At first glance, it seemed like delightful news for Japan: economic output growing at a 12.7 percent annual rate for the first three months of the year - the fastest growth in 23 years. Not bad after a four-year recession.
In fact the numbers mean economic growth for the fiscal year that ended March 31 was 2.3 percent, twice government targets.
But no chorus of joy arose from Tokyo economists.
The world second-largest economy, far from being on the fast track and pulling the world economy along in its wake, is still struggling. The stellar first-quarter growth, it turns out, is not only unsustainable but somewhat illusory, many economists say.
Much of the growth is attributable to a 27 percent increase in public spending that is now petering out, and to "front-loading" of capital investment by the semiconductor industry as the calendar year opened.
Moreover, the very fact that Japan's gross domestic product (GDP or national output) hit such a high official level makes it harder to achieve a statistical improvement in the next quarter.
"PKO in GDP" is how one respected private-sector economist sums up the Economic Planning Agency (EPA) data. PKO refers to "pricekeeping operations," (a play on the peacekeeping operations of the UN), in which the government orchestrates stock purchases through public pension funds to bolster share prices.
"Obviously, it is not sustainable," says Mineko Sasaki-Smith, an economist at Credit Suisse First Boston. "We did see strong growth, but 12.7 percent is so much more than all of us economists in Tokyo had forecast." He and other analysts note several additional distorting factors that make the EPA data unreliable:
*The EPA may have been too generous in its seasonal adjustment, by adding back into this year's actual figures the amount of the estimated loss in GDP in the same quarter last year caused by the devastating Kobe earthquake.
*The seasonal adjustment for a leap year is inadequate, analysts reckon. (The EPA does not explain how it adjusts its figures.) The issue may not seem like much, but an extra day in February extends the time for economic activity by 3.6 percent that month.
*The EPA has leeway, also undefined in public, to adjust private consumption data.
*The household spending survey is itself notoriously selective and unreliable. It covers only 8,000 households, which must constitute families of four. It takes no account of people living alone, couples with two incomes and no children, or single wage-earners living with working parents. The tedious research is compiled mainly by part-time workers who are paid $9.20 for each multipage questionnaire completed.
"It's rubbish in, rubbish out. Something here is anomalous, and I suggest it is the GDP numbers, not the rest of the world," says Chris Calderwood, an economist at the Tokyo office of Barclays de Zoete Wedd.
He blames the "feudal nature" of the Japanese statistical system, wherein the EPA merely compiles and distributes data that are grudgingly surrendered to it piecemeal by other ministries jealous of their own fiefdoms.
The sum of this information - the GDP data - loses credibility against smaller monthly data points. Compare, for example, the 5.1 percent year-on-year growth in private consumption claimed by the EPA's GDP report with other data showing a 2.5 percent rise in household spending, a 1.7 percent rise in passenger-car registrations, and a 1.8 percent decline in consumer-goods shipments for that period. Then last Thursday the Trade Ministry said Japanese firms increased production by 2 percent in May from the previous month, when the figure grew 3.2 percent. Economists had foreseen 4 percent growth.
The Japan Center for Economic Research, the research body that came closest to the EPA's figures with a forecast of 9.1 percent growth in the first quarter, predicts a 2.1 percent drop in GDP for the April-June quarter, and just 1.4 percent growth for the entire fiscal year.