Shinzo Abe didn’t fool any of the people any of the time.
Japan is now in a recession by the standard definition of two quarters of declining GDP in a row. The timing leaves little doubt that this recession was caused by the 3% increase in the consumption tax this April fools day. Japan’s GDP declined at an annual rate of 7.3 percent in the second quarter and at an annual rate of 1.6 percent in the third quarter. The 2014q2 decline is similar to the decline in US GDP in the 4th quarter of 2008 — a falling off the cliff decline.
This corresponds to a really impressively large multiplier. Consumption was only 61% of GDP in thrifty Japan so the increase in tax receipts should have been about 1.8 % of GDP yet GDP declined 1.8%. This isn’t normal. for one think tax changes normally have a smaller multiplier than spending changes. For another, GDP changes slowly with momentum (note the further small decrease in the third quarter).
Abe (and central bank governor H Kuroda) have been making trouble for me. They seem to have called the expected inflation imp from the vasty deep, and I said it couldn’t be done. Now they are forcing me to consider forward looking intertemporal optimizing consumers.
The long announced consumption tax increase implied that consumers faced a quite unusual real interest rate. Prices increased 2.1% in April (to me the surprising thing is that this is so close to 3 %). Price increases on that order were predictable and predicted. That corresponds to a real annual interest rate of oh about -24% (since nominal interest rates are almost exactly zero). Japanese consumers had a huge incentive to buy anything which didn’t rot in March.
Some goods are quite durable (OK my car is not at the moment quite as durable as it was Friday but I think it just needs a new battery). It is possible that well before March, Japanese demand was artificially high because of the threatened tax increase. There was an impressive 1.5 % increase in the first quarter of 2014.
So I think the Abe experiment provides evidence of forward looking consumption savings choices. Now Japanese consumers didn’t have to look far forward and everything was announced in advance. Still I am a bit impressed. I think such similarity between standard theory and data is very unusual.
The other interesting result of the experiment is the 2.1% consumer price increase. Now this is less than 3/105 (the old tax rate was 5%). But it is oddly close. In standard theory, producers consider future costs when setting sticky prices — and indirect taxes like the consumption tax are part of costs. Basically it seems that merchants did not smooth the prices they charged consumers when facing a very predictable indirect tax increase. In theory, and it seems also in Japan, this should cause extremely high demand followed by extremely low demand. That should be somewhat inefficient if the production function has short term increasing marginal costs.
I use the weasel word “somewhat,” since a fairly large spike is sales serves the joint interest of consumers and merchants to send Abe as few Yen as they can. Still I think it is hard to reconcile the huge decline with foresight, profit maximization, and standard assumptions about production functions.
Again 2.1 < 3/105. Consumer prices were rising before the tax increase. Before it was scheduled they were declining (there was deflation). The change was ascribed to monetary policy by just about everyone (that means including me, and I hated to so ascribe it). But smoothing prices charged consumers given the anticipated tax increase might have played a role. This should include merchants accepting slightly lower pre-tax prices immediately after the tax increase — in other words the missing fraction of a percent of consumer price increases might have been moved forward to May as well as pulled back to March.
An anticipated sharp decline in demand should have lead to a very low ratio of manufacturing inventories to shipments. But in 2014q4 the index was 1.0265 (it is set to average 1 in 2010 and is usually about 1 except during recessions). It then jumped up to 1.07685. This is the sort of thing which usually happens during recessions (which are usually not predictable). I’d say that it almost looks as if Japanese manufacturers were not as forward looking as their consumers.