788499385673507noreply@blogger.comtag:blogger.com,1999:blog-17232051.post-38202514330230094242013-02-14T00:44:36.602-05:002013-02-14T00:44:36.602-05:00I also believe there is no mystery in the Koizumi ...I also believe there is no mystery in the Koizumi instigated revival. In aggregate, the Koizumi years were characterized by structural reform, as Koizumi's mantra was "no growth without reform". <br /><br />The observation that there was no large currency depreciation is incorrect. While JPY continued to appreciate against USD, by over 10% during the period, the real effective exchange rate as measured by the BoJ depreciated some 40%, to a level not seen since the 1960s. Thus Japan's exports became significantly more competitive.<br /><br />Secondly, the observation that there was no significant increase in inflationary expectations is also off the mark.The BoJ was in the midst of quantitative easing essentially during the entire Koizumi years, expanding bank current account balances from JPY5 trillion in March 2001 to JPY35 trillion by March 2006. This QE was never about stimulating growth, but at keeping the banking system afloat as the Koizumi Administration forced banks to write off/write down their bad loans.The banks parked their ballooning piles of excess cash (from both massive current accounts and deposit growth versus shrinking loan demand) in JGBs, depressing yields. When it became clear after Q2 2003 that the worst of the bank balance sheet crisis was over,the BoJ was confident enough in the viability of the banking system to abandon QE, and collateralized overnight call rates immediately began discounting positive instead of negative inflation.<br /><br />Finally, very few highlight the role the Koizumi reforms had in reversing negative consumer and business sentiment. In addition to removing balance sheet risk uncertainty in the banking sector, the introduction of a REIT market where foreign investors provided massive amounts of non-recourse loan capital fostered a mini-property boom. In addition, promotion of venture capital/entrepreneurial activity, fostered a mini boom in emerging company listings,later vilified by the Livedoor debacle, which came to a head in January 1996 and killed the small cap boom. <br /><br />A close look at the components of nominal GDP growth during the period shows private housing investment and private capital expenditures as showing a significant positive contribution to GDP growth,aided to a lesser degree by net exports and household consumption.In other words, households and businesses felt confident enough to make longer-term commitments.This despite nominal GDP showing basically flat growth overall because of the significant shrinkage in government fixed capital formation (public works expenditures). <br /><br />After Koizumi, the reform initiatives fizzled and began reversing, the real effective exchange rate jumped sharply, capital investment and housing investment fizzled, and nominal GDP began to fall again, despite now-positive growth in government expenditures. www.JapanInvestor.comhttps://www.blogger.com/profile/18234844156740622561noreply@blogger.comtag:blogger.com,1999:blog-17232051.post-14182675099587887532013-02-13T15:16:56.327-05:002013-02-13T15:16:56.327-05:00Yes, I am suggesting that a significant piece of t...Yes, I am suggesting that a significant piece of the relative performance could be explained by productivity. It doesn't take much to increase productivity from a depressed underutilized starting point. <br /><br />This sounds extreme in your example (can we please modify it so it isn't me that dies); but how about an example of a firm with 100 employees that can produce 100 widgets but has only been producing 95, now one person retires and the following year the firm receives orders for 97 widgets. It doesn't seem so implausible to me that the firm would not necessarily have to hire another person.<br /><br />Danhttps://www.blogger.com/profile/15136541075745913165noreply@blogger.comtag:blogger.com,1999:blog-17232051.post-8748679111505372872013-02-13T08:29:03.043-05:002013-02-13T08:29:03.043-05:00"Sure, but birth rate changes take a couple d..."Sure, but birth rate changes take a couple decades to show up in the working-age population numbers! ;-)"<br />... well, the crisis started in Japan back in 1989 ... moreover migration can react much quicker - thats I guess what explains to a large extent the US working age population numbers ...