(FORTUNE Magazine) – A FEW THINGS we all used to know about pay: Wage earners got paid strictly by the hour. Salary earners got paid by the year. And executives -- only executives -- got bonuses. It was a simple system, seemingly logical, and it worked for decades. Who would want to change it? Answer: Thousands of managers who have decided it doesn't work well enough. Companies in industries from steel to electronics to banking are experimenting with a panoply of radical new companywide pay concepts in an attempt to increase productivity -- and often to cut costs. Lump-sum bonuses replace base wage increases. So-called gain-sharing plans reward improvements in quality and productivity. Pay-for-knowledge systems pay workers for the skills they master rather than the tasks they perform. Such innovations are suddenly sweeping through U.S. business. The American Productivity & Quality Center reports that 75% of employers now use at least one form of nontraditional pay plan -- and roughly 80% of the plans have been adopted in the past five years. Most of these programs share two characteristics: They put more of each employee's pay at risk, and they link that pay more closely to performance. Cynics say that most CEOs need a lot of nerve to adopt such programs, since their own pay is so often unrelated to profits. The point may be valid, but in the past few years some corporate boards have joined in the move to incentive pay by mandating strict pay-for-performance plans for CEOs. The results can be dramatic. When Reebok International earned record profits last year, CEO Paul Fireman, 44, got a bonus of $15,066,700 (see box in following story). That makes him one of America's highest-paid executives. Investors aren't complaining. Even after paying that mammoth bonus, the company earned a standout return on equity of 28%. ''This is the dream that America and capitalism have to offer,'' says Fireman. ''When I grew up, I thought the dream was all gone, used up. The world should know that this is still possible.'' Managers are spreading the word to employees about the lure of incentives. The reason is competitiveness. Not only does incentive pay induce workers to produce more, it also holds down wages and wage-related benefits and lets compensation costs rise and fall with the company's fortunes. In Japan workers receive an average 25% or so of total pay in the form of a flexible bonus. In America the average is still only 1%. Economist Martin Weitzman of MIT believes that variable pay used nationwide could ease inflation and reduce unemployment, as companies are apt to hire more and fire less. He advocates raising variable pay for the U.S. work force to around 20% of total compensation. Contributing to the enormous growth in variable pay has been America's shift toward a service economy. Compensation costs in service companies can eat up 60% to 70% of the operating budget, says Jerrold Bratkovich of Hay Group, a compensation consulting firm. Partly as a result, incentive pay is spreading into hospitals, banks, and other labor-intensive service providers. Bankers Trust, for example, has slowed down salary raises and cranked up incentives to compete for top talent with the likes of Goldman Sachs and Salomon Brothers, where large bonuses have long been the rule. Says Bankers Trust compensation chief Gerard Hickey: ''There is a certain amount of greed in everybody, and that's something we think we can work with.'' Certainly most companies need a new approach. A survey by the Public Agenda Foundation of 845 blue- and white-collar workers showed that 45% believe there is no link between pay and performance. Often there is not. The so-called merit raise often has little to do with merit. Some union contracts dictate that practically everybody get one. In 1982, reports the compensation consulting firm Towers Perrin, merit increases spanned a wide range, from none to 17%. By contrast, in 1988 planned increases were concentrated between 4% and 5%. In other words, almost everyone, sluggard or star, was getting about the same raise. That's motivation? As more managers see the need to adopt new pay plans, they thirst for advice on how to do it right. Jude Rich, president of Sibson & Co., a human resource consulting firm, estimates that this year U.S. industry will spend $125 billion on employee incentives, and much of that money will be wasted. Experts say that roughly half the incentive plans they see don't work, victims of poor design and administration. Executives seeking counsel flock by the hundreds to the holy shrine of incentive pay, Lincoln Electric in Cleveland. Over the past five years more than 3,000 visitors from companies including Motorola, TRW, 3M, and McDonnell Douglas -- as well as several busloads of executives and union leaders from Ford Motor and GM -- have turned up at this manufacturer of industrial electric motors and welding equipment renowned for an incentive pay system that has produced spectacular