The US economy is punctuated by an increasing number of low-wage employment even as the proportion earning of middle-level wages continue to decline. This observation tends to support the general theory of a number of economists regarding the growing polarization of wages. Despite the warning issued by economists that flexible production eventually leads to job polarization, many business leaders continue to favorably support flexibility as a way to boost organizational productivity.
Adaptability, flexibility, and speed to market are among the terms that business leaders are using to describe their efforts to improve organizational quality and productivity. But as this excerpt from Lean and Mean shows, flexibility can come at a very dear price.
It is the summer of 1983. The president, Congress, business leaders, economists, and the media are joyously celebrating the fact that, at long last, the national economy is coming out of the back-to-back recessions of the previous three years. How infelicitous that one journalist should choose that moment to cast a pall on the festivities.
According to a widely read article by Robert Kuttner, published in the Atlantic, the share of Americans earning middle-class wages was declining over time and would likely continue to do so. The reasons included the growing vulnerability of the economy to foreign competition, stagnating productivity growth, and the long-term shift of the economy's center of gravity from manufacturing to services. To a weary establishment, both inside and outside of government, Kuttner's challenge was simultaneously too well argued to dismiss out of hand, and too scary to take seriously.
The idea of a declining middle America circulated within Washington throughout the remainder of the decade. But as the economy continued to grow after the trough of the 1982 recession, thanks to a combination of Reagan-era military spending and unprecedented borrowing by consumers and businesses, it became more and more difficult, and seemingly unnecessary, to consider the idea that America was becoming an increasingly polarized society. After all, inequality historically had always fallen during periods of economic growth, as the unemployed returned to work and selective labor shortages began to appear in one or another corner of the country.
As it turned out, Kuttner proved to be right on the money. From the vantage point of the 1990s, it is hard to find any serious observer who does not agree that inequality is on the rise. The polarization of the jobs that employers are making available to people searching for work is cleaving the whole population, white and black, Anglo and Latino, into highly paid haves and more poorly paid, increasingly insecure have-nots.
What the M.I.T. economist Lester Thurow has called the "surge in inequality" may, at least in part, be connected with industrial restructuring and business reorganization. Lean production, downsizing, outsourcing, and the growing importance of spatially extensive production networks governed by powerful core firms and their strategic allies, here and abroad, are all part of businesses' search for "flexibility," in order to better cope with heightened global competition. But this very search for flexibility is also aggravating an old American problem--economic and social dualism. This is an institution that was widely thought to be disappearing along with those dinosaurs of the industrial past, the vertically integrated giant corporations.
The popular pronouncement of the imminent demise of the big firms was premature. It is they and their partner companies, not small business, per se, that account for most of the jobs, sales, and output in U.S. industry, year in and year out, in both mature and high-tech sectors. But the ways in which big business has been reorganizing itself to become more competitive are proliferating low-wage, insecure employment. The trend toward our becoming an hourglass economy proceeds. This is the dark side of flexible production.
THE GROWING POLARIZATION OF WAGES
In a paper originally commissioned by the U.S. Congressional Joint Economic Committee and published in late 1986, Barry Bluestone and I presented the results of an analysis of official Census Bureau data on the earnings of individual U.S. workers. We showed that since at least the late 1970s, it was the lowest-wage and the highest-wage groups of Americans that were increasing in number most rapidly, while the proportion earning middle-level wages was falling. This message, in the form of a qualification to the Reagan administration's bullish self-promotional claims for the wonders of its track record in creating jobs, was subsequently published as an opinion editorial in The New York Times in February 1987.
The result was a firestorm of public and academic debate. With the passage of time, the clumsy attempts by the conservatives then running the government, and by journalists sympathetic to their cause, to dismiss these findings as either "absurd" or (worse) "cooked up" in order to suit the political needs of the Democratic party opposition now seem fairly amusing. What was more important was the debate that this work triggered among academic economists. Over the next four years, literally dozens of books and papers were published, conferences held, interviews granted to the media, and careers made over the question of whether, during the sustained macroeconomic expansion after 1982, inequality among wage earners was in fact continuing to grow--let alone to polarize.
Economists and statisticians justifiably worry whether, even using the same original data, different researchers might come to different conclusions about some phenomenon because of, say, how they account for the effects of inflation, or how they choose to define the variable they are measuring (income? wages? compensation, which includes both wages and benefits?). The quality and content of the four-year debate about income inequality improved as the protagonists increasingly focused on these technical questions and stopped attacking one another on ideological grounds. Gradually, the facts did come to speak for themselves, more or less, and nearly all economists looking at the problem came to agree.
They agreed that average post-World War II earnings (or at least the rate of growth of earnings) peaked back in the early 1970s. Their computer analyses told them that inequality was growing procyclically--that is, even as the economy expanded, in contradiction to all previous evidence and to the predictions of standard theory. They concurred that the proportion of Americans earning poverty-level wages was increasing among men and, by some accounts, among women as well, but that women were finally catching up to men, both because of the changing structure of the economy (more service, office, and white-collar jobs, and fewer factory jobs) and because men's condition had worsened so much. Finally, the economists agreed that the dimensions along which U.S. workers were drifting farther and farther apart from one another were work experience (how many weeks and hours of employment a person had) and, most of all, hourly wage rates. This growing polarization was observable even among workers with comparable personal traits, education, and years of experience.
The appearance in 1992 of three widely read technical documents published in prominent places, inside and outside the government, marked the convergence of assessments on the extent, if not the causes, of the phenomenon of secularly widening earnings inequality. One document was a report published by the Census Bureau on workers with low earnings. Looking only at the roughly half of all people employed year-round and full-time (as Bluestone and I had done in most of our own earlier writing), the government found "a sharp increase over the past decade in the likelihood that a year-round, full-time worker (or a worker with a year-round, full-time attachment to the labor force) will have . . . annual earnings [below the poverty level for a four-person family]. In 1979, 7.8 million or 12.1 percent of all year-round, full-time workers had low annual earnings. By 1990, the number of year-round, full-time workers with low annual earnings was 14.4 million and the proportion was 18 percent.
