Japan’s economic statistics are, by and large, rated highly for their diversity and accuracy. So it comes as no surprise that Japanese experts are helping developing countries improve their own statistical systems. Recently, however, that reputation seems to have been somewhat tarnished because of media speculation that the Economic Planning Agency might have juggled GDP figures as a result of political pressure.
In a preliminary GDP report for October-December 1999 released in May, the EPA gave its own estimate of financial-sector capital spending. Just several days later, however, it issued a correction replacing that estimate with an actual figure, which turned out to be far more negative. The media speculated that the initial May figure — the one calculated by EPA economists — might have been used to play down the steep drop in capital investment.
Confidence in the EPA also suffered in the spring of last year when there were rumors that key GDP figures might have been leaked in advance of the release of an official report. That time, however, the use of an estimated figure in the initial report, released one month before the general election, fueled speculation that the agency might have doctored an extremely negative actual figure to make it politically favorable to the government and the ruling parties.
The EPA offered a different story. During the compilation of GDP data, the agency learned that capital investment by financial institutions in the last quarter of 1999 had dropped by more than 30 percent from the previous three months. Given the unusually sharp decline, it decided to further analyze the figure. However, with little time left before the scheduled release of the report, the agency had no choice but to use the same internal estimate that had been announced a few months before. But the analysis revealed that capital spending had in fact plunged by more than 30 percent. So, a correction was made to update the estimate.
The clarification seems reasonable enough. However, as the government office in charge of the most important economic statistics, the EPA should have responded more quickly and more persuasively to the charges of “manipulation.” As a former EPA official pointed out, “The agency explained on a piecemeal basis, adding fuel to the speculation.”
In fairness to the EPA, it should be noted that GDP statistics are compiled by standard methods adopted by the United Nations. The OECD member states, including Japan, publish their GDP figures on a quarterly basis. There is very little — if any — room for juggling numbers. So it would be discourteous, to say the least, to question outright the integrity of those who engage in the compilation of GDP data.
It should also be noted that a GDP report is not the product of a purely mechanical process of crunching numbers. Raw numbers are adjusted at the discretion of experienced statisticians before final figures are worked out. The human factor plays an important part in measuring quarter-to-quarter changes in GDP growth. Mechanical work is not sufficient to accurately grasp changes in the industrial structure.
In the United States, where the system of basic statistics is not as sophisticated as that in Japan, bold estimates by statisticians are not uncommon. This is justified on condition of transparency and consistency of estimation methods. In Japan, GDP is calculated on the basis of 48 types of statistics, including surveys on family budgets and corporate finances.
Improving economic statistics in terms of both quantity and quality is a constant challenge. In the case of family budget surveys, for example, the number of people sampled at random is being increased. Improvement is needed particularly in the disclosure of data on government-financed public-works projects. This data is published once a year because it is compiled on an annual basis after the end of each fiscal year. So the figures on these programs in quarterly GDP reports are estimations. With these projects coming under increasingly critical public scrutiny, better disclosure by local governments is also an urgent necessity.
Another thing that needs improving — or correcting, to be more precise — is the loose manner in which the EPA director general and other economics ministers issue “personal projections” of GDP growth rates prior to formal announcements. Actually, those projections are “off-the-cuff” estimates. Given that Japan is the second-largest economy, politicians should know that their premature references to key numbers can have disruptive effects on world financial markets. Their insensitivity may be the biggest problem that must be dealt with.
In a time of both misinformation and too much information, quality journalism is more crucial than ever.
By subscribing, you can help us get the story right.