productivity since 1934. Rather than paying an hourly rate, Lincoln rewards its factory workers on a piecework basis: For each acceptable piece they produce, employees receive so many dollars. In addition, each worker receives a yearly merit rating based on his or her dependability, ideas, quality, and output, which serves as the basis for a year-end bonus. Employee bonuses average 97.6% of regular earnings. The payoff to Lincoln is 54 years without a losing quarter, 40 years with no layoffs, and, according to a study cited by economists William Freund and Eugene Epstein, workers who are up to three times more productive than their counterparts in similar manufacturing settings. WHY DO some incentive plans work and others do not? To find out, Marc Wallace, a professor of management at the University of Kentucky, conducted a two-year study, sponsored by the American Compensation Association, in which he visited 45 companies that have made radical changes in their pay systems. His conclusion: The most critical factor is the existence of a well-defined business strategy. Of the six companies he studied that put in incentive programs for no clear reason, all six failed. The blunder at a manufacturer of engine blades illustrates the trap companies can fall into. The company adopted a plan that rewarded workers for increased productivity. Trouble was, the plan didn't take quality into account. Workers immediately began producing carloads of blades -- some of which lost their twist once inside an engine. Customers such as General Electric and Pratt & Whitney were not amused, and the company ended up doing much costly rework. The most widely used form of incentive compensation is profit sharing: More than 30% of U.S. companies have it. About 85% of these defer at least part of the bonuses by putting them into employee retirement funds, but more and more are paying employees in cash. The immediate reward helps hold down salary and wage increases. Corporations commonly adopt profit-sharing programs in the belief they will increase productivity. But they are not effective motivators, say most compensation experts, mainly because most employees have little influence on profits. Says Jerome M. Rosow, president of Work in America Institute, a nonprofit research firm: ''Many things that affect profits, such as pricing policy, the market environment, and taxes are unrelated to workers' performance. One financial decision, such as taking on a lot of debt or making an acquisition, can wipe out anything the workers can do.'' For companies that are serious about using incentives to motivate workers, gain sharing is the best bet. It is most effective in plants or business units comprising 500 people or fewer, and rewards employees for results they can directly influence: for instance, producing more steel per hour or collecting on more delinquent accounts. When designed correctly, gain-sharing programs can cut labor costs, absenteeism, and turnover, and improve quality and service. Although gain-sharing plans have been around for 50 years, they have lately become the fastest-growing form of incentives. Edward Lawler at the University of Southern California's Center for Effective Organizations says that 26% of U.S. companies use some form of gain sharing. About 75% of these plans have been adopted since 1980. The best evidence that gain sharing works comes from the companies that use it. At Carrier, a subsidiary of United Technologies that manufactures heating and air conditioning equipment, productivity in six plants in Syracuse, New York, was 24% higher in the first eight months of 1988 than in 1986, and the reject rate was way down. Why? Since January, all personnel in those plants have been participating in Improshare, a popular gain-sharing program developed in the Seventies by consultant Mitchell Fein. It is fairly simple to run. In 1986, the benchmark year, Carrier workers needed about 1.8 man-hours to get a finished product out the door. Under the Improshare program, when workers produce more acceptable quality goods per hour than they did in 1986, the resulting savings in labor costs are split fifty-fifty between the company and the employees. Everybody in the plant, from maintenance workers to machinists to managers, gets the same percentage bonus, giving them a strong incentive to work together as a team. How well does Improshare work? Early this year disaster struck a Carrier unit when a water main broke, undermining the floor. Several machines fell | into a sink hole. Employees labored all day and night to get the machines up and running in a different plant so they wouldn't miss their weekly bonus. Says Improshare program manager William Forkhamer: ''That would never have happened in the past.'' So far this year, 2,500 employees in Carrier's Syracuse operation have received a total of $3 million in bonus pay. Workers are currently in a contract dispute with management, but Improshare is not the issue.

