Why has deflation returned to Japan?

Japan Times reports:

Consumer prices excluding fresh food fell 0.1 percent in August from a year earlier, the first drop since April 2013, the same month Kuroda embarked on a campaign of record asset purchases to rid Japan of its “deflationary mindset.”

My goodness is economics a difficult subject.  (Scott Sumner is implicitly surprised too.)  So why is this happening?

Many of you might be tempted to utter some version of the words “liquidity trap.”  Even if this is one of the more reasonable versions of the liquidity trap arguments, there remains a problem.

Liquidity trap arguments imply that someone’s marginal utility of holding money is basically flat, whether that be the banks, the bank shareholders, the customers — someone.  And the flatness holds for a bunch of relevant someones, not just a few people.  (Or is that a flat marginal utility curve for “money plus safe short-term bonds“?  Whatever.)  With a flat marginal utility curve of money there are multiple equilibria, just as multiple equilibria more generally plague liquidity trap models.  Velocity could be something other than what it is, because at higher or lower levels of cash balance holdings the rate of return on those holdings still would be the same.

Institutional frictions may play a role in setting the equilibrium.  So why an equilibrium with falling prices?  Prices are sticky to some extent, which tends to militate against those of the multiple equilibria where prices are falling.  One might expect the equilibrium where the aggregate of prices rises, if only slowly.  But then why would a slightly higher rate of price inflation turn back down to a lower rate?

A second view is that the money supply/credit supply is endogenous, a’ la Fischer Black, Basil Moore, and others.  Until the real economy does better, the force of M times V will be weak.  This view involves no particular commitment to the slope of the marginal utility of money schedule.  Post 2013, prices went up for a while, because people thought Abenomics might work, but now that they see it doesn’t prices are sliding back down again.

Yet a third view is that the Japanese simply haven’t tried hard enough yet to debase their currency, see for instance Krugman on credibility or various Scott Sumner posts.  In this context I would myself cite gerontocracy rather than credibility issues.

My best guess is that some version of #3 makes #2 true at the relevant margin, but I don’t think such matters are well understood.

Addendum: Scott Sumner comments, for any plausible measurement I still say the rate of price inflation is relatively low in a puzzling manner, relative to asset purchases.  Large increases in money are in principle capable of offsetting relative small declines in food and energy prices, and if they do not that is simply another way of restating the puzzle.


It's not the inflation rate you want to follow but the nominal GDP growth path.

Nominal GDP for Japan in yen:
Since Shinzō Abe has been Prime Minister of Japan, you see positive growth in GDPn. Where's the liquidity trap?

Looks like its all due to energy.

Let's look at Japan's recent inflation history, or rather lack of it:

2014 2.74%
2013 0.4%
2012 0%
2011 -0.3%
2010 -0.7%

Why has deflation come back? It never went away. It just hid under the kotatsu for most of 2014 until concensus flatulence over how inflation would lead to disaster drove it out again. Japan is full of old men who are convinced that inflation will eat away at their own savings and middle-aged men who are convinced of the same thing and the young men are badly outnumbered and really have no choice but to agree. Meanwhile the women who are in charge of saving decisions have been raised to believe that virtue is socking yen away into their post office account and believe this virtue should be rewarded by a positive real return and probably aren't going to come down on the side of reflation either. While they know they were much better off in the pre '88 days of strong economic growth the dots between solid, consistent inflation and getting back to reasonable growth for a mature economy have definitely not been connected in the popular consciousness.

Just as how over the past generation Japan would raise interest rates when ever there was the slightest hint of a recovery and crush it, as soon as they scraped together almost three percentage points of inflation they gave up on maintaining it and let it be eliminated, presumably just by not doing anything to offset falls in commodity prices. Coal, oil, LPG, LNG, and iron ore are all down.

" Japan is full of old men who are convinced that inflation will eat away at their own savings and middle-aged men who are convinced of the same thing"

Are they wrong?