<br />Mario Holznerhttp://www.wiiw.ac.at/?action=staff&opt=details&id=6noreply@blogger.comtag:blogger.com,1999:blog-17232051.post-47428542386766319192013-02-13T08:25:34.519-05:002013-02-13T08:25:34.519-05:00i'l send you my calculations by emaili'l send you my calculations by emailMario Holznerhttp://www.wiiw.ac.at/?action=staff&opt=details&id=6noreply@blogger.comtag:blogger.com,1999:blog-17232051.post-52756722323379021342013-02-12T13:31:45.433-05:002013-02-12T13:31:45.433-05:00if you compare real GDP PPP per working age popula...<i>if you compare real GDP PPP per working age population, the euro area did by far best throughout the 2000s</i><br /><br />Let's see some #s...Noah Smithhttps://www.blogger.com/profile/09093917601641588575noreply@blogger.comtag:blogger.com,1999:blog-17232051.post-78502737943844546272013-02-12T13:30:54.728-05:002013-02-12T13:30:54.728-05:00that depends inter alia also on demographics ... b...<i>that depends inter alia also on demographics ... both of which are not exogenous to GDP development and fiscal policy</i><br /><br />Sure, but birth rate changes take a couple decades to show up in the working-age population numbers! ;-)<br /><br />Especially in Japan, which has always been a place where most teenagers do not work.<br /><br /><i>and on the distribution of productivity among working age population</i><br /><br />Ah, but this strengthens the puzzle. As the working-age population shrank, it also aged. Older people are less productive.Noah Smithhttps://www.blogger.com/profile/09093917601641588575noreply@blogger.comtag:blogger.com,1999:blog-17232051.post-84081271622317628542013-02-12T12:43:22.207-05:002013-02-12T12:43:22.207-05:00ad 1) that depends inter alia also on demographics...ad 1) that depends inter alia also on demographics ... both of which are not exogenous to GDP development and fiscal policy ... and on the distribution of productivity among working age population ...<br />ad 2) on the contrary ...Mario Holznerhttp://www.wiiw.ac.at/?action=staff&opt=details&id=6noreply@blogger.comtag:blogger.com,1999:blog-17232051.post-14742173959697837972013-02-12T12:35:06.017-05:002013-02-12T12:35:06.017-05:00"But the comparison will be the same if you c..."But the comparison will be the same if you compare Japan to Europe."<br />... interestingly enough not - but that again depends on which indicator you look at: if you compare real GDP PPP per working age population, the euro area did by far best throughout the 2000s and if you look at real GDP PPP per employed it was basically a catastrophy for the Euro area ...<br />Mario Holznerhttp://www.wiiw.ac.at/?action=staff&opt=details&id=6noreply@blogger.comtag:blogger.com,1999:blog-17232051.post-79105156780965841352013-02-12T12:15:03.983-05:002013-02-12T12:15:03.983-05:00I don't think there's any mystery at all. ...I don't think there's any mystery at all. The standard monetary explanation fits to a t. There's a nominal shock in early nineties, which coupled with tight money and sticky wages and nominal debt effects lead to rgdp slow down. All the fiscal intervention does nothing to overcome the tight monetary policy, leading to a lot of debt and nothing else. Eventually, the relative prices/wages adjust to the new ngdp path and rdgp growth resumes. In other words, the nominal shock had been processed by the price system. Nothing to explain. Money is neutral in longer term. Anonymoushttps://www.blogger.com/profile/00758784432507739207noreply@blogger.comtag:blogger.com,1999:blog-17232051.post-80374928105213898902013-02-12T11:44:01.519-05:002013-02-12T11:44:01.519-05:00real GDP PPP per employed
Think about what happen...<i>real GDP PPP per employed</i><br /><br />Think about what happens to that indicator when unemployment falls... ;-)<br /><br /><i> finally, if we compare overall GDP PPP</i><br /><br />i.e., ignore demographics? I refer you to Mr. Krugman.Noah Smithhttps://www.blogger.com/profile/09093917601641588575noreply@blogger.comtag:blogger.com,1999:blog-17232051.post-61596813697908163482013-02-12T11:41:54.744-05:002013-02-12T11:41:54.744-05:00It seems as likely that you are measuring an incre...<i>It seems as likely that you are measuring an increase in the rate of relative population decline</i><br /><br />Not really, no...because Krugman's numbers mean that the remaining working-age population raised its output quite a lot.<br /><br />Imagine if you and I both earn $50,000. GDP per working-age population is $50,000. Now suppose you die. Unless my income increases substantially, GDP per working-age population will remain at $50,000.