The American people love to believe that the United States is unique. The Constitution is unique, not warmed-over John Locke. U.S. foreign policy is unique, not conventional big power self-interest. And above all, the real “shining city on a hill,” is the U.S. economy, which naturally has the fastest productivity growth in human history, greater than any other country.

At a seminar held by four leading U.S. economists of differing political persuasions last week, there were differences on the size of the projected tax cut, but unanimous agreement that the New Economy has put productivity growth on a permanently higher plateau, and that conventional views of stock valuation and the business cycle were thus outdated and worthless.

It is still their belief that the Internet and its outgrowths have raised the United States’ potential for economic growth in a secular fashion, forming a revolution equivalent to the coming of railroads, electric power or the automobile.

Is it not odd, then, that there are still only four substantial companies that have arisen from the Internet, Yahoo, E-Bay, Amazon.com and AOL, and that the four, taken as a whole, have yet to show a profit?

In the fourth quarter of 2000, the four companies together showed sales of $3.5 billion, and a net loss of $76 million — before deducting losses on various misguided dot-com investments they had made.

This is hardly the stuff of a new industrial revolution, and given the huge amounts of capital investment that have been devoted to Internet development, must make one seriously question whether the Internet has had a positive effect on the U.S. economy at all.

Maybe it is simply an “interesting new wholesaler with a nice sideline in pornography” as James J. Cramer of TheStreet.com once wrote, but if so, it is an interesting new wholesaler that has managed to absorb several hundred billion dollars of investors money.

Another odd thing about the surge in productivity growth is that it appears to have taken place entirely without any external help. Yes, Greenspan has run a lax monetary policy, but Burns and Miller did that in the 1970s, and it did nothing for productivity.

Companies have been downsizing, and de-layering middle management — but the real surge in that came in 1984-92, not productivity’s golden years.

Taxes have been cut, but very marginally, and they are considerably higher than a decade ago, before the 1990 and 1993 increases.

Government regulation has been on a substantial upward curve during the Clinton years.

The only real changes that could theoretically have caused a productivity surge have been a huge stock market boom, and an enormous increase in executive stock options. However, similar stock market booms in 1927-1929 and 1964-1968 produced no great surges in productivity, as far as we can tell.

As for stock options, believing that they are the cause of the gains requires one to believe two very unlikely things: (i) that top management is motivated at the most basic trained-canary level to ring the bell only when it gets a very large peanut, and (ii) that lower management and shop floor workers, not typically recipients of much stock-option largesse, are inspired to great efforts by the sight of their corporate superiors getting rich.

As for technology changes, the Internet, as discussed above, has been at best neutral in its effect on productivity, though this may change as the false starts go bankrupt and the true efficiencies kick in.

However, far more important than the Internet’s effect on office life was the personal computer, introduced in the early 80’s. Older readers may remember that economists were puzzled by there being no sign of a productivity surge at that time. Yet the day-to-day effect on office workers’ lives was immense, and layer upon layer of secretarial and administrative assistance became redundant.

At this point, the level of puzzlement rises to a boil, and one’s suspicion is drawn to the figures themselves. They are, after all, produced by the Bureau of Labor Statistics, a government body peopled by doubtless well-meaning employees most of whom, one can venture to guess, are sorry that Al Gore is not President.

In the 1980s, one can further guess, they took the view taken by a majority of historians at that time, that Ronald Reagan was among the least satisfactory of U.S. presidents.

If indeed they held the views imputed to them here, one can further guess that they would not have wished to see a surge in productivity caused by the hated “Reaganomics,” but that conversely President Clinton’s invocations of a beautiful “New Economy” caused by benign Third Way policies would have struck something of a chord in them.

Traditions of the bureau, of course, would have prevented these political sympathies from affecting their work, but as the 1980s wore on, and the middle-level staffers were increasingly baby-boomers, it is possible that they also held the usual baby-boomer disrespect for tradition.

All very well as a conspiracy theory, of course, but is there any actual evidence for it? Well, as it happens, there is.

The BLS updated its productivity statistics methodology in 1995, and again in 1999, using in the latter year new gross domestic product methodology developed by the U.S. Department of Commerce’s Bureau of Economic Analysis.

The changes in the methodology, taken as a whole, are considerable, and the individual effect of each change is unclear. (Although a very good Ph.D. thesis could no doubt be written by studying the effect of each change, one at a time.)

However, there is one change whose effect, at least qualitatively, is analyzable quite easily: In the GDP methodology, business software is now counted as investment, whereas previously it was consumption.