I think so. Currently Japanese people might get 0.2% return on their bank deposits. If deflation is half a percent that makes for a real return of 0.7%. If instead Japanese people could earn a real return of 2% or more as is often seen in economically healthier developed countries then they would be significantly better off. Then there's the issue that the Bank of Japan has let their interest rate fall to 0.1%. No sensible central bank would allow their interest rate fall that low as it leaves almost no room to respond to negative shocks with interest rate cuts. And being unable to respond quickly to economic shocks with cuts in interest rates can have significant downsides that can leave the entire nation worse off than it would otherwise be.

The argument that we need to have higher interest rates now so that the Central bank can lower them later if need be seems to be completely circular.

Jorgensen, higher interest rates for Japan now would be disastrous for the country. I strongly recommend the opposite. There's just a problem with getting them below zero. It's like my German geography teacher used to tell me, "You can't push on a piece of string. Unless you invade Russia, then the string will freeze in the winter. But it's probably best to find a simpler solution."

Japan is full of old men who are convinced that inflation will eat away at their own savings and middle-aged men who are convinced of the same thing and the young men are badly outnumbered and really have no choice but to agree.

You mean it's the demography, stupid.

Demographics is involved but Japan's median age is about equal to Germany's and Japan has has averaged low economic growth since 1992. Germany's economic growth has been considerably better while the two country's have had similar population growth rates, so the reason for Japan's poor performance is clearly not just demographics.

" Germany’s economic growth has been considerably better while the two country’s have had similar population growth rates"

Two countries reduced to ashes after a serious international argument both re-emerge under the occupation of the winner of the argument, although one is divided in half as a reward to a soon-to-fail ideology. One continues to dominate its immediate surroundings, the other flourishes, then stagnates. What accounts for this?

Well, German economic growth this century and a bit and has been: 2000 3%, 2001 1.7%, 2002 0%, 2003 -0.7%, 2004 1.2%, 2005 0.7%, 2006 3.7%, 2007 3.3%, 2008 1.1%, 2009 -5.6%, 2010 4.1%, 2011 3.6%, 2012 0.4%, 2013 0.1%, 2014 1.6%

That's an average of 1.2% over that time period. Significantly better than Japan but not that much better and pretty lousy in the last few years. So maybe Germany just hasn't been as efficient as Japan in slowing economic growth, but from now on they will be neck and neck in the race to have the lowest economic growth of a former wonder economy.

As for why, to me it seems pretty obvious. Japan didn't do what was necessary to keep their reserve bank interest rate from falling below 3%. Keep it at 3% or more and when there's a negative shock, bang, you lower that sucker for a quick boot to the economy to prevent or reduce a downturn and then you work on getting it back up to 3% or more again. But instead they decided to have their rate sit around 0%, which in this world is like going naked into a hedgehog throwing fight. And Germany, which used to have at least a couple of percentage points to cushion shocks has been at 0.05% for over a year now.

Jason's model for the price level (P) looks like this:

P(N,M) = alpha*k(N,M)*(M/M0)^(k(N,M)-1)

Where N is NGDP, M is currency in circulation (as opposed to the full amount of base money), alpha and M0 are constants, and k is the information transfer index, which goes as log N / log M

It's based on the information equilibrium concept described in the paper by Peter Fielitz, Guenter Borchardt:

"A general concept of natural information equilibrium: from the ideal gas law to the K-Trumpler effect"

funny, few ideas when there's an actual economic question that cannot be blamed on blacks, mexicans, muslims or gays.

hmmm, where are all the third rate economists now? average ain't over.

krugman really doesn't need his own blog. tyler will keep posting it for him.

tyler. what a childs name. wheres tyler? playing with garbage again. ohhh tyler!

Stoned at 9am, eh?


There seems to be a pattern: rising inequality, asset bubbles, financial crisis, aggressive monetary intervention, wage and economic stagnation. Probably only a coincidence.

Obligatory dismal science comment.