<br /><br /><i>or a change in US productivity</i><br /><br />But the comparison will be the same if you compare Japan to Europe.Noah Smithhttps://www.blogger.com/profile/09093917601641588575noreply@blogger.comtag:blogger.com,1999:blog-17232051.post-801717352992900062013-02-12T11:34:00.187-05:002013-02-12T11:34:00.187-05:00... exactly - what is being measured here is the q...... exactly - what is being measured here is the question - my last posting was probably somewhat unclear ... I have now compared the following three indicators and depending which one you chose you can tell at least three different stories: a) real GDP PPP per capita - no major difference in growth between the US & Japan in the 2000s, b) real GDP PPP per working age population - Japan is growing faster than the US in the 2000s, c) real GDP PPP per employed - US was growing faster in the 2000s than Japan ... finally, if we compare overall GDP PPP we find the US having strong growth and 175% of the 1988 level in 2011, while Japanese growth was anaemic and in 2011 only at 133% of the 1988 level ... in these comparisons you indirectly always say something about productivity, the demographic and labour market developments in the two countries ... Japan has a stagnating and aging population with falling employment but traditionally low levels of unemployment, while the US has an ever increasing population but cares less about unemployment figures - it is crucial to note that productivity seems to be increasing faster in the US if measured by GDP per employed ... as a result the overall US economy is developing better than the Japanese economy ... one obviously comes at an other conclusion if one looks at GDP per working age population where I would however claim that this rather shows the Japanese love for low unemployment rates ... still Japan's zombie firms effect?Mario Holznerhttp://www.wiiw.ac.at/?action=staff&opt=details&id=6noreply@blogger.comtag:blogger.com,1999:blog-17232051.post-51705419839439650372013-02-12T08:30:12.279-05:002013-02-12T08:30:12.279-05:00I still have no idea what is being measured here.
...I still have no idea what is being measured here.<br />It seems as likely that you are measuring an increase in the rate of relative population decline, or a change in US productivity as you are anything else. Japan's working age population decline has been accelerating, using the same assumptions used in calculating the productivity numbers would be a clue how much of this is at play. <br /><br />Its also not so clear to me that there was Austerity for the full period, tax revenues quite possibly declined more than government spending though its hard to tell. The intra period rate of change of government borrowing happened at the same point as credit expansion turned positive for the banks, the numbers are percentages so you can't compare relative impact on GDP.<br /><br />I would guess that Japan's private sector began spending more either in slowing rate of paying down debt for corporates or in increasing borrowing for households in response to the boom in Global GDP and this was slightly offset for part of the period, though probably not the whole period by marginally less government spending. This would be consistent with manufacturing productivity rising and the trade story.Danhttps://www.blogger.com/profile/15136541075745913165noreply@blogger.comtag:blogger.com,1999:blog-17232051.post-53177549371129955832013-02-12T04:42:23.555-05:002013-02-12T04:42:23.555-05:00Dear Noah,
thank you (and also Matt) for this pos...Dear Noah,<br /><br />thank you (and also Matt) for this post which I reposted with some remarks under<br />http://snbchf.com/2013/02/liquidity-trap-proble/<br /><br />For me you are the leader in thoughts on Japan. See here my ideas<br />1) Why the currency war is not the main driver behind the strong yen http://snbchf.com/2013/02/why-the-currency-war-is-not-the-main-driver-of-the-weak-yen-or-what-determines-fx-rates/<br /><br />2) An overview on economic opinion leaders from Kyle Bass oder Adam Posen to Paul Krugman and Richard Koo based on Poscar/Mc Culley terminology.<br />http://snbchf.com/2013/02/overview-japanese-currency-debasement-and-the-way-to-private-leveraging/<br />and some more...<br />George Dorganhttp://snbchf.comnoreply@blogger.comtag:blogger.com,1999:blog-17232051.post-84649914956956748242013-02-11T16:58:06.790-05:002013-02-11T16:58:06.790-05:00Yup, the trade data is not consistent. Think the ...Yup, the trade data is not consistent. Think the basket weights and deflators are different. Also Japan has fiscal year end in March, which may also have an effect on measurement. I don't know Japanese, so I don't know how to reconcile the various metrics. As they say, it's all Japanese to me...