Critics of gain sharing say that improvements such as the ones Carrier is making are often due to a pay program's novelty and usually do not last. The folks at Firestone Tire & Rubber's tire plant in Wilson, North Carolina, would disagree. Gain sharing has been in place at Wilson -- a non-union, all- salaried shop -- since 1977 and wheels along smoothly. In contrast with most other Firestone plants, which are not using gain sharing, the Wilson plant has had no layoffs. Annual turnover is less than 0.5%. Says Wilson plant manager Lawson Maddox: ''I don't think there is any question that gain sharing increases productivity and profitability. But the plan doesn't run itself. If you don't constantly reevaluate your formula, you'll put yourself on the road to ruin.'' The Carrier and Firestone plans have something that consultants say every incentive pay plan needs: employee involvement. Mitchell Fein says, ''With gain sharing, management must be willing to do two things -- share and listen.'' At Carrier, for instance, plant productivity information is posted daily on a bulletin board, and workers are encouraged to tell plant managers about any and all ideas. At Firestone's Wilson plant, Maddox meets with all 1,700 employees once each quarter, in groups of 75 to 80, to share information on the budget, business conditions, and the economy. Ironically, increased input from employees can complicate incentive plans in practice. Middle managers often feel threatened by employee participation, sometimes for good reason. A supervisor at one of Carrier's Syracuse plants was dismissed after hourly workers complained to the plant manager that five supervisors were too many. ''We seem to have created a shop floor full of managers now,'' says Forkhamer. ''Many first-line supervisors are afraid of losing control of the shop floor. The old-timers are saying, 'Give me back my baseball bat.' '' The greatest danger companies face when putting in an incentive pay plan is that it may degenerate into nothing more than an entitlement program -- or, as Rich calls it, ''salary in drag.'' Many companies make the mistake of simply layering expensive incentives on top of already competitive salaries, and then paying out those extras regardless of performance. Says Marc Wallace: ''I find a tendency for management to speak in stern, macho tones when adopting incentives. Then, a few quarters down the road, they turn into Father Christmas.'' At Kodak plant workers have kissed Santa goodbye. The company has temporarily abandoned production incentives because they had become automatic: Employees were earning 15% ''bonus'' checks in good weeks and bad. For companies using gain sharing and other small-group incentives, the entitlement issue is thorny because it raises the question of whether divisions or groups that are meeting their individual performance objectives should continue to earn bonuses when the corporation is not making a profit. Berkshire Hathaway Chairman Warren Buffett, who calls himself ''the ultimate believer'' in well- designed incentive compensation plans, believes they should. ''If I had the job of running a football team that was last, but had two outstanding players, I wouldn't pay them less just because they had a lot of clods around them,'' he says. ''Then we'd stay last. Basically, you've got to pay for talent in this world.'' Last year Buffett's three top managers earned bonuses equal to more than five times their base salaries. Two companies, Du Pont and Nucor, have adopted vastly different approaches to ensuring that employees share risks as well as rewards. Starting in January, all 20,000 employees of Du Pont's $5.4-billion-a-year fiber division will start contributing part of their annual wage and salary increases into an ''at-risk pot.'' The process will continue for three to five years until each employee (with the possible exception of some union members, who have yet to vote on the plan) is putting 6% of his or her annual pay into the pot. Bonuses will be based on how close the division comes to its annual target for earnings growth: 4% per year after inflation. If the division meets the objective, employees get their 6% back; if the division exceeds the target by 25%, employees get their 6% plus a bonus of another 6%; if they beat the target by 50% or more, they get the maximum bonus of 12%. As in the most effective incentive plans, there is a downside: If the division doesn't meet at least 80% of its objective, employees lose everything they put into the pot. While Du Pont's plan has teeth, critics accuse it of having the same flaw as corporate profit sharing. Employees, who work in departments as diverse as intimate apparel, flooring, and home furnishings, will have little control over the results as well as vastly different performance objectives. Says Jim Nottke, a laboratory director: ''One concern I have is that the division will now pay too much attention to the short term. In research we must look at the long term.'' At Nucor, a $1-billion-a-year non-union steelmaker, Chairman F. Kenneth Iverson has developed a plan that compensation consultants say represents the best of several systems: profit sharing on top of small-group incentives, with a dash of employee involvement thrown in. ''It is the classic example of an incentive program that works,'' says compensation connoisseur Warren Buffett. ''If I were a blue-collar worker, I would like to work for Nucor.'' So, apparently, would a lot of other people. When Nucor's mill in Darlington, South Carolina, advertised to fill eight openings last fall, over 1,300 applicants showed up, creating such a traffic jam that state police had to be called out. Unfortunately, the force was a bit thin -- three officers were already at Nucor, applying for jobs. WHY THE FUROR? Money. Factory workers at Nucor's five steel mills earn weekly bonuses, based on the number of tons of acceptable quality steel their production team has produced, that average well over 100% of their base pay. Wages are meager, ranging from $5.80 to $9.02 per hour, less than half the union rate. Yet because of bonuses, the average worker at Nucor earns $32,000 per year -- roughly $2,000 more than unionized counterparts. Workers who are late lose their bonus for the day; workers who are more than 30 minutes late lose their bonus for the week. Department managers at Nucor earn yearly bonuses based on return on plant assets; plant managers receive bonuses based on Nucor's overall return on equity. In slow times managers can suffer even as workers continue to earn bonuses based on output. At year-end the company distributes 10% of pretax earnings to all employees, with the exception of corporate officers, who receive a package of cash and stock. Last year Nucor turned out more than twice as much steel per employee as did far larger steel companies, such as USX and Bethlehem Steel. Iverson attributes half that advantage to Nucor's status as a minimill, making | different products than the big guys do, and half to its system of incentives. Says David Redick, a supervisor at Nucor: ''Each crew wants to make more money than the crew in the previous shift. So production goes up, up, up.'' The company hasn't had a losing quarter since 1965 and hasn't laid off anybody in 20 years. Companies are also remodeling compensation programs to complement design changes on the factory floor. Last year Northern Telecom upgraded half its Santa Clara, California, PBX switch plant from an assembly operation where workers repeat the same step over and over to one based on teams of workers who are expected to understand the whole manufacturing process. To support the change, the plant this June adopted a pay-for-knowledge system, which awards raises to workers only when they learn new ''skill blocks,'' such as circuitboard preparation or system testing. Gone are their automatic yearly increases. Says Dick Dauphinais, a compensation director at Northern Telecom: ''To have flexible manufacturing, you must have flexible compensation.'' Money motivates, or at least it can. The challenge is getting it to motivate the right kinds of behavior, which is harder than it sounds. Designing an effective incentive compensation program is far from easy. But as companies such as Lincoln Electric, Nucor, and Firestone will attest, the effort can be well worth the trouble. Changing your pay plan is a big risk, but not changing it could be a bigger one.