Also, does this mean that I can soon afford to take the family on a vacation to Tokyo?

If you have American dollars, then yes, the US dollar has already massively strengthened against the yen over the past couple of years. Time to book that trip. And no doubt Hawaii has been seeing fewer Japanese vistors as a result. But Australia, having seen its dollar fall with the price of its exports, will see an increase in Japanese tourism.

There's a nobel out there waiting to be claimed for the academic who can correctly pin down why advanced economies like the US and Japan aren't able to make their inflation target despite currency debasement. It will be on par with explaining the mechanism behind stagflation.

Run away accumulation of money in the hands of: the wealthy, corrupt Chinese bureaucrats, Russian oligarchs, third world dictators, sovereign wealth funds, tax exempt endowment funds, and American multi-nationals.

Japan buys oil on international markets. Oil is used in agriculture. Oil prices are falling. Therefore the oil component of food production costs are falling.

Is 0.1% unreasonable for all that?

Are economists staring at their own navels?

I think economic policy to hide gains from lower energy costs are pretty silly, shooting of feet.

Navels is a charitable way of putting it.

Yeah, the truth is they are examining their navels from the inside.

Point 4: The Japanese demographic change (more people getting older and now less people) is becoming so large that it is impossible to have inflation. (ie as a population gets over 50 their desire to save is higher than desire to spend.)

It is difficult to have inflation many places these days, but it is especially difficult to have "inflation" measured as prices while energy is in a great fall.

Collin, it is the easiest thing in the world to have inflation. Just like it's very easy to hit the ground at the base of a cliff. The only hard part is making yourself jump.

In a fractional reserve banking system it is important to know the impact of money supply on prices.

But just because something is important to know doesn't it is easy to know.

The best answer is the bundle of goods thing, and it works after a fashion, but it strikes me as one of the biggest stupidities in the economic profession to think that a move in a price bundle is inflation first, other things second.

Imagine that someone perfected a new technology. Call it "fracking." Imagine that this technology increased both immediate production and calculated reserves for oil world-wide. Imagine that this new production and those new reserves impacted prices. Imagine that the prices affected bundles of goods in every industrial country.

Should you try to extract that effect somehow? Or should you just play dumb and call it all "deflation", as if monetary supply made it happen, and not this "fracking?"

is that true? Shouldn't a rational person spend more money (on other people) as they get older?

Japan's gross saving rate as a percentage of GDP was 30% in 1984 and 22% in 2013.

When an individual ages they quit working and spend the money that they've saved while they were working so there's nothing abnormal or even enlightening about those figures.

Please go right ahead and do a better job of enlightening er and answering his or her question. (I'm afraid I'm not very good at this sort of thing and do enjoy it when a skilled individual steps in to save the day.)

For those who think that bundles-of-goods correctly measure inflation in the monetary sense, just be patient. Lower energy prices will change consumption patterns, the bundles-in-use will change, and "inflation" will return.

"For those who think that bundles-of-goods correctly measure inflation in the monetary sense"

How else would you measure inflation?

That's the rub, as I say in my longer comment. Bundles of goods are the method we have, but we should recognize their real limitations. "Bundles" are a moving target affected by far more than money supply. In the case of oil, it doesn't even take technological change. A misjudgment by oil producers about future consumption will make future oil prices higher or lower. That's not money supply either.

Basically, I think we should be slow to recognize CPI or PPI as "inflation" and recognize that it is certainly not a direct measure.

When the coastal valleys of southern California filled, rents and home prices started to rise strongly. Surely this lack of available land is "monetary inflation" as well.

" Surely this lack of available land is “monetary inflation” as well"

I think that goes to the core of why our discussion of inflation is so wrong. If prices of desirable pieces of land go up because population is rising and becoming wealthier then that should not count as inflation. That sort of increase because of scarcity does get counted towards inflation. On the other hand improvements in quality do not get taken into account. The result is that we are having ferocious debates over inflation falling between zero and two % when it is not defined or measured within an accuracy of plus or minus 1%-2%. (that is that all of our debate is within the error bars).