<br /><br />REER is typically a trade weighted basket, adjusted for inflation differentials.<br />http://en.wikipedia.org/wiki/Trade_weighted_index<br /><br />If memory serves, the BIS or World bank once found the that average half-life for currencies to mean revert to its REER is ~7 years.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-17232051.post-78110605055269353672013-02-11T16:19:07.813-05:002013-02-11T16:19:07.813-05:00Hmm...this chart shows Japan experiencing a trade ...Hmm...this chart shows Japan experiencing a trade deficit for one quarter in '03...really??<br /><br />I had thought that net exports boomed in the early 2000s, but every data source I could find said otherwise...<br /><br />That "real effective exchange rate" data is interesting...is that trade-weighted?Noah Smithhttps://www.blogger.com/profile/09093917601641588575noreply@blogger.comtag:blogger.com,1999:blog-17232051.post-74299098375544621192013-02-11T14:34:31.785-05:002013-02-11T14:34:31.785-05:00Well, the Yen cheapened vs PPP, right? So even th...Well, the Yen cheapened vs PPP, right? So even though nominally it was range bound, on a real basis it weakened. <br /><br />Re: Real Net exports, I am using data from Bloomberg, which is sourced from the Economic and Social Research Institute of Japan. I can't upload the bbg data, but it is reproduced here:<br />http://ycharts.com/indicators/japan_real_net_exports_of_goods_and_services<br /><br />There are obviously a lot of issues regarding the proper measurement of net trade. Imports of iPads into the US was cited as a recent example. So I'll allow that this issue probably can't be settled fully at this time. But even if we agree that net exports did not change much, the surge in the current account surplus was supportive, as you noted. <br /><br />As you noted, there is a lot of data that suggests not much changed in Japan domestically during the time frame, even as global growth surged. At the same time, Japan's Real Effective Exchange Rate declined to the lowest level in recent memory in 2007. (Page 2 here:<br />http://www.alliancebernstein.com/CmsObjectABD/PDF/EconomicPerspectives/EPJP_110923_GB.pdf)<br /><br />For an open economy like Japan, this is a significant event. Even if you disagree with my other points, any explanation for growth during the 2000-7 period would be incomplete with addressing that, IMO. Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-17232051.post-17361387676482513632013-02-11T13:08:10.372-05:002013-02-11T13:08:10.372-05:00Skeptical of your data. Net exports didn't cha...Skeptical of your data. Net exports didn't change much over the period, and the yen/dollar exchange rate didn't either.Noah Smithhttps://www.blogger.com/profile/09093917601641588575noreply@blogger.comtag:blogger.com,1999:blog-17232051.post-59674112904346555272013-02-11T12:48:00.103-05:002013-02-11T12:48:00.103-05:00If you look at the change in the composition of GD...If you look at the change in the composition of GDP over that period, the trend is clear. Net Exports as a % of GDP increased by 4pts, while Private consumption decreased by 1.8pts and Investment decreased by 2.7%. So my view is that the Yen carry trade, leading to a weaker Yen and higher net exports was a large driver of Japanese growth over that time frame. <br /><br />From 2007 - 2012, the biggest changes to GDP composition has been Private consumption (+2.5%) and Investment, (-2.5%) while Net exports was broadly flat. <br /><br />IMHO, I think the last 5 years are 'normal,' in that the trends in GDP composition will continue, given the demographic challenges. 'Abeconomics' will support growth somewhat, but with net exports already at 2.5% of GDP, and the aggressive monetary policy being discussed a 'one-shot' weapon, the Japan reflation story will probably have a shelf life of not more than two years. Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-17232051.post-8504505955791501232013-02-11T10:28:31.866-05:002013-02-11T10:28:31.866-05:00"Also, inflation expectations increased very ..."Also, inflation expectations increased very slightly, but stayed around 1%. So if quantitative easing helped, it did so through means that are not well-understood."<br /><br />Maybe QE helped by preventing inflation expectations from dropping further? It's like how the ARRA/stimulus was canceled out by 50 little Hoovers.Peterhttps://www.blogger.com/profile/08272747870634233567noreply@blogger.comtag:blogger.com,1999:blog-17232051.post-59505489650502806892013-02-11T10:15:59.261-05:002013-02-11T10:15:59.261-05:00I don't find your question intuitive at all.
