Productivity is easy to understand, but tough to pin down in a service industry like trucking. Even if you get good numbers, what do they tell you?

As a management concept, productivity improvement is irresistible. It's simple to understand -- get people to do more work without increasing their workload or workday. It offers big payoffs for everyone involved -- employers can cut operating costs while employees can generate the revenues needed to increase their pay. And it boosts both competitiveness and profitability without setting off inflation.

The problem is that few people seem to agree on just what constitutes truly improved productivity in the real-world workplace or how to measure it. Take the federal government, for example. The U.S. Dept. of Labor says that productivity for all American workers except for those involved in farming grew at an annualized rate of only 0.6% in the second quarter of 1997. They also say that they were wrong about initial productivity figures for the first quarter and have revised them significantly downward, from 2.6% to 1.4%.

According to a number of economists quoted in places like The New York Times and The Wall Street Journal, those numbers have to be wrong. In their view, the low inflation rate and soaring corporate profits and stock prices of the past few years can only be explained by far more robust improvements in American business productivity. The improvement is there, they say, it's just that the Labor Dept. is using outdated standards to measure it.

If you make something, measuring gains in productivity is relatively straightforward. Critics of Labor's productivity measuring point out that the numbers for manufacturing show big gains this year -- 3.2% for the second quarter, up from 2.5% in the first. By extension, the far larger service and financial sectors of our economy must have lost a lot of ground to give us the department's low overall numbers.

Given the strength of our present economy, the only logical explanation for the low growth numbers, the critics say, is that the government hasn't figured out how to get good productivity measures in these more abstract areas.

As a service industry, trucking falls into that hard-to-measure category. Common wisdom is that the industry as a whole is far more productive since deregulation, although just how much more productive is difficult to determine. Certainly, different types of fleets have come up with relatively accurate productivity gauges for their particular operations. Truckload carriers, for example, can track revenue miles, while less-than-truckload fleets can measure shipments per person. Private fleets, at least those that operate most like TL carriers, can also watch loaded miles, while vocational fleets can look at product delivered or hauled per person.

Unlike government statisticians who can't devise a single productivity measure that fits the entire trucking industry, you can do a fairly good job when it comes to tracking performance of your own fleet's operations. But you still share the Labor Dept.'s basic problem -- what do the numbers mean, and what do they tell you about the success or failure of your management initiatives?

For example, suppose you decide to invest in a full-blown automated dispatch system complete with wireless communications. Revenue miles, or whatever other productivity yardstick you use, climbs by x%. Cost justification for the investment is as simple as dividing the increased revenue or loaded miles by the cost of the system. You've made a sound business decision that's supported by the numbers.

But if you're honest with yourself, is it really that simple? First, how can you be sure that the productivity gain is really due to the new technology and not to some minor change in the operation brought about by installing the system? Maybe it's due to simply putting company-wide attention on dispatch. And even if the entire gain is due to the new system, how do you factor in the potential loss of business from all those shippers who said they only want to use fleets with wireless tracking capabilities?

Try as hard as you might to quantify it with numbers, managing a fleet for improved productivity is still far more an art than a science.