I'd start by not allowing substitution of "similar" items into the bundle. That alone makes inflation measurements into at best an art as opposed to a science. You'll never catch Cowen substituting his Harvard degree for one from Cornell. An Cornell's a lot more like Harvard than beef is like chicken.

You substitute anyway, as you adjust the bundle to consumer habits. People eat more chicken? People prefer chicken. That is the new normal, and inflation is the change in price of chicken.

Bundles of goods do not measure inflation in the monetary sense. They measure price inflation. Monetary inflation is measured by how much the Fed is "quantitatively easing."

Excluding food and oil the August Japanese CPI rose 0.8% according to the article you linked to.

It seems premature to jump to the conclusion that deflation has returned.

food and oil don't count because they are the biggest expenditures right? "outliers?"

Food does not count because it is so volatile.

Oil does not count because (1) it is volatile and (2) rise and fall in the price of oil which is driven by increasing or decreasing supply, rather than monetary factors, is not part of the phenomena that we are trying to encapsulate in "inflation" or "deflation"

TC - 1, Sumner - 0. What do you expect when a chess master plays a patzer?

As for Japan's deflation, it's masochism by Japanese seniors that's to blame. The more financial repression the more they like it. I kid you not, the Japanese love to save. That said, I saw a stat that showed Japanese savers have gone from China like 30% savings rate to single digits in the last generation, which I find hard to believe. But, if true, then it means Japan will implode since no more savers are left to finance the excessive JP gov't debt.

It remains consistent with my model that predicts essentially a flat price level (approximately zero inflation) for years to come:


Now with an update showing current core inflation and some more background:


Nobody ever replies to Jason S.

Is he a krank?

I'm not going to bother finding out if Jason is a krank or not as long as he posts blind links. If he can't be bothered to give a one or two sentence summary of his model then I'm certainly not going to bother to go through the massive effort of clicking on his link.

Jason's model for the price level (P) looks like this:

P(N,M) = alpha*k(N,M)*(M/M0)^(k(N,M)-1)

Where N is NGDP, M is currency in circulation (as opposed to the full amount of base money), alpha and M0 are constants, and k is the information transfer index, which goes as log N / log M

It's based on the information equilibrium concept described in the paper by Peter Fielitz, Guenter Borchardt:

"A general concept of natural information equilibrium: from the ideal gas law to the K-Trumpler effect"

A draft paper Jason put together detailing the application of the principle of information equilibrium to various subjects in economics, including the justification for, empirical analysis of, and derivation of the above price level model are explained in his draft paper on the subject here.

Here's a collection of predictions that follow from Jason's ITM framework:

It'd be interesting to compare the performance of those predictions against predictions from a different framework.

The velocity of money goes down as the interest rate goes down. I think this is the core issue for a "liquidity trap". If the velocity of money goes down then inflation is down. So far the Japanese have been able to keep pushing interest rates lower. I think any sort of shock, like oil prices going back up, could set the velocity of money going up, which could cause inflation, and cause the velocity of money to go up more etc.. So a shock at this point could set off a deadly feedback loop in Japan.

I have some graphs to show how velocity of money changes with interest rate:


A while back I posted a comment at econlog to Scott Sumner's remark about the level of deflation in Japan has actually been at 0% or slightly positive for years while at times slipping to very mild deflation (-0.5%). Sumner replied that he thought the GDP deflater was the better measurement. But now he is using core inflation, which makes sense to me.

Would it be that he doesn't have recent GDP deflater numbers so uses core inflation?

Japan accepted 11 refugees in 2014.

Now they are talking about accepting dozens.

If the deflation is demographic, this could help.

If central banks, through quantitative easing, can monetize national debt without consequences, then why not?

Previous investment has led to lots of worldwide production, probably even over-production, when the distribution of increased income from the increased production is so skewed towards the already rich financiers.

Real goods are getting better, and cheaper. But real wages aren't going up, so the vast number of middle income folk don't increase the quality/ price of the goods they buy as much as the increased production reduces prices.

A helicopter drop of cash would likely cause a bit of an increase. In Japan or in the US. I'd advocate a $1000 tax credit for all who pay taxes, and keep "dropping" that reduced tax burden on the taxpayers until inflation was higher.

I oppose fiscal stimulus of gov't spending more cash, but I DO support lower taxes / tax credits or refunds, so earners have more cash in pocket to spend / invest / deleverage (pay back debt, like cars, houses, credit).

> are in principle capable of offsetting relative small declines in food and energy prices

Umm, energy is -15.0% in the US and I'm guessing oil is a bigger part of Japan's energy mix than it is in the US. A year ago this week is literally when crude fell off a cliff.

Japan consumes just over half as much oil per capita as the United States does. However, unlike the US it imports all its coal, LPG, and LNG and the prices of those commodities have also fallen. Seabourne coal is half what it was at its peak in 2011. The price of iron ore and coke are also down, which is relevent since Japan still has a large steel industry. And the Japanese steel industry must be very glad that China has devalued its currency since it was not doing well.

Or: Monetarists continue to be befuddled by reality, continue to seek increasingly convoluted explanations that wouldn't require them to take off their monetarist glasses.

Of course energy prices have been deflationary as everywhere, but also QE is doing nothing to spending or even to the exchange rate since late 2014. The slight NGDP growth was not driven by stimulus to credit, it was driven partly by the devaluation boost to exports, and partly by the initial "Abenomics" fiscal expansion, neither of which are continuing.

Not just the marginal utility of holding money for someones is flat, the market consensus is that holding yen is more valuable than spending it despite expectations of indefinite QE continuation. In Japan's situation, the choice to QE or not is just choosing the instruments of the financialization of future incomes occurring through public deficits - deciding between issuing currency or issuing bonds. Focusing on the moneyness of currency vs bonds makes no sense here.

I recall Tyler Cowen hosting on this website an active exchange on a related subject back in April 2013, following Martin Wolf's "Japan's Unfinished Policy Revolution" in the FT. Grant Lewis from Daiwa subsequently had a meaningful rebuttal a few days later on FT as well. I can't recall if it was you Tom who surfaced the notion that if marginal utility of holding money was zero, then there would be multiple equilibria , which in itself is a messy thought. All very insightful intelligent exchanges.

The corporate savings in Japan, unlike Household/consumer savings, is a strange and sustaining phenomenon. In part it can be explained by the large mostly-well-funded corporate employee pension plans (imagine that!), as well as shareholder interests being under-served by corporate leaders and perhaps a lack of successful shareholder activists.

The other part of that argument is that Japan's QE and savings are not being deployed into productive fixed investments and that capital formation is not being multiplied out. I would conjecture that in any "normal" OECD country, if the corporate leaders weren't able to find those investments in their country.... they would invest abroad. Relative to other countries, perhaps that's not happening in Japan?

Of course, we saw Rockefeller Center at one point bought by Mitsubishi. But why aren't there today any Japanese real estate global players like Brookfield, Lend Lease, Blackstone, etc? Why is it that Japanese consumer banks aren't more dominant in their backyard of Hong Kong, Singapore, Australia or Taiwan? Why isn't there a host of Japanese companies that compete today with Nestle, Procter & Gamble, H&M, Starwood, etc? Blaming the lack of exported success on colony history perception, language, or culture seems inadequate. Its more about the lack of interest, motive, and risk-taking by corporate leaders in Japan. Afterall, if there was a will, there would be a way to hire foreign managers, have equity/JVs in a compatible structure, and leverage the synergies within their own organizations.

Whatever the reason, the Japan QE and savings could be very attractively put to use in Japanese companies... even if Japan's GDP was stagnant. Afterall, Japan doesn't live alone in a bubble.... or does it?

Comments for this post are closed