"Young workers have the highest likelihood of receiving low earnings . . . but the rate has increased since 1979 for all age groups below 65 years of age. . . . The rate for persons with one or more years of college [also] rose from 6.2 percent to 10.5 percent."
At the Economic Policy Institute in Washington, DC, the research director, Lawrence Mishel, and a staff economist, Jared Bernstein, looked at all workers, not just those employed full-time and year-round. Mishel and Bernstein used the same data as had the government and virtually identical methods of analysis. They concluded that, between 1979 and 1989, from one business cycle peak to another, the proportion of U.S. workers earning hourly wages below the poverty line--about $6.50 in 1991 purchasing power--rose from 25.4 percent to 28 percent. Over the same decade, the fraction of the workforce earning three times the poverty line and above--about $19.50 per hour--rose from 7.3 percent to 8.7 percent.
The third major milestone of 1992 was an exceptionally exhaustive review of the debate, published in the American Economic Association's official review medium, the Journal of Economic Literature, written by M.I.T.'s Frank Levy and Harvard's Richard Murnane, two well-known economists specializing in the study of work, education, and income. Levy had himself been a major participant in the debate, so people were especially eager to read his updated views. That various drafts of this paper circulated among economists all over the country, with correspondence flying back and forth, reveals much about how seriously the research and policy communities had come to take a subject that, only half a decade earlier, had been almost completely dismissed.
Drawing heavily on a doctoral dissertation originally written at Yale by the economist Lynn Karoly and completed by her at the RAND Corporation, Levy and Murnane concluded in their review that "Nineteen-hundred-seventy-three marked the end of rapid real earnings growth and the beginning of slower growth bordering on stagnation. Nineteen-hundred-seventy-nine marked the beginning of a sharp acceleration in the growth of earnings inequality, particularly among men.
"[Between] 1979 and 1987 . . . the proportion of men earning more than $40,000 (in 1988 dollars) increased, while the proportion of men earning less than $20,000 increased as well. The combination of increased earnings inequality around a slow-growing average means that significant numbers of workers particularly younger, less educated men--now earn less than their counterparts of the mid-1960s.
". . . [In other words, at least] the male annual earnings distribution has 'hollowed out,' leaving larger percentages of workers at the top and bottom of the distribution, and a smaller percentage in the middle. At least for men, it is now clear that there were fewer middle class jobs in the mid-1980s than a decade earlier."
This was, in all important essentials, the central proposition of the original magazine article by Kuttner and of that report to Congress written by Bluestone and myself six years earlier, which had triggered the debates and research projects over income inequality to begin with. Subsequent popular books by the political commentator Kevin Phillips and others only repeated the bad news, to the point where the polarization of earnings and the crisis of the middle class became the most salient issues of the 1992 presidential campaign.
WHY CONVENTIONAL EXPLANATIONS DO NOT FILL THE BILL
One possible explanation for the growth in earnings inequality, at least within the context of U.S. institutions, dominates the thinking of mainstream economists. This is the straightforward, plausible idea that the market value of a college education rose during the 1980s. This could have occurred because, even in the older manufacturing industries, employers were requiring more and more technical skill from their workers, whatever the color of their collars. If so, then people with advanced degrees would be expected to be in a position to command a higher rate of pay per hour than others. And indeed, several economists initially thought their data did show evidence of such a shift in the structure of managers' demand for labor.
There surely was a widening wage gap during the first half of the 1980s between college-educated workers and those who never progressed beyond high school. But Mishel and Bernstein demonstrate that this growing gap resulted primarily from a "precipitous decline of wages among the non-college educated work force and not [from] any strong growth of the college wage." And in any case, by 1987--well before the onset of the 1990 recession--the absolute inflation-adjusted wages of even college graduates began to decline. Moreover, as Harvard's Larry Katz, currently the chief economist of the U.S. Department of Labor, has himself reported, inequality has grown among even people with the same levels of education. Katz and the University of Chicago economist Kevin Murphy estimate that, since 1970, there has been a 30 percent increase in such within-group inequality--a rise of stunning proportions.
For Levy and Murnane, as for other economists who study income distribution and the structure of labor markets, "the most important unresolved puzzle concerns the reasons for the almost 20-year trend toward increased within-group earnings inequality." After all, standard economic theory predicts that, after accounting for differences in such ascriptive traits as race, gender, and age, workers with the same education and work experience should receive about the same rate of pay. (This expected leveling is driven--at least according to the theory textbooks--by the mobility of labor. Thus, anyone with skill will move from a "bad" job to one with more attractive prospects and pay. And firms will gladly pay to attract "good" employees.) To add even further to the conundrum, wage rates in the 1980s varied significantly even among factories or offices within the same industries, and even among ostensibly identical jobs within the same business establishment. How could this be so?
Harvard's Richard Freeman is the director of labor studies at the National Bureau of Economic Research and the dean of empirically oriented U.S. scholars on labor. Freeman opens the explanatory door one step beyond the conventional education-skill story (to which he otherwise subscribes). He finds that the long-run decline of union density in the United States--the proportion of workers belonging to unions or covered by a collective bargaining agreement--has contributed to the growth in earnings inequality, both among and within industries. Unions, he writes, have reduced wage dispersion in every country and time period for which data are available. Thus, we should not be surprised that steadily declining union membership in this country contributes to growing inequality. But, he hastens to add, declining unionization is only one of many developments promoting inequality, "a supporting player in the story. . . not the main character: Rosencrantz or Guildenstern, not Hamlet."
