Productivity is not easily understood by economists. That probably is why virtually no one expected productivity to actually decline for the first time in six years during the winter.

As economists, we are trained to recognize that workers can be more efficient either because they are more knowledgeable about their work, can use their education to improve output, can use better tools, can organize their work more efficiently, be more motivated by their management, or have better equipment to accomplish their tasks.

Furthermore, an economy can be more efficient even if none of the above happens if the work is being done where workers have higher efficiency. If economic growth is concentrated in labor intensive services such as nursing care, productivity may slow or even decline. If growth is occurring most significantly in computer or communication equipment production, productivity can soar even if the production of computers or communications equipment no longer is experiencing productivity gains.

Even when we use all the above explanations to measure productivity gains, economic studies usually have something left over -- the residual-- that has not been explained. Unfortunately, this residual usually is fairly large, about a third of all the explanations for productivity gains.

We also know that changes in the rate of growth cause changes in productivity. This is the so called cyclical effect. If orders do not enter the door, machines are idled and workers are shifted to less productive activities, such as refurbishing the facilities or re-organizing the inventory. Because of data limitations, our studies tend to use machines available rather than machine hours of use. Similarly, we use hours paid for rather than hours working at peak efficiency tasks.

Obviously, less is produced from an idle machine than an operating one. Furthermore, it takes time to adjust machinery to the new reality of orders. Increasingly, time also is spent before hard to hire employees are laid off. Indeed, the speed by which excess labor is shed is inversely related to the effort of hiring those workers in the first place.

A tight labor market into this year probably delayed the reduction in employment that was required if orders remained weak. Thus, employment grew even as growth slowed in the winter. Only in the spring did employers outside of manufacturing decide that they had to reduce their employees.

Certainly, a large portion of the reduction in productivity during the winter was the result of this reduced use of equipment and lowered effectiveness of workers,

While economists know why productivity deteriorates when the economy slows, we have no good formulas for telling us how much of the slowing is related to the slowing and will be restored when orders rebound. Indeed, most models use the trend growth in productivity that is established over a long enough period to preclude these shifts in the idling of equipment and employed people.

If we knew what was happening to the trend rate of productivity gains, we could make some conclusions about what might happen to the growth of this economy after the weakness subsides. Will productivity rebound to the 2.5 percent annual rates of growth assumed by the Bush administration in its projections of surpluses as far as the eye can see? If not, higher interest rates may be in store for bond buyers.

As interest rates remain an important determinant of the value of future company earnings, that might also imply that stock values could grow at a substantially slower rate than virtually everyone other than Warren Buffett currently is assuming.

Also, a temporary dip in productivity usually is borne by profit recipients. That first quarter dip means that profits will again fall sharply. However, a permanent downward shift in productivity gains without any relief in hourly wages will mean a higher inflationary environment in the future.

In short, our ignorance is in an area that requires more understanding than the economics profession currently has on what determines productivity changes.

My own guess is that most of the slowing is cyclical and will not be reflected in higher inflation. However, the slower growth in technology based activities and the slower spending on equipment will cause lower than trend productivity gains for several years. This means significantly slower profit growth, somewhat higher inflation, and substantially lower government surpluses than currently assumed.

And as all my faithful readers know, those items mean a great deal to the performance of the economy, our economic well being and the return on our investments.