Scott Sumner  

Japan adopts an NGDP target

Innumeracy Illustrated... Friday Night Video: Government...

Marcus Nunes points to a report that Japan has adopted an NGDP target:

Japanese Prime Minister Shinzo Abe vowed on Thursday to raise gross domestic product by nearly a quarter to 600 trillion Japanese yen ($5 trillion), pledging to refocus on the economy after the passage of controversial security bills that eroded his popularity.

Abe unveiled the plan at a news conference marking his election to a second three-year term as ruling Liberal Democratic Party leader and hence, premier.

Abe stopped short, however, of setting a timeframe for the new GDP target, which could raise doubts about the goal.

. . .

"The goal is to create the largest economy in the post-war era to help make people's livelihood the most affluent. I'll make GDP worth 600 trillion yen as our clear target."

(Actually the increase would be 20%, not 25%). There are two positive aspects to the announcement, and one very strong negative. The positive aspects obviously include the precedent of a major country setting a specific NGDP target. A second big positive is that the target is expressed as a level, not a growth rate.

But, as I'm sure you already noticed, none of this means much without setting an explicit date. Still I view this as a slight positive, despite the preceding flaw. In recent months there had been complaints from people in the Abe administration that the central bank policy was too expansionary. Some called for reducing the inflation target to 1%. If a 1% inflation target had been set by Abe in 2012 then that would have been defensible. Switching to NGDP targeting would also be defensible. But reducing a recently set 2% inflation target to 1% is a truly bizarre and silly idea. If you are going to do that, there's really no point in having any sort of inflation target. With the implied support provided by the new $600 trillion yen NGDP target, the head of the BOJ will be able to brush off any criticism of his QE program.

In the rest of his post Marcus has a nice discussion of why target rules are superior to instrument rules.

Mark Sadowski has a guest post at Marcus's blog, where he argues vigorously against the view that NGDP targets are less reliable than inflation targets:

Which means the claim you frequently hear that NGDP revisions are larger than inflation revisions is pure grade A horse manure. You will never see any evidence supporting this mindlessly repeated spurious claim, because no such evidence exists.

And, finally, inflation is a totally artificial construct requiring that we come up with an estimate of the extraordinary abstraction known as the "aggregate price level." To see how preposterous this is imagine equating the aggregate price level between what it is now and what it was in say 14th century England. The goods and services are so different it requires the complete suspension of one's disbelief.

I'm no expert on GDP revisions, although I hope others will look at this question and try to confirm Mark's conclusion. But I love the comment on inflation since the 14th century. Any estimate of the change in the "cost of living" (living how?) since 1300 would be almost meaningless. On the other hand an estimate of the change in NGDP since that period would have a clear meaning, even if undoubtedly there'd be some measurement error.

PS. Health care is now 18% of US GDP. What was the price of "health care" in 1300? How about "education"? Cars? Cell phones? "Houses"? Gym memberships? Your gas, electric and water bills?

Suits of armor?

Comments and Sharing

CATEGORIES: Monetary Policy

COMMENTS (10 to date)
E. Harding writes:

Gary North pointed out that while looking at the CPI since 1913 is pretty much worthless, looking at it since the Vietnam War is more defensible, as there's more goods to compare today that existed in 1967 than in 1913.

ThomasH writes:

The superiority of NGDPL trend target over an inflation rate is mainly that it is a level target. Secondarily it presumably preforms better to negative supply side shocks. Only as a third order of magnitude is the problem of index number construction. And setting a NGDPL trend target itself depends on a notion of what an optimal level of inflation is.

Airman Spry Shark writes:


Debatable. NGDP growth rate target could be superior even to price level target, simply because NGDP is well-defined (i.e., measurable) and P is not. Even if a price-level target is agreed upon, such a regime would spawn metadebates about the deviation from target (cf. PCE inflation vs. CPI inflation vs. GDP deflator; chained vs. unchained); NGDP, either rate or level, is subject only to measurement error, not definitional ambiguity.

Scott Sumner writes:

Thomas, I don't believe that you need to know the optimal rate of inflation in order to set an NGDP target, only the optimal rate of NGDP growth.

Meets writes:

Scott, do you think this is why US stock markets rallied Friday?

How does one know what part of Nominal GDP growth is Real GDP growth?

Zimbabwe is an extreme case. Its NGDP soared with runaway inflation, but its real output decreased as businesses failed.

What has to happen for RGDP to increase when NGDP does? What is the evidence that this policy works, and does it help everyone, or are there costs which fall unequally on the populace?

NGDPLT may be superior (I'm convinced it is), but pointing out that inflation cannot be compared across centuries is a spurious argument. As long as inflation is a decent measure across a period of 2-10 years, then inflation targeting cannot be truly critiqued on the basis that it breaks down across longer time periods. Any period longer than a decade is probably too long for nominal rigidities to be a big deal.

How many unemployed are still waiting to get paid as much as their forefathers did in 1353? How many people are waiting to sell their houses back to what they perceive of as the fair prices of 1345 + 20%?


(unrelated: how much do nominal rigidities in the housing market figure in your defense of NGDPLT? Perhaps because you have an American perspective, you are underestimating the effect of this factor in Southern Europe, for example).

bill writes:

Another nice feature of NGDPLT is that it forces the Fed to really try to boost the economy when RGDP growth falls. This results in countercyclical inflation. So NGDP ties the dual mandate to a single target.

Scott Sumner writes:

Meets, Probably not, the rally seemed to start in Europe.

Andrew, In the long run RGDP growth is determined by supply-side factors. That's why Zimbabwe did so poorly during their hyperinflation.

In the short run, NGDP and RGDP are correlated due to sticky wages and prices.

Luis, I agree that inflation is more accurate over shorter time periods. And yes, other arguments for NGDPLT are more important. But the "1300 to today" thought experiment really drives home the point that there is no "true" rate of inflation.

I'm afraid I don't know much about housing prices in southern Europe. What do you think causes them to be so sticky?

Bill, Exactly.

I think house prices are sticky for the same reasons that wages are sticky: nominal illusion. People have a very hard time selling below what they paid in nominal terms. If you are under water, you often literally cannot afford move away, but I think the typical case is that people refuse to take the nominal hit even at large personal cost. I know of someone who made an offer on a Lisbon house for 60% of the asking price a few years back. The person selling felt insulted and said so in no uncertain terms. My friend told me this story saying he had just checked up and the house was still for sale and the price has dropped (but not yet to 60% of the initial asking price, he laughed that when it did, he'd repeat his offer).

I have heard of stories from the other side too, people who refuse to sell their houses/apartments because they cannot get what they paid for it. I have heard of people refusing better jobs elsewhere or tolerating ridiculously long commutes because they refuse to move and "eat" the loss.

Even in the rental market, houses stay on the market very long before the asking price drops.


Housing prices do not figure in the EU's harmonized price index, for the simple reason that there was no political agreement on how to harmonize them. So, deflation is Southern Europe is under-estimated.


In Europe, housing is much more of an issue than the US (most big cities have a housing market like SF/NYC and a larger fraction of the population lives in these cities), so these issues impact quality of life much more (40-55% of young people still live with their parents in Southern Europe: some of it is culture, ie, choice, but the increase in the last few years would not have happened if prices had dropped as they should instead of this decade-long slow adjustment to a new reality; alternatively, if inflation was targeted to 4% or NGDPLT to 5% per annum, the adjustment would have been much faster).

Comments for this entry have been closed
Return to top