Corporate loyalty -- repeatedly pronounced dead -- survives. It has diminished and disguised itself, but it has not disappeared and may have just gone into hiding. Of course, we've all heard the obituaries. Companies, increasingly beholden to Wall Street, care little for workers. Everyone can be fired on the spot. Workers reciprocate. The shrewd among us cultivate job skills, anticipating a pink slip or searching for a better offer. Everyone is, potentially, a free agent.

As these ideas seeped into popular culture, they inspired some extravagant theorizing. Sociologist Richard Sennett of New York University argues, for example, that the fraying of long-term jobs is destroying people's broad capacity for trust. In "The Corrosion of Character: The Personal Consequences of Work in the New Capitalism," he writes: " `No long term' means keep moving, don't commit yourself, and don't sacrifice."

Theories like this are more seductive than plausible. On reflection, their logic is questionable: Do people mainly form personal trust based on their jobs -- or do relations with parents, friends, relatives and teachers matter more? Then there's the problem of facts. The death of corporate loyalty falsely presumes that long-term jobs have virtually vanished.

It's true that, in the 1980s, companies became more willing to fire. But this hardly obliterated career jobs. The effect has been more modest. In 1983 about 38 percent of men 25 and over had worked at one company for 10 years or more; by 1998, that was 33 percent, reports the Labor Department. Job tenure rises with age, and among men the loss has spanned the age spectrum. In 1983 about 66 percent of men 55 to 59 had been with one company at least 10 years; by 1998, that was 57 percent. By contrast, job tenure has risen slightly among women, as more pursued careers.

What's also untrue is that middle-aged workers -- stereotyped as costing more and doing less than younger workers -- have suffered most from layoffs. This class of disemployed has inspired many hard-luck stories. Being middle-aged, I'm sympathetic. But studies refute the anecdotes. Consider one study of 51 big companies that are clients of Watson Wyatt Worldwide, a consulting firm. It found that job losses in the 1990s focused mainly (as they always have) on the young. About 70 percent of workers with 15 to 19 years at a single company stayed for the full study period of five years (and, among leavers, many left voluntarily).

Finally, most workers aren't acting like free agents. If they were, wages would be exploding. Labor markets are tight. The unemployment rate has been below 5 percent since mid-1997. Free agents -- believing their time at any company is brief -- would exploit the situation. They'd demand big raises or go elsewhere. In general, that isn't happening. Over the past year, labor costs have risen only 4 percent. People value stability or fear change. Maybe they're loyal.

"We are not moving to an economy made up only of short-term jobs, indifferent employers and disloyal employees," notes economic historian Sanford Jacoby of the University of California, Los Angeles. Why is this?

Perhaps pressures from Wall Street are exaggerated. This is doubtful. Economists Henry Farber of Princeton and Kevin Hallock of the University of Illinois studied almost 3,000 corporate job reductions between 1970 and 1997. Early in the period, job cuts mainly reflected a weak economy or dire company circumstances that forced plant closings. Later, the reasons became a vaguer need to control costs -- aka, raise profits and stock prices.

The real reason that career workers haven't become extinct is that it doesn't make any economic sense. Companies depend on people who know their products, customers, markets and work methods. This takes time and costs money. High turnover is disruptive. People without long-term attachments are more mistake-prone, being more ignorant and indifferent. Random layoffs poison morale.

These considerations, of course, count for more in today's taut labor market. Companies are desperate to attract and keep good workers. But the logic never disappeared. Consider Intel, the huge computer-chip maker. In 1998, it spent $312 million to train workers. Training is considered so important that every senior manager teaches at least four classes a year. "When you invest this time and money, you don't want to lose workers," says Nancy Lindsay, an Intel staffing manager.

Still, this sort of approach wins little overt gratitude, because few companies now promise lifetime employment, even if they often provide it. ("We did layoffs in 1986 and 1987," says Intel's Lindsay. "We vowed to avoid them again -- if we could.") Job cuts occur even in prosperous times, driven by mergers, shifting market conditions and management mistakes. As absolute security vanished, anxiety and suspicion rose. Not surprisingly, the share of employees saying management "is generally respected" fell from 58 to 50 percent between 1990 and 1997, reports Jacoby.

But the larger reason that corporate loyalty seems dead when it isn't is that few people will acknowledge it. It's not smart or cool. In a perceptive essay, John Clancy of Washington University notes that people won't speak well of most major institutions, private or governmental. "Loyalty and other related aspects of social character began to change well before the era of downsizing," he writes in Across the Board magazine.

No one wants to vindicate the writer Ambrose Bierce's biting comment that loyalty is "a virtue peculiar to those who are about to be betrayed." But people also "have an innate need to become attached to something larger than themselves," argues Clancy. What to do? At work, says Clancy, people often resolve the conflict by projecting their loyalty onto the "product they helped create," be it software, an auto design or a magazine story. This saves them "from the abyss of cynicism" -- and it's a socially acceptable form of corporate loyalty.