BOX: A GUIDE TO INCENTIVE PAY PLANS

PLAN TYPE Profit sharing

HOW IT WORKS Employees receive a varying annual bonus based on corporate profits. Payments can be made in cash or deferred into a retirement fund.

WHAT IT REQUIRES TO BE EFFECTIVE -- Participating employees collectively must be able to influence profits. -- Owners must value employees' contributions enough to be willing to share profits.

ADVANTAGES -- The incentive formula is simple and easy to communicate. -- The plan is guaranteed to be affordable: It pays only when the firm is sufficiently profitable. -- It unites the financial interests of owners and employees.

DISADVANTAGES -- Annual payments may lead employees to ignore long-term performance. Factors beyond the employee's control can influence profits. -- The plan forces private companies to open their books.

PLAN TYPE Gain sharing

HOW IT WORKS When a unit beats predetermined performance targets, all members get bonuses. Objectives often include better productivity, quality, and customer service.

WHAT IT REQUIRES TO BE EFFECTIVE -- Objectives must be measurable. -- Management must encourage employee involvement. -- Employees must have a high degree of trust in management.

ADVANTAGES -- The plan enhances coordination and teamwork. -- Employees learn more about the business and focus on objectives. -- Employees work harder and smarter.

DISADVANTAGES -- Plans that focus only on productivity may lead employees to ignore other important objectives, such as quality. -- The company may have to pay bonuses even when unprofitable.

PLAN TYPE Lump-sum bonus

HOW IT WORKS Instead of a wage or salary increase, employees get a one-time cash payment based on performance or a union contract. The bonus does not become part of base pay.

WHAT IT REQUIRES TO BE EFFECTIVE -- Employees must have a sense that their prosperity mirrors the company's. -- Management must have a good relationship with employees.

ADVANTAGES -- The plan lets companies control fixed costs by limiting pay raises and attendant benefit increases.

DISADVANTAGES -- Management sometimes awards bonuses subjectively, so employees may resent awards they consider unfair.

PLAN TYPE Pay for knowledge

HOW IT WORKS An employee's salary or wage rises with the number of tasks he can do, regardless of the job he performs.

WHAT IT REQUIRES TO BE EFFECTIVE -- Skills must be identified and assigned a pay grade. -- The company must have well-developed employee assessment and training procedures.

ADVANTAGES -- By increasing flexibility, the plan lets the company operate with a leaner staff. -- The plan gives workers a broader perspective, making them more adept at problem solving.

DISADVANTAGES -- Most employees will learn all applicable skills, raising labor costs. -- Training costs are high.

BOX: MAKING INCENTIVES WORK

1. Start with a clear strategy. Think hard about your goals. Focusing just on customer service, for example, could also raise costs.

2. Focus on jobs that can be measured. Make sure that managers know what constitutes peak performance and how to value it.

3. Decentralize incentives. Allow individual business units to tailor reward plans to their specific situations.

4. Separate incentives from base pay. Deliver incentive rewards in a lump sum to highlight the link between rewards and performance.

5. Create a ''sunset'' provision. State conditions, such as the introduction of new technology or a change in business plans, under which the program will be modified or terminated.

SOURCE: PROFESSOR MARC J. WALLACE JR., UNIVERSITY OF KENTUCKY

BOX: TESTS OF AN INCENTIVE PLAN

1. Does the plan capture attention? Are people talking more about their activities and taking pride in early successes under the plan?

2. Do employees understand the plan? Can participants explain how it works and what they need to do to earn the incentive?

3. Is the plan improving communications? Do employees know more than they used to about the company's mission, plans, and objectives?

4. Does the plan pay out when it should? Are incentives being paid for desired results -- and being withheld when goals are not met?

5. Is the company or unit performing better? Are profits up? Has market share grown? Have gains resulted in part from the incentives?

SOURCE: PROFESSOR MARC J. WALLACE JR., UNIVERSITY OF KENTUCKY