Modern economics is predicated on the idea that people are rational. Yet if you spend any time with traders and active market participants, it quickly becomes apparent that they're a superstitious lot. Good-luck charms, rally rituals and totemic ties are commonplace. When you think about it, doesn't technical analysis sometimes border on numerology, imbuing certain chart points with significance unjustified by any fundamental analysis? And traders seem compulsively drawn to conspiracy theories.

Many market participants have often suspected that governments, especially the Japanese government, manipulate economic data for political advantage. It was therefore with more than a little cynicism that we learned from

The New York Times

that Japanese officials purposely excluded data when calculating the October to December 1999


. That data, had it been included, would have made the contraction steeper than the 1.4% that was reported. Many of the sources in the story suggested the deception was for crass political gain, an understandable conclusion given that Japan has an important election coming up next month.

The established facts are clear. When the government estimated GDP, it deviated from its usual practice of replacing estimates of capital expenditures with survey data. The survey data found that capital investment among financial institutions was substantially less than the initial estimates suggested. Specifically, the survey data showed a decline of 37.7% compared with the initial estimate of a 3% dip.

An official from the

Economic Planning Agency

, which makes the GDP calculation, reportedly explained that the survey data were not used because they were not consistent with other economic reports. No explanation was given for the stark discrepancy. Subsequent data confirmed the weakness of the survey data. That means that rather than a 1.4% contraction, the Japanese economy contracted by 1.6%.

This is not significant. Calculating GDP to tenths of a percentage point is the way economists demonstrate they have a sense of humor, distinguishing them from their professional brethren, accountants.

Some have insinuated the numbers were fudged so that it would be easier to achieve the government's fiscal-year growth target of 0.6%. In order to achieve this, the Japanese economy needs to have expanded by 2.4% in the January to March quarter rather than 2%, which would have been required if the earlier capital expenditure estimates had proven accurate.

As it turns out, estimates for January to March GDP, which is due out in early June, range between 1.8% and 3.0%. A recent


survey found a median forecast of 2.1%. (Because Japan will not make adjustments for leap year, it's getting a little extra help.)

It stretches credibility to suggest that Japanese officials were so determined to reach their growth target that they manipulated the data by less than a quarter of a percent and even then cannot be assured that their target will be met. Remember, this is a country that surveys taxi-cab drivers and hairdressers to fine-tune data on personal consumption. What are we to think will happen if the growth target is not met? Although short-term speculators may disagree, it doesn't really change anything if growth for fiscal-year 1999/2000 comes in at 0.4% or 0.5%. The economy expanded, and by some reckoning, may have indeed bottomed in the first half of last year.

Nor can it be argued that missing the target will cause an economic or political backlash. With weak personal consumption and more than a whiff of deflation -- April consumer prices were 0.8% below year-ago levels -- conditions for an end of Japan's near zero interest rate policy are not yet ripe. A couple of tenths of a percentage point on an annualized growth figure is inconsequential for next month's lower house election. Prime Minister Yoshiro Mori's approval rating has plummeted. Yet, the most likely scenario is for a continuation of the status quo, whereby the LDP remains the largest party, but forms a governing coalition.

Let me conclude by making two points. First, recall that U.S. economic officials (like

Federal Reserve


Alan Greenspan

) have often warned of the deterioration of U.S. statistics due to a lack of proper funding. Outsourcing some economic reports, such as the

leading economic indicators series, does not appear to be possible for many reports.

Second, and more importantly, economic data are not objective. Values are imbedded in the very way economic activities are defined. Who ought to be considered unemployed, for example? How should people in the armed services be counted? What about their output or the output of prisoners? Should the savings rate include capital appreciation?

The ways these questions are answered are not immune for political considerations. Consider the monthly trade report. When

Ronald Reagan

was president, record merchandise trade deficits were embarrassing his administration. Sounds familiar? The solution: combine the chronic merchandise deficit with the persistent surplus on services. The result: a smaller monthly trade deficit. Europe and Japan continue to report merchandise trade separately from services.

In fact, I have proposed that Japan ought to copy the way the U.S. reports its monthly trade figures. Its merchandise trade surplus would be pared by its significant deficit on service trade. This might help reduce the political attention the report currently attracts. Eventually, I think the Japanese will adopt this proposal.

Of course, they might very well be accused of cooking the books.