Using the Labor Dept.'s Displaced Workers Survey, which is conducted every two years, Farber finds that the percentage of workers reemployed after a spell without a job rose sharply during the 1990s (chart). By February, 2000, the date of the most recent survey, 75% of workers who lost their jobs from 1997 to 1999 had landed new ones. In contrast, only 61% were reemployed by early 1992 after losing work from 1989 to 1991.
Workers were also increasingly able to find new jobs at a pay level closer to that of the job they lost. Farber estimates the average real weekly earnings loss at 12% for reemployed full-time workers who were displaced from 1989 to 1991. Yet those who were displaced from 1997 to 1999 lost an average of only 2% upon finding new employment.
But Farber's work also shows that losing a job exacts a price even in boom times. Although those workers who were reemployed in the most recent survey had gotten back to near their previous wage level, they missed out on wage gains enjoyed by those who didn't suffer a spell of unemployment. Workers who stayed employed during the 1997-1999 period saw their weekly earnings rise an average of 10% during that time--an increase that the displaced workers never realized. Conventional wisdom says that giving nonsupervisory employees greater independence and responsibility is good for both them and their employers. Companies that place workers in self-managed teams, rotate jobs, and encourage employee feedback, it's believed, achieve greater workplace flexibility, higher productivity, and better business performance. The benefit for employees is greater job stability, because these practices should make their companies more successful, so they don't need to lay off workers or use temps.
But a study just released by the National Bureau of Economic Research finds that such practices usually yield quite a different result. According to its authors, Peter Cappelli of the University of Pennsylvania and David Neumark of Michigan State University, flexible work practices reduce layoffs and employee turnover only in the shrinking manufacturing sector, not in the rest of the economy. What's more, employees do not show any additional loyalty to nonmanufacturing employers that use these practices: Workers usually quit or retire at a higher rate in those companies.
Cappelli and Neumark based their review on a 1997 Census Bureau survey of more than 1,800 plant and business managers. They concluded that flexible work practices were most successful in manufacturing because of the nature of the work. Jobs are more narrowly defined in manufacturing, so flexible practices, especially job rotation, produce greater improvements in a factory than in a retail store, Cappelli says.
Overall, the authors find that flexible work practices are being used in combination with layoffs and temporary employment, not as a way to avoid them. Even manufacturers increase the use of temporary help while using practices that make work more flexible.
Offering greater job stability might make workers more loyal, but that doesn't seem to be a big consideration for most companies that are using these work practices. Instead, the goal seems to be to employ workers so that "just the right amount of labor is fit to the exact amount of demand," says Cappelli. And this goal isn't likely to change any time soon. As the U.S. economy slows, companies are likely to face increasing pressure to transform their payroll expenses from a fixed to a variable cost. Whether or not the U.S. economy dips into recession this year, the critical factor in keeping growth strong in the long term is innovation. And innovation, in turn, depends critically on having a steady stream of new scientists and engineers. On that score, there's some good news. According to the National Science Foundation, enrollment of science and engineering graduate students rose 2% in 1999--the most recent year for which statistics are available--after five consecutive years of decline. The biggest gain, 12%, was in computer science. Engineering enrollment ticked up 1%, for the first gain since 1993. The numbers of African American, Hispanic, and Asian American graduate students rose, as did the number of students who were enrolled in graduate school for the first time.
The bad news? Foreign students on temporary visas were the main factor in the 1999 hike. Their numbers rose by nearly 8,000, to nearly 110,000, while students who are U.S. citizens and permanent U.S. residents fell by a little more than 1,000, to just over 300,000 (chart).
Foreign students are less likely to stay in the U.S. and contribute to its economy. The last time the NSF studied the topic, in 1994, it discovered that only 48% of foreign students who received U.S. doctorates in science and engineering were still in the U.S. a year or so later.