One day Don Miller, CEO of Roppe Corp., in Fostoria, Ohio, was listening to a local factory owner gripe about his workers. Once the workers had reached their production quotas for the day, the colleague said, they would sit around and do nothing until quitting time. Why not let them go home, Miller suggested. "He said I was crazy," Miller says. "So I decided to try it myself."

Roppe manufactures rubber baseboards, tile, stair treads, and the like. Before the program was begun, some eight years ago, the $50-million-plus company's four production lines had averaged about 75% of their quotas; Miller figures that workers feared management would just raise the quotas once they met them consistently. Hoping that the prospect of leaving early would override that disincentive, Miller experimented with one production line. He set a new quota that was 10% higher -- 29,000 feet of rubber baseboard a day instead of 26,000 -- and in return, he agreed to pay those workers 10% more per hour and to let them go home when they reached the new goal. (The workers would still receive eight hours' pay.)

Within a week workers on that line were meeting the new quota and leaving work about an hour early. Soon Miller extended the program to all lines, with similar effects. As a result, he found himself with happier workers and a 35% increase in productivity, for a 10% raise. Since then, he's noticed ancillary benefits: teams won't tolerate sabotage efforts by new workers; supervisors no longer must time employees' lunch breaks; maintenance workers, who can work between shifts, charge less overtime.

One danger of such a program is that quality might slip, but Miller insists that hasn't happened, partly because of strict production standards and partly because workers don't want to spoil a good thing.

There's another danger: that Miller will get greedy and up the quota when he sees that workers could produce more in a full day. Success depends on good faith in setting the quotas; if the goal seems unattainable, workers will revert to cynicism. So Miller negotiates new quotas only when new, more efficient equipment is installed or improvements are made to operations.

Changes like those occur less than once a year. When they do, in-house experts conduct a time study of the improved process and recalibrate the quota. And they set it so that workers should be going home about an hour early. As proof of the trust that the arrangement inspires, Miller points to the fact that Roppe's work force remains nonunion. Meanwhile, the company's cost of labor as a percentage of sales has dropped, and wages have risen.

* * *