HOW LEAN PRODUCTION AND NETWORKED FIRMS PROMOTE THE POLARIZATION OF WAGES: THE NEW DUALISM
The decade between 1965 and 1975 was a golden age for mavericks in economics, as it was for the rest of U.S. society. Out of that period emerged a school of political economists who devoted themselves to examining the evolution of the organization of work over the course of the twentieth century. Trained at the leading graduate schools, yet grounded in practical work in the civil rights movement and in urban community development, these young economists eventually produced a body of ideas about "dual," or segmented, labor markets. It turns out that these old ideas have much to say about the new forms of flexible production, in general, and about the surge in inequality, in particular.
In brief, a "primary labor market" was said to be dominated by large, vertically integrated companies that, over the course of the century, had acquired some degree of oligopolistic power and operated in markets that were relatively sheltered from foreign competition. These corporations earned above-average profits, and their employees--especially if they were organized into labor unions--earned above-average wages. In the interest of achieving stable labor relations and promoting on-the-job learning, and in interaction with unions seeking job security for their members, managers created within their firms vertical career ladders, so-called internal labor markets (ILMs). Workers in this segment of the economy could progress up through the rungs of these ladders, learning by doing, under the watchful eyes of supervisors who would periodically evaluate their progress as part of a complex administrative process of "bureaucratic control."
By contrast, outside the boundaries of these generally big firms lay a world of mostly smaller enterprises. Often, these firms acted as subcontractors to the big firms in the core of the economy. Unsheltered from intense price competition with one another, these firms formed the principal actors within the so-called secondary labor market. Here, production tended to be more standardized, wages were lower, new employees could be more easily recruited off the street, employee benefits were limited or nonexistent, and opportunities for much valuable on-the-job learning were generally hard to find.
That was then; this is now. Today, business is operating in a world in which best practice increasingly entails vertical disintegration, downsizing, outsourcing, and the formation of networks of companies in order to operate across national borders and sectoral boundaries. It would be ludicrous to suggest that all of the features of post-World War II labor markets have somehow miraculously survived intact.
And yet, when we characterize the prototypical business organization of the new era as a lean and mean flexible firm, embedded within networks made up of partners and dependent suppliers and subcontractors, we are implicitly recognizing that the workforce in the new economy is, arguably by managerial intent, being systematically divided into insiders and outsiders. Some people are employed on full-time, year-round schedules, receive health insurance and paid vacations, and experience continuous formal and informal job training. A few can even still look forward to having access to structured, more or less predictable opportunities for upward mobility, although in an era of flexible production networks, moving "up" more often means changing jobs by moving from one company to another.
But in this evolving system, at least as many people now confront job "opportunities" that, increasingly, are limited to involuntary part-time or part-year work, low wages, few benefits, and frequent job changing that fails to provide a rising standard of living. This is life on a treadmill. Smaller firms that act as suppliers to the big firms are being similarly stratified, according to the degree to which they collaborate with their large industrial customers and are technically or financially supported by those.
To be sure, some part-time workers choose that form of employment, and some fraction of those who do are well paid for it. This is especially the case for professional white women who work, for example, as freelance editors or computer programmers. But one of the most knowledgeable experts on these changes in labor market institutions, the economist Chris Tilly, of the University of Massachusetts at Lowell, concludes from his extensive econometric and ethnographic research that "secondary, 'bad' part-time jobs greatly outnumber ['good'] part-time jobs." How does he know? Because "of the 2.9 percentage point climb in the rate of part-time employment between 1969 and 1988, 2.1 points were due to increasing involuntary part-time employment."
The increasing polarization of wages is clearly consistent with the reinstitutionalization of labor market dualism. Indeed, several European scholars have already spied this connection. More than a decade ago, the Oxford University sociologist John Goldthorpe suggested that corporations' search for flexibility could actually heighten labor market and social duality. Another British commentator who recognized early the reinstitutionalization of dualism inherent in the core-ring or network transformation of business organization was John Atkinson. Working out of the corporate-sponsored British Institute for Manpower Studies, Atkinson predicted that the largest multinational firms would attempt to retain their control over global economic affairs by transforming themselves in such a way as to create a new and far more sophisticated set of hierarchies and labor market segments, both within the core company itself and outside the firm, along its network of suppliers of products and labor. Indeed, as a management consultant, Atkinson was advocating the reproduction of such dualities, in the interest of promoting his clients' very survival.
Several years later, building explicitly on Goldthorpe's reassessment but coming at the problem from the left of the labor movement, the University of Milan sociologist Marino Regini conceived what he termed the "new dualism." Across the world, argued Regini, companies were combining "informalization" (mainly by relying increasingly on outsourcing) with what he called "Japanization," the introduction of teamwork and just-in-time management within the core labor processes of the leading firms.
Meanwhile, back in the United States, the Stanford University sociologists Jeffrey Pfeffer and James Baron may have been the first Americans to state that "the increasing separation of the workforce into [core] and [peripheral] workers may increase wage inequality, as workers who are considered to be more essential to the firm will tend to be paid more, while those used as buffers will have less market power and obtain lower wages . . . . We suspect [this] inequality may be based increasingly on hours worked and the stability of employment, the dimensions which [most] distinguish core and buffer employees."
Thus, it seems, while the economists who had first invented dual, or segmented, labor$theory were occupied elsewhere, the sociologists and planners were telling us that the institution was alive and well, embedded within the very flexible production systems that were supposed to be replacing the old order.
THE DEVOLUTION OF INTERNAL LABOR MARKETS AND THE EROSION OF EMPLOYMENT SECURITY
The new dualism differs from the original version in at least one important way: Even high-level jobs are no longer secure. In the age of flexibility, even the most profitable big finns are inclined to shed even white-collar employees. The shrinkage in the number of safe, stable, secure occupations seems to be keeping pace with the big finns' downsizing.
