Choosing Inequality

Barron's , June 1993

When Cypress Semiconductor faced its first sales and profits downturn, the company took a common step: It froze salaries, giving no raises in 1992. Going into 1993, however, the chief executive, T.J. Rodgers, made a radical decision: Regardless of how well Cypress did, it would once again give raises, at a corporate average of 5%, even if that meant laying off lower- performing employees.

“Imposing an indiscriminate salary freeze,” says Rodgers in his new book, No-Excuses Management, “is like placing a vat of crude oil on an open flame. For a while nothing much happens. But the longer you maintain the freeze — the longer you heat the crude — the more quickly you boil off the valuable, volatile elements like gasoline. Boil the crude long enough and all that’s left is the tarry sludge.”

In other words, Cypress had a choice: Keep the salary freeze and lose its most valuable employees to competitors. Or use merit-based layoffs to free money for raises.

Cypress is an extreme case. Determined to stamp out “the mediocrity of the mean,” the company has an official policy of ranking employees in each job from top to bottom by performance, trying to pay the top-performing employee 50% more than the bottom-ranked person, giving no two people with the same job the same merit raise. No employee gets the average raise.

Cypress’s policies sound radical but its meritocratic philosophy is hardly unique. Since the late 1970s, a “pay-for-performance revolution” has swept American business, fundamentally altering notions of what is “fair” when it comes to paying employees.

“The definition of fairness has changed,” says Craig Ulrich, a principal in the New York City office of William M. Mercer, a human-resources consulting firm. More companies consider performance the main criterion for raises and promotions, ahead of such traditional factors as seniority. And from line production workers to executive-committee members, an increasing number of American employees collect much of their pay in performance-based bonuses.

A July survey of 56 member companies of the Council for Continuous Improvement found that 66% of hourly workers were eligible for quality-oriented bonuses. For white-collar workers, Ulrich estimates, 10 years ago half of all companies offered performance bonuses to those making at least $65,000 a year. Today, three-quarters offer bonuses, and they’ve dropped the threshold to about $45,000 in constant dollars.

The effects of this compensation revolution are obvious within the business world, where consultants and managers constantly experiment with new ways to reward good work. But in the world of public policy, the revolution is barely recognized. And the notion that employees who do the best work should get the highest pay has profound implications for the political debate over fairness in economic policy.

By rewarding achievement above other factors, performance-based pay systems create income differences between individuals doing “the same job,” between individuals with the same experience, and between individuals with the same educational credentials. These systems in fact push companies to pay more valuable employees more even if they have less education or seniority than other workers. (Performance-based differences also appear, though perhaps in diluted form, when incentive pay rewards work teams rather than individuals.)

Paying more valuable employees more means paying less valuable employees less. “The people who are rated in the bottom half won’t like it,” says Jude Rich, president of Sibson & Co., a Princeton, N.J.-based human-resources consulting firm. “Whether that’s different teams or individuals, that will be true. But the real question is, what do they do about it? In more cases than not, what happens is they work harder. They try harder. Because they don’t like being at the bottom of the pack.”

One of the biggest changes is in how companies value seniority. Ten or 15 years ago, Ulrich notes, seniority played a much bigger role in compensation. Younger employees were expected to pay their dues, to bide their time at low salaries in anticipation of payoffs later on.

But today, he says, “If you’re getting the job done and if you’re of tremendous value to the organization, it really doesn’t matter whether you’re 30 years old or 40 years old or 50 years old. You’re going to be paid accordingly.” That seems only fair to the young, but it looks wrong to their older colleagues, who may have waited for rewards that never came. In candidate Bill Clinton’s phrase, they “played by the rules,” only to find that the rules had changed.

Consider the raging debate over the “fairness” of income distribution in America. If we ignore the considerable effects of income mobility — the “poor” today aren’t the same people as the poor in 1983 or 1973—there is some truth to the statement that the rich are getting richer. But the trend has nothing to do with anything as transitory as tax policy. And it doesn’t imply that the American economy has suddenly become “unfair.”

From at least 1963 on, if you divide full-time working men aged 25 to 54 by income, you find that real hourly wages for those in the bottom half of incomes are slowly creeping down and those in the top half are slowly creeping up. Why?

A major reason seems to be that the payoff to higher education jumped sharply in the 1960s and has skyrocketed since 1978, once the older baby-boomers began to settle down. Whether this trend is good news or bad news depends on how you feel about the market spontaneously rewarding brains over brawn.

As Kevin M. Murphy of the University of Chicago and Finis Welch of Texas A&M noted in a paper at January’s American Economics Association meetings: “The Wall Street Journal splashed the good news across its front page that college pays more than ever before — the evidence was that the wages of college graduates were high and rising relative to wages of high school graduates....Newsweek bemoaned increasing inequality, pointing to the low and falling wages of high-school graduates relative to college graduates.”

But general wage dispersion isn’t the end of the story. Inequality is increasing even among people with the same number of years of schooling or experience, Murphy and two other economists find in a more recent paper published in the Journal of Political Economy.

Taking wages as the prices employers pay for certain attributes, they conclude that while the demand for both education and experience has indeed increased, the demand for what they call “unobservable ability” — how well people do their jobs — is even more important. What matters, it turns out, are those performance differences that managers, looking at individual workers, can notice and reward. Economists, looking at Census Bureau aggregates, can never hope to identify them.

To beat back rising inequality, egalitarian politicians can’t just establish apprenticeship programs or urge kids to stay in school. They would have to undo the systems managers have created to reward different people differently. They can hammer well-paid employees with higher taxes. And if that doesn’t work, they might choke pay differentials directly, by making them illegal.

For now, people can still argue about the best way to pay employees, and employers can put their theories to the test in the marketplace. To some people in some circumstances, rewarding loyalty and experience—seniority—seems more important than paying simply for today’s measurable achievement. For others, especially the young and ambitious, that seems unfair. Managers at AT&T Universal Card, believe giving everyone the same bonus, based on achieving companywide quality measures, is the best way to build a successful team. At Cypress Semiconductor, such an indiscriminate award would violate the company philosophy.

There surely is no single right way to pay employees. There are only different customers, different employees, different products, different markets. Corporate pay systems are not, after all, measures of moral worth but of economic reward. As such, they require constant tinkering and experimentation.

But politicians, even those who praise business innovation and new technologies, rarely notice, much less appreciate, the experimental process that invents new ways of managing employees. And when those experiments succeed and spread, as the pay-for-performance revolution has, they decry the change.

Paying for performance is indeed an affront to the civil-service philosophy of rewarding credentials and years on the job rather than judging performance. But for employers and employees who believe pay should reflect productivity, not arbitrary criteria or the boss’s whim, the new inequality is quite fair.