...I don't find your question intuitive at all.<br /><br />Why would you not compare inflation to capacity?<br /><br />I don't see a compelling reason to think that inflation would be driven by a double ratio of (Japan output/capita)/(US output/capita).<br /><br />If you were looking at changes in prices why wouldn't the first place you look be demand compared to capacity? It's not hard to imagine a situation where capacity was flat, total output was falling, per capita output was rising and working population was declining.Danhttps://www.blogger.com/profile/15136541075745913165noreply@blogger.comtag:blogger.com,1999:blog-17232051.post-25856430981063966682013-02-11T06:02:36.793-05:002013-02-11T06:02:36.793-05:00Hi,
using WDI data and constructing indices starti...Hi,<br />using WDI data and constructing indices starting with 100 in 1988 I have tried to compare the development of US and Japan real GDP PPP per working age population as well as working age population itself. While the other three lines are more or less constantly upwardsloping quite similarly, the line of the Japanese working age population index is first stagnating and then declining starting in the early 2000s ...<br />Could this be the explanation?<br />Best regards,<br />mario holznerMario Holznerhttp://www.wiiw.ac.at/?action=staff&opt=details&id=6noreply@blogger.comtag:blogger.com,1999:blog-17232051.post-12321483892507814872013-02-11T01:09:17.680-05:002013-02-11T01:09:17.680-05:00Thanks for pointing that out! I thought you were d...Thanks for pointing that out! I thought you were dead btw...Noah Smithhttps://www.blogger.com/profile/09093917601641588575noreply@blogger.comtag:blogger.com,1999:blog-17232051.post-56022018400773433202013-02-11T00:03:07.403-05:002013-02-11T00:03:07.403-05:00The upswing in household debt would be the most li...The upswing in household debt would be the most likely candidate, but unfortunately the graph you posted has data on debt-to-income for the UK, not Japan. (See, for instance, <a href="http://www.thecommentator.com/article/667/what_s_holding_the_british_economy_back_debt_debt_and_more_debt" rel="nofollow">here</a>.) It turns out that the debt-to-income ratio in Japan was <a href="http://www.oecd.org/std/financialstatistics/" rel="nofollow">relatively steady over the expansion</a>, robbing me of one of my favorite explanations...<br /><br />Given all this, my view is that the Japanese experience is best interpreted through the lens of demand. Suppose that we have a crudely Keynesian take on the Japanese slump, in which the downturn was simply a shortfall of aggregate demand. Then the "mystery" of the Japanese expansion is not much of a mystery at all; after all, it wasn't an actual acceleration in GDP growth, just an acceleration in GDP growth <i>relative to labor supply</i>. And in a demand-constrained economy, perhaps there's no reason we should expect changes in supply to matter at all! If households smooth their consumption, then the retirement of elderly workers will leave demand (and therefore demand-constrained GDP) unaffected, while the working-age population decreases. This is roughly what we see in the data.<br /><br />All this is so crudely Keynesian that I'm a bit embarrassed to be writing it, but I do think that it can be embedded in a much more sophisticated model. Envision a riff on Krugman's 1998 BPEA suggestion that the natural rate of interest was low due to the saving needs of an aging population; as the population ages and we approach the new steady-state age distribution, the liquidity trap will naturally abate.Matt Rognlienoreply@blogger.com