One consequence of these practices is that, in sharp contrast to their Japanese and German competitors, U.S. managers are coming to devalue internal labor markets (ILMs) as organizational mechanisms for developing loyalties and productive skills. The decline of ILMs simultaneously marks an important change in the particular character of labor market dualism and constitutes one of the sources of that growing polarization of earnings about which so many raised so much fuss during the 1980s.
For a number of years, the urban planner Thierry J. Noyelle, of Columbia University, has been studying what he explicitly refers to as the "dismantling" of ILMs, especially by companies in the retailing, public utilities, and financial sectors. His argument, in a nutshell, is that the rapid growth of a highly educated urban workforce, together with successful equal employment opportunity laws and programs, has made it possible for managers to "externalize" much of the skill training they used to conduct, mostly informally, in-house. The advent of information technologies, argues Noyelle, complements this trend by leading to "a kind of universalization or homogenization of skills demanded across a wide range of industries, allowing for [even] greater externalization of training for many middle-level workers. Many occupations have now become more generic and less firm specific than they once were . . . [and] computer-oriented algorithmic logic has replaced many firm-specific idiosyncratic practices. [All of this amounts to a pronounced shift] away from on the job training [that has] further undermined the raison d'etre of the old internal labor market."
The growth of an educated, more diverse urban labor force and the standardization of so much office and other white-collar work through what Noyelle calls "algorithmic logic" are permissive developments, while the corporate search for flexibility and cost savings has been the driver. In any case, these are complementary forces, and their net result is the decline of the ILM as an organizational solution to the problem of training.
One serious methodological problem that has plagued policy research on labor market structure is that what is being observed in the data are people (workers, families), when what is often wanted are observations on jobs. The economists Maury Gittleman of New York University and David Howell of the New School for Social Research have made a considerable contribution in this respect. In their operationalization of these institutional constructs, they also provide strong support for the hypothesis that it is indeed those middle-level jobs--and, with them, middle-class incomes--that have been disproportionately shrinking in recent years.
Using data originally constructed at the NBER by Katz and his fellow Harvard economist (and Clinton administration official) Lawrence Summers, Gittleman and Howell have studied a matrix of 621 different combinations of industries and occupations, drawn from the 1980 Census. In this way, they have been able to distinguish, for example, between custodians working for the automobile industry, where they tend to be well paid, and custodians employed in fast-food restaurants, where they are not. It is these interactions that have enabled Gittleman and Howell to inform us that about half of all high-wage blue-collar workers in 1980 did not work in manufacturing, and that more than two-thirds of all low-wage blue-collar workers were employed in services. This ability to finally break the false (but popular) identity between manufacturing and blue-collar work is a major achievement.
These 621 combinations accounted for fully 94 percent of total nonagricultural employment in the United States in 1980. Each of these "jobs" is then characterized along seventeen different dimensions, from average hourly wage rates to the extent of involuntary part-time employment to employer-assessed physical strength requirements for performing the job. The application of a standard statistical technique, cluster analysis, to this 621-cell matrix has led the researchers to identify in the data a latent structure according to which jobs fall into one or another of six categories.
The two categories whose jobs generally pay the highest wages, offer the greatest opportunities for on-the-job training, and so on are "professional, managerial, and sales," and "public sector." The two categories that generally pay the lowest wages and offer the fewest employee benefits are "low-wage blue-collar" and "contingent" (part-time, casual, part-year employment). Sandwiched in between are two categories consisting of "routine white-collar" and "high-wage blue-collar" jobs. This empirically derived (rather than assumed or imposed) taxonomy extracted from 1980 data looks remarkably like the labor market structure discovered by the economists David M. Gordon, Richard Edwards, and Michael Reich from historical data on the post-World War II era, which only reinforces the impression of continuity with the past.
Comparing the distributions of actual employees in 1979, 1983, and 1988 against these benchmarks, Gittleman and Howell conclude, "The job structure has become more bifurcated in the 1980s, as 'middle class' jobs [the share of the workforce employed in those middle two categories] declined sharply and the workforce was increasingly employed in either the best. . . or the worst. . . jobs. White women became much more concentrated at the top, while white men and black and Hispanic women were redistributed to both ends of the job structure. Black and Hispanic men, however, increased their presence only in the two [lowest] job clusters. At the same time, the quality of [these lowest category] jobs worsened considerably, at least as measured by earnings, benefits, union coverage, and involuntary part-time employment. As these results would suggest, we found that earnings differentials by cluster, controlling for education and experience, increased in the 1980s. The male-female racial wage gap also increased."
THE GROWTH IN CONTINGENT WORK
Richard Belous of the private corporate-sponsored National Planning Association (NPA) estimates that in 1988 the number of workers employed in part-time, temporary, contract, and other forms of what has come to be called "contingent" labor in the United States was between 30 and 37 million--roughly a quarter to a third of the civilian labor force. Research on the changing incidence of contingent work was, for many years, the province of a small group of scholars and labor unions (one of those scholars, the economist Katherine Abraham, is now the commissioner of the U.S. Bureau of Labor Statistics [BLS]). At the Washington-based Institute for Women's Policy Research, Heidi Hartmann and Polly Callaghan have documented that, since 1982, temporary employment in the United States has grown three times faster than employment as a whole (Exhibit 1). Between 1976 and 1990, the number of part-timers (defined as those averaging fewer than 35 hours of waged employment per week when they did work) increased by 7 percent, compared to a 2 percent growth of full-time workers--those usually employed for 35 hours a week or more.
One study that has been enormously influential in shaping the public debate about contingent work was conducted at Cornell University's School of Industrial and Labor Relations by the labor economist Ronald Ehrenberg and his students. Using econometric methods, Ehrenberg's team was able to show "there has been a tendency towards increased employment of part-time workers in the United States in recent years, a trend that is observed after one controls for cyclical factors. Moreover, this trend has come from an increase in 'involuntary' part-time employment, not from an increase in voluntary part-time employment. Searches for explanations for the recent growth of part-time employment in the U.S. should therefore focus on the demand side of the labor market."
