SEATTLE/ Conservative Monitor -- With all the mixed economic signals one wonders whether the economy is in recovery or in a continued contraction? Will the stock market make a recovery or experience new lows? Will the Feds efforts to revive the economy result in continued growth, inflation or stagflation?

While economists and others try to make forecasting a science, forecasting is at best an art. Few are successful, and it is not much comfort that Greenspan's record of forecasting before becoming Chairman of the Federal Reserve was less than stellar.

Many forecasters are like the morning paper's astrologer. Their goal is to elicit an emotional response. The most popular is the feel-good response. The economy is improving, which it does most years, and that stocks will increase in price, which they do more often than not. Just keep predicting that the good times will continue, ahem return.

The best record for forecasting is the one that predicts that the next period will be similar to the current period. Next year will be like this year. It is odd, but most people are more comfortable with their current situation, so forecasting that things will be the same offers a certain amount of reassurance to the audience.

For those few who understand how the economy works, their models are handicapped by the use of government data. The government is more interested in direction and trend. It's as if they are concerned that data that bounces around may undermine their credibility. As if help was needed.

The initial employment sample data may not have a good response rate. In a business downturn, personnel departments may be busy planning for and reducing the work force, so they report late. At the local and state level, large firms that report layoffs may skew the data and therefore may be excluded from the sample that makes up the initial reports because they seem to fear that the data may be incorrectly calling a change in direction. The data may be massaged by discarding outliers, using moving averages or other methods.

With inflation data, changes in methodology have reduced the historic measure of inflation by nearly a percentage point. Quality adjustments create problems in measuring price increases. (Should the change in price of a vehicle that resulted from adding an air bag be included in the measure of inflation?) This has become more of a problem as computers and pharmaceuticals have become a greater part of the daily budgets of individuals.

Productivity changes, the sum output adjusted for inflation less the sum of inputs adjusted for inflation, are suspect. How do you explain the recent sudden drop in productivity if technology was having the impact claimed? Are the changes a problem of measuring inputs and outputs, inflation or both? My analysis suggests that both inflation and productivity have averaged closer to the historic norm than previously thought.

Although the record for forecasting the economy in general and turns in particular has been historically bad, the record for forecasting specific industries has usually been reasonably good. The major exception has been the tech sector contraction.

The analysts that followed the tech sector should have been able to forecast the drop in demand for computer and communication equipment. The analysts failed to recognize the rapid increase in demand as a function of Y2K. Rather they assumed these growth rates would continue, they extrapolated the growth over the period 1998-2000 into the future. Basic analysis should have warned that companies would scale back investment in capital expenditures on technology after January 2000, especially as Y2K was a bust, while trying to digest the new equipment. With double the normal investment, it may take several years before these industries begin to invest at the old growth rates.

The same could be said of the growth in dot com companies. The market for Internet sales was already somewhat defined, although there was some opportunity for expanding this marketing and sales vehicle. The defined market was existing catalog sales, which should shift, albeit gradually in some cases, to the Internet for convenience and speed of making a purchase. While some growth could be expected once people became more comfortable with purchases over the Internet, analysts that extrapolated sales increases into the future didn't seem to bother to check with the consumer as to why they prefer shopping in stores. They ignored the social aspects of mall shopping.

The problem with the above examples is that the analysts rarely have advanced training in economics or business experience that make them creditable witnesses to future economic events. They extrapolate because it is an easy tool. The investors and the public can relate to this approach since they observed trends continuing. In addition, the effort to sell securities biases their opinions.

It is still unclear what character the reduction in economic activity will take (slowdown in growth, recession or something more) the data seems to suggest an economy in trouble. Added to the apparent world-wide contraction, it is unlikely we will have a quick recovery to previous levels of growth in economic activity or in profits. While it is constantly claimed that basic economic principles do not apply in "this case" or that "things are different this time", the business cycle is alive and well. While the duration and magnitude may have changed, we are in a business correction. It's not possible to have competitive price pressure, rising wages and energy costs, and increasing profits, especially if productivity is not increasing. The computer and communications sector products (technology) appear more and more like commodities which means prices will continue to deteriorate. This correction is going to be deeper and longer than most people expect.