The effect of this is to increase GDP by the amount of business software purchased each year, while decreasing it by the depreciation charge or, as the government statisticians call it, the Consumption of Fixed Capital.

Naturally, as software is an increasingly important part of the economy, new investment generally exceeds depreciation, so GDP is increased, and GDP growth is increased as business software becomes increasingly important to the economy as a whole.

The effect is quite substantial. According to the Bureau of Economic Analysis, 1998 current dollar GDP was revised upwards by $248.9 billion, 2.9 percent of the previous figure, while 1992 GDP was revised upwards by only 1.2 percent.

Thus the 1991-1999 GDP growth rate was revised upwards from 3.1 percent to 3.5 percent, while the 1998 personal savings rate was revised up from 0.5 percent to 3.7 percent. (It has of course since sunk again so as to be below zero even on this new estimate.)

Since business software has been a substantial part of the U.S. economy since at least the 1970’s, one is immediately surprised that this change had not been implemented before. While the effect of the introduction in the 1980s would have been less than today, it would significantly have increased announced growth rates during the Reagan years.

Of course, under the conspiracy theory, the BEA officials did not want to do any such thing.

An additional factor affecting the calculation is the assumed life of software investments, which, in the GDP, are 3 years for package software and 5 years for custom software. These depreciation rates are taken as constant over the whole period.

However, as we know from thinking about it, this is totally unrealistic. The IBM 360 series of computers, introduced in 1965, was not replaced by the 370 series until 1975, so its software had a life of more or less 10 years.

In the 1980s, software had to be compatible with DOS, which was the industry standard from 1981 until 1990 or so. Thus, while several generations of PC software were introduced, a life of 5 to 6 years was probably realistic, higher than that assumed by the BEA, while custom software, still at that time running on mainframes, had a life of 7 to 8 years. (There would, after all, have been no Y2K panic if there had not been considerable pre-1990 software running on U.S. mainframes.)

Conversely, software today is replaced more or less on the Moore’s Law cycle of 18 months — after that, it becomes incompatible with other users’ newer versions.

Internet system software, also, doesn’t have an expected life of 5 years — even had Pets.com not gone out of business, its Internet interface would not still have been in use in 2005.

About 1 ½ years is a reasonable expectation today of a software’s life-span.

Using the same depreciation rates for 1985 as 2000 has a clear effect; it depresses GDP growth rates for the 1980s, by artificially writing off long-life 1980s software too soon, while it inflates 2000’s GDP and GDP growth substantially, by not writing off today’s ephemera quickly enough.

To quantify this, note that gross software investment in 1999 was $180 billion, and in 2000 was $230 billion. Assuming mid-year investment, and taking a 1 ½-year life, CFC — on 1999-2000 software in 2000 — should have been $196 billion, compared with approximately $74 billion on the BEA’s official figures.

Thus you can calculate that 2000’s GDP is wrongly inflated in BEA’s figures by $122 billion, or 1.25 percent, and 2000’s GDP growth, and labor productivity growth, is wrongly inflated by about 0.5 percent.

Of course, all this statistical finagling affects only labor productivity growth, the figure to which most attention is given, and that which is most treasured by devotees of the New Economy.

If, instead, we look at multi-factor productivity growth, a figure also produced by the Bureau of Labor Statistics. then artificially inflating the capital stock doesn’t help you, since it inflates the denominator as well as the numerator.

This figure is enlightening indeed.

From 1982 to 1988, through most of an upswing, multi-factor productivity grew from an index of 87.1 to an index of 95.4, a growth rate of 1.53 percent per annum.

From 1992 to 1998, in a similar economic cycle, it grew from 96.6 to102.4, a growth rate of 0.98 percent per annum.

Not much of a productivity miracle there!

If the reasons for the GDP methodology revisions were innocent, the much trumpeted growth surge would be a myth. It would still be dangerous, because the figures convince mainstream economic decision-makers to plan on the non-existent growth spurt continuing, thus intensifying the over-investment slump that is on the way.

However, while I can’t prove it, the above analysis strongly suggests that the motivation for the mid-1990s revisions was not entirely innocent. Hence the productivity surge is not simply a myth, it is a political scam.

-0-

(The Bear’s Lair is a weekly column that is intended to appear each Monday, an appropriately gloomy day of the week. Its rationale is that the proportion of “sell” recommendations put out by Wall Street houses remains far below that of “buy” recommendations. Accordingly, investors have an excess of positive information and very little negative information. The column thus takes the ursine view of life and the market, in the hope that it may be usefully different from what investors see elsewhere.)

This article originally appeared on United Press International.