"Demand side of the market" means the decisions of managers. The explosive growth of involuntary part-time work, compared with full-time employment, is graphed in Exhibit 2. These developments have serious implications for personal well-being. For example, between 1979 and 1989, the share of the private-sector workforce covered by pension plans fell from 50 percent to 43 percent. The incidence of employee coverage by health insurance at least partly provided by employers declined from 69 percent to 61 percent. These are very substantial declines for such a short period of time.
Exhibit 3 shows just how much the average decline in benefit coverage is wrapped up with the shift toward lean production and contingent work. As depicted in that exhibit, the proportion of voluntary part-timers without health insurance has actually been declining since 1983, while a steadily rising share of involuntary part-timers have no health insurance. This growing division of the U.S. workforce into those who are entitled to employee benefits and those who are not is one of the more unfortunate aspects of the growth of insider-outsider employment arrangements--truly a dark side of flexible production.
In the late 1980s, government agencies and business research institutes began to take the phenomenon seriously. Thus, in 1989, the NPA published a collection of case studies authored by Belous. In the same year, two economists at the Bureau of Labor Statistics announced a semiofficial definition of contingent work as including "any job in which an individual does not have an explicit or implicit contract for long-term employment and one in which the minimum hours worked can vary in a non-systematic manner." We can expect more such research and standard setting from the Clinton administration, with Abraham now at the helm of the BLS, and with Karen Nussbaum directing the Labor Department's Women's Bureau. (Nussbaum was the cofounder of 9-to-5, the clerical workers' association that was eventually absorbed into the Service Employees International Union [SEIU]. More than any other organization, SEIU has spoken up on the subject of contingent work for more than a decade.)
Other government agencies have been looking more closely at still other dimensions of contingency. In one of the more gruesome, the General Accounting Office (the investigative arm of Congress) reported in June 1992 that the number of illegally employed minors--children under the age of 14--had almost tripled since 1983. William Halperin, then the associate director for surveillance at the National Institute for Occupational Safety and Health, was quoted as characterizing these findings as "astounding," yet probably only "the tip of an iceberg."
The real question remains whether managers are consciously creating (or re-creating) dual labor markets by deliberately externalizing and otherwise transforming work. At least in the United States, the answer is yes. U.S. companies are deliberately creating both peripheral and internal low-wage labor markets for employees on contingent work schedules.
This conclusion comes from a 1988 survey of American managers conducted by the Conference Board, the New York-based think tank for the Fortune 500. The survey was conducted with representatives of 521 of the country's largest manufacturing, financial, and nonfinancial services corporations. Respondents were asked to describe their use during the previous year (1987) of internal pools of temporary workers, of outside agencies that provide "temps," of independent contractors (for example, editors or computer programmers), of so-called flextime arrangements, of regular part-time employees, of job-sharing arrangements among employees, of compressed work weeks, of phased retirement, and of home-based work.
From the Conference Board data, Kathleen Christensen discovered that a significant fraction of the firms sampled did use part-time labor as a deliberate "contingent staffing alternative," along with outside temps, independent contractors, and internal temp pools. In some companies, "firms hire as independent contractors former employees who have been laid off, taken early retirement, or left the firm to go out on their own"--a practice that is also increasingly common in such jobs as equipment maintenance in steel mills. This sort of outside contracting is common both in manufacturing and elsewhere in the economy. According to Christensen, nearly 100 percent of the finance and insurance companies surveyed reported using part-time workers as a regular activity. Managers also told the interviewers that they expected to increase their use of contingent work schedules in the future.
DUALISM IN THE NIKE PRODUCTION SYSTEM: A CASE STUDY
Dual labor markets--the dark side of flexibility--can exist even within the boundaries of a single firm, especially when that firm is organized as a geographically extensive network with its core operations at one location and tiers of parallel or lower-level activities at other sites. For a U.S.-based example, take Nike.
No product sold in America (or Europe or the Far East) today is more well-known than Nike's running shoes. Like cars, hamburgers, and furniture, Nike shoes (the name Nike stands for the goddess of victory in Greek mythology) come in literally dozens of models, from the relatively cheap to the very expensive, from the (literally) pedestrian to the stylish. Although Nike is legally registered as a U.S. corporation, not one of the 40 million pairs of running shoes that Nike produces annually is manufactured within the United States: Everything is subcontracted from elsewhere.
As reconstructed by Michael T. Donaghu and Richard Barff, two Dartmouth College geographers (one of whom worked for the company for a time), Nike's global division of labor is an exemplar of the principle of what I call concentration without centralization. That principle is intended to capture the continuing dispersal of production, but ultimately under the technical and financial control of managers in a relatively small number of big multiregional, multisectoral, multinational corporations and their strategic allies. Nike is such a network firm, whose management has found ways to connect the lowest-paying unskilled jobs with the highest-paying skilled R&D jobs, classic mass production with flexibly automated technology, and the First World with the Third.
Bill Bowerman and a partner founded Blue Ribbon Sports (BRS)-- the company that would later become Nike--in 1964. Initially, BRS was purely a U.S. distributor of athletic shoes made in Japan by Onitsuka Tiger. Bowerman was a good designer, and the product began to capture the growing domestic U.S. market as the public's interest in stylish running shoes grew. In 1971, BRS's relationship with its first Japanese partner ended, and the Americans entered into an agreement with Nisso-Iwai, a large trading company with access to its own manufacturing contractors--again, inside Japan. Research and development took place in BRS's facility in Beaverton, Oregon, under Bowerman's supervision. Production remained in Japan, under the management of the Nippon Rubber Company.
In the wake of the first oil shock of the 1970s and a major revaluation of the yen, Nippon Rubber decided to relocate much of its production operations from Japan to Taiwan and South Korea (just as the high exchange value of the yen of the early 1990s is now driving many Japanese companies to move production offshore). BRS itself opened an assembly facility in Exeter, New Hampshire, sourcing components from U.S. and German suppliers. Within a short time, this network of factories was turning out a wide range of products, from inexpensive mass-market footwear to stylish running shoes. By the late 1970s, BRS was marketing its products in Europe, South America, Southeast Asia, and the United States. BRS factories in Britain and Ireland came next. In 1978, BRS legally changed its name to Nike.
The Nike production system is organized into two broad tiers. In the first of these tiers, "developed partners" located mainly in Taiwan and South Korea work closely with the R&D personnel in Oregon to make the firm's most expensive, high-end footwear. (By partners, Nike managers mean contractors and suppliers who share some joint responsibility with the core firm, for design or for evaluation of production methods.) The Asian partners contract out most of the work to local low-wage subcontractors.
"Volume producers" are considerably larger, more vertically integrated companies with their own leather tanneries and rubber factories that manufacture more standardized products and sell to several buyers, of whom Nike is only one. Production and sales are highly variable from one month to the next.
Finally within this tier, Nike has created what it calls "developing sources"--producers located in Thailand, Indonesia, Malaysia, and China. These are the lowest-wage, low- and semiskilled operations that Nike is gradually upgrading ("bringing along"). Technicians from the United States, Taiwan, and South Korea are assigned to work in these "developing" facilities, often on a rotating basis (Nike calls this its "expatriate program"). Such sources often take the form of joint ventures between the Taiwanese or South Korean "big brothers" and the less industrialized Southeast Asian "little brothers."
Nike managers explicitly acknowledge the advantages of this spatial division of labor within the first tier. According to Donaghu and Barff, apart from "hedg[ing] against currency fluctuations, tariffs and duties [and] political climate change...in the long run [this arrangement] keeps pressure on the first tier producers to keep production costs low as developing sources mature into full-blown developed partners."
The second tier of the Nike production network consists of the many material, component, and subassembly sources. Predictably, the least complicated elements may be produced at any of the network's locations in the United States, Europe, or Southeast Asia. But some elements, such as the air cushions that pad all modern athletic shoes, either require skilled labor to turn them out or make use of proprietary technology or designs. These tend to be located in the vicinity of the Oregon headquarters, where engineers and others can be exchanged easily and frequently--an example of "specialization subcontracting." As the physical infrastructure and human capital of the "little brothers" is upgraded over time, Nike anticipates that a growing fraction of its second-tier suppliers will be located in Asia.
Clearly, the Nike production system has undergone a substantial evolution since the 1970s. Yet it remains a dualistic system, combining high-wage and low-wage, specialized and standardized production, core and periphery. As Donaghu and Barff put it, "When Nike's shoes were first produced in Asia, Japan represented the core and other nations constituted the periphery. Today, Nike's intraregional division of labor in South East Asia simply based on labor costs has S. Korea and Taiwan as the core and China, Thailand, Malaysia, and Indonesia as the periphery. Japan, with average manufacturing hourly wages very close to those of the United States, is still a supplier of materials, but is no longer a site of production for Nike athletic shoes.... [However,] there is no evidence to suggest that production has shifted as markets have developed. In fact, there is evidence to the contrary. The English, Irish, and American plants all ended production of Nike athletic shoes in the mid 1980s, even as demand for Nike running shoes was accelerating in all three countries."
Interviews with the U.S. managers revealed that they clearly perceived the flexibility of this production system as residing in a combination of low cost, spread risk, and--most of all--"the speed by which a design for a new model of shoe is transformed into a product at the market." This turnaround time is partly a function of using highly disciplined labor and partly of deploying "mass production techniques by simplifying work stations and updating existing machinery." For example, note Donaghu and Barff, "a new machine was introduced that could automatically position the needle and trim excess thread for the operator. This kind of labor saving change is hardly revolutionary and can be specifically associated with the continuation of Fordist means of production. The only elements of Nike's system that could be classified as flexible machinery [are] the computer-aided design and computer-aided engineering used in the Beaverton [U.S.] R & D facility, and some numerically-controlled molding machines used by one or two South Korean subcontractors."
The other aspect of flexibility in this system is the core firm's adeptness at shifting productive capital from one place to another. Nike "has opened plants and begun contracts only to end them within only a matter of a year or two. It also utilizes capacity subcontracting methods to meet variable market demand."
Throughout the first tier, as much as 80 percent of sales to the final assemblers of the footwear are "prepurchased," meaning that Nike effectively advances capital to the members of what, after all, clearly has features that resemble a far-flung putting-out system. And while Nike's subcontractors themselves tend to be vertically disintegrated and may in some cases be locally agglomerated, and while customer-supplier relations may sometimes be described as reciprocal, or at least not confrontational, "by any definition, [Nike's] method of physically producing athletic shoes is mass production." At the level of the global network, "the company still relies on large volume production by [mostly] semi- and unskilled labor," linked to high-tech R&D and sophisticated financial management situated in the United States.
Nike is a neo-Fordist firm, whose flexibility derives mainly from its managers' ability to construct and govern a dualistic system characterized by concentration without centralization. Its profitability derives directly from its managers being so cleverly able to manage the dark side.
PLAYING OFF INSIDERS AGAINST OUTSIDERS
To recapitulate: Collaboration among producers within networks may well help business to cope with the uncertainties and the heightened competition that are part and parcel of the post-1970s global economy. Yet at the same time, vertical disintegration of the older big firms, the devolution of sheltered internal labor markets with their opportunities for lifetime employment, and the increasing use of outside subcontractors standing in varying degrees of dependence and independence vis-a-vis the core firms in the network all sharpen the divisions between insiders and outsiders. This in turn reinforces the long-run trend toward the polarization of U.S. earnings.
The danger is that, instead of promoting collaboration among the more and less powerful organizational actors within a network--and, for that matter, among managers and workers within individual companies and establishments--the managers governing the system may yield to the temptation to use their power to exploit these differences in relative power to play one group within the network off against another. Thus, for example, in acknowledging the recent growth within General Motors of various forms of teamwork, the Harvard Business School's Joseph Badaracco reminds us that "to stress only the new and important participatory elements of [General Motors'] relationship with the [United Auto Workers union] is an oversimplification. Greater cooperation has arisen against the background of a shift in power away from the UAW and toward GM. Indeed, the movement toward greater participation for workers and unions over the last ten years has been paralleled by another, perhaps even more widespread effort to limit their power through overseas sourcing, givebacks, and manufacturing strategies that expand facilities in nonunion states."
Another of the leading theorists of the flexible network firm, the University of Arizona sociologist Walter Powell, is equally cautionary. "Practices such as subcontracting," he notes, "have a double edge to them; they may represent a move toward relational contracting ... with greater emphasis on security and quality; or they could be a return to earlier times, a part of a campaign to slash labor costs, reduce employment levels, and limit the power of unions even further." Moreover, "some firms are seeking new collaborative alliances with parts suppliers while at the same time they are trying to stimulate competition among various corporate divisions and between corporate units and outside suppliers.... Are companies really as confused as it seems?"
Powell's answer is, perhaps not. Perhaps these contradictory combinations of integration and disintegration, of collaboration and competition, of coherence and duality are built into the very nature of how network forms of organization grow and develop. If so, he acknowledges, this poses an extraordinary challenge for those who must live with such a system.
Certainly, the revival of labor market segmentation further weakens the bargaining power of labor unions, making it more difficult for them to organize new workers and to pressure companies to innovate continually in order to generate the additional productivity out of which to meet a rising wage bill. This is the "high road" to economic growth. From the perspective of the national--perhaps even the international--- economy as a whole, growing income stratification between capital and labor, and the growth of a pool of low-wage workers act as a drag on systemwide technological progress and, therefore, on long-run economic growth. The reason is that dualism encourages all too many firms to build their activities on a foundation of cheap labor, thereby taking the "low road" to company profitability. This may ultimately be the most serious consequence of the dark side of flexible production and business reorganization.
THE HIGH ROAD OR THE LOW ROAD TO LONG-RUN ECONOMIC GROWTH?
In the United States of the 1990s, the average worker brings home a paycheck that, depending on how you measure "average" and how you account for the ravages of inflation, is anywhere from 7 percent to 12 percent lower than what it was at the end of the Vietnam War, a generation ago. The rich have been getting richer and the poor, poorer. And the great middle class is at best treading water, if not actually sinking below the surface of the decent life that our leaders and teachers had led them to expect as the reward for hard work. Once, these facts about the secular stagnation and polarization of the U.S. economy were hotly contested. Today, hardly any reputable scholar would disagree. In popular journalism, stories of the two-tiered society and the hourglass economy now appear regularly on the nightly television news and in the newspaper Sunday supplements.
Secularly increasing inequality raises obvious normative concerns: Who wins? Who loses? Who is most in need? But there are also macroeconomic considerations that motivate the earnings dispersion debate. The French economist Alain Lipietz is but one among many to suggest that the impaired capacity of a growing fraction of the workforce to consume out of current income could create problems for effective aggregate demand, were the trend sustained for a sufficient period of time and were it not offset by government deficits or household dissaving. Indeed, in the United States during the 1980s, short-term aggregate demand was sustained despite growing inequality, partly through the accumulation of more than $2 trillion in added federal government debt and close to $500 billion in additional consumer credit.
But beyond the macroeconomic impact, there are negative microeconomic consequences flowing from the revival of inequality in general, and of labor market dualism in particular. A growing pool of low-wage labor sends precisely the wrong signal to firms, encouraging them to compete on the principle of cheap labor rather than on the basis of technological improvement and the upgrading of their employees' skills. As the British economists S. Deakin and Frank Wilkinson explain: "Dependence upon undervalued labor provides a way by which inefficient producers and obsolete technologies can survive and compete. Firms become caught in low productivity traps from which they have little incentive to escape." Moreover, the process may become viciously circular, since, having deprived themselves of a technically competent workforce to begin with, these firms are subsequently ill-equipped to innovate. Hence, "when these firms are subjected to competition from more efficient [companies], improved technology and products, their only hope of survival is further to reduce wages."
New research from Richard B. Freeman and Lawrence F. Katz suggests that inequality and the erosion of wage standards may now be occurring in countries besides the United States, although to very different degrees and with less pernicious implications, given their generally stronger social safety nets. Nevertheless, in both Japan and Europe, over the course of the post-World War II era, growth-oriented companies with long-term planning horizons worked together with governments committed to the practice of industrial policy. Especially in Europe, organized labor's demands for higher wages and better working conditions, and its progrowth stance in the political arena, contributed to pressing big business and the government to take the high road to economic growth and development. Along this path, companies invest continuously in their employees' skills and in manufacturing, office, warehouse, highway, railroad, and aircraft equipment, embodying new technologies. The bigger firms help to upgrade the technical capabilities of their generally smaller suppliers. Through this combination of technology, training, and technical assistance, the productivity of the national economy increases--and, with it, the standard of living of the mass of the population.
But there is also that other, less admirable path: the low road to company profitability. Along this path, managers try to beat out the competition by cheapening labor costs. They move whatever operations they can to low-wage rural areas or to Third World countries. They scrimp on training. They routinely outsource work that used to be performed in-house to independent subcontractors who will not (usually because they themselves cannot afford to) pay decent wages, let alone provide even the most basic benefits such as health insurance premiums or paid sick leave. Low-road companies try to squeeze the last ounce out of older capital equipment, rather than steadily retooling and upgrading their technical capabilities. They play off their suppliers against one another to get the cheapest price today, with no thought to the negative impact this can have on the quality of tomorrow's deliveries. At the last extreme, a company that once made its own products, using domestic workers and paying them a living wage, now hollows itself out, abandoning manufacturing altogether to become more or less strictly an importer of things made by foreign companies--or by their own overseas subsidiaries.
Throughout the twentieth century, U.S. business has been at war with itself over whether to travel the high road or the low one. Since the 1970s, the tension between these alternatives has become even more pronounced. The consequence of a generation of managers taking the low road to a restoration of profits is the cultivation of the habit of competing mainly on the basis of cheapening labor power, rather than upgrading technology and skills. Not all firms in all U.S. industries are pursuing a low-road strategy. But the evidence that a large number are doing so seems so compelling that we should worry about the future prospects for a restoration of the historic U.S. economic pattern of growth at high wages with declining inequality.
Why have so many U.S. (and, increasingly, foreign) managers elected the low road to resolving the profit squeeze? Surely the current weakness of the U.S. labor movement is part of the answer. Drawing on data from the BLS, Freeman and the University of Wisconsin political scientist, Joel Rogers, observe that the fraction of private sector workers in the United States who belong to unions has fallen to a pre-1935 low of 12 percent, and that a rate of just 5 percent is a reasonable forecast for the year 2000. The problem is compounded by the absence of any other structures of worker representation that could compensate for the decline in union density. A strong (or, as in the 1930s, re-emerging) trade union movement effectively forces employers to make decisions that enhance productivity in order to contain unit labor costs.
But what explains the weakness of U.S. labor unions and the upsurge in corporate restructuring strategies? The answer lies in the sheer suddenness with which the U.S. economy found itself inserted into the international trading system in the 1970s. Between 1969 and 1979, the share of U.S. gross national product (GNP) accounted for by imports doubled, and merchandise imports as a share of total GNP originating in the manufacturing sector nearly tripled. The old oligopolistic mechanisms available to the leading U.S. corporations for absorbing cost increases by raising their price markups were, thus, seriously undermined by the threat (and the reality) of foreign competition.
Add to this the continually fluctuating exchange rates, which created chronic uncertainty and wreaked havoc with investment decisions, and a portrait emerges of an environment in which it was simply easier for many firms to attempt to contain their own labor costs than to seek enhanced profits through investments in expensive new plants and equipment. Too many U.S. firms abandoned revenue-enhancing strategies to boost profits and turned sharply toward tactics that emphasized cost reduction, instead.
The growing dependence of productive enterprise on equity financing, managed by Wall Street intermediaries, contributed to the already developing tendency in U.S. industry for decisions to be made with an eye toward short-run profit. In contrast to industries in Japan, Sweden, and Germany--where the merchant banking systems have been so fundamental to the financing of long-term capital projects--U.S. industry has become more and more the prisoner of impatient capital. The surge in technologically or synergistically unsupportable mergers and acquisitions, along with rank speculation in land, currencies, and futures markets, was made that much more extreme in the country whose institutions were least well suited to providing a counterweight.
Finally, reinforcing all of these developments have been three decades of high real interest rates, with relief having come only in the 1990s. The Vietnam era spawned ever more powerful pressures for future inflation, as the Federal Reserve eased the money supply to validate competing claims on real resources in an effort to head off further political turmoil at home. The attempts by unions and other organized groups during the 1970s to maintain the living standards of their members in the face of this inflationary bias eventually convinced the government, even before the election in 1980 of Ronald Reagan to the presidency, that a tight monetary policy was essential to containing inflation and to weakening the claims of labor and of the social movements.
But the monetarist experiments, first in the United States and then in the United Kingdom, only depressed first national, then international economic growth. That, in turn, suppressed whatever private investment in new plant and equipment might otherwise have been forthcoming. Now, in the 1990s, synchronized recessions in Japan and Europe have only made matters worse, by greatly weakening foreign demand for American exports.
Consequently, influenced by slow growth of demand and high real long-term interest rates throughout most of the period, the annual rate of growth of industrial capacity has fallen steadily in this country, from 3.5 percent per year in the early 1980s to about 1.5 percent per year in 1992. About this trend, the chief economist of the Wall Street investment banking firm of Morgan Stanley and Company, Stephen S. Roach, has expressed the concern that "smokestack America may have gone too far in hollowing out its industrial base in order to achieve short-term efficiency gains."
This is a particularly vicious circle, since slowly growing industrial capacity means that even a modest spurt of economic growth will quickly push up rates of capacity utilization. To the extent that this gets read by the always jittery Federal Reserve as a signal of impending inflation, the central bankers will be inclined to step on the monetary brakes, slowing the economy further. In this way, excessively cautious macroeconomic policy reinforces the signals to business to opt for the short run-- to go for the low road.
WHITHER THE DARK SIDE?
This is a conundrum--what economists call a low-level equilibrium trap--that the search for more "flexible" production and working arrangements on the part of individual companies, their suppliers, and their strategic partners seems only to be making worse. What is needed is a combined effort by a critical mass of farsighted private businesses, a government able and willing to overcome its own equally shortsighted fear of budget deficits and to deliver serious domestic macroeconomic stimulus, and a much greater degree of international coordination of fiscal and monetary policies among the G-7 nations, that is, the United States, the United Kingdom, Japan, Canada, Germany, France, and Italy.
Without such interventions, we run the very grave danger of finding ourselves trapped on the dark side for years and years to come.
Atkinson, John. "Recent Changes in the Internal Labour Market Structure in the U.K.," in Technology and Work, ed. Wout Buitlaar (Aldershot, U.K.: Avebury, 1988), pp. 133-49.
Badaracco Jr., Joseph L. The Knowledge Link (Boston: Harvard Business School Press, 1991).
Badaracco Jr., Joseph L. "Changing Forms of the Corporation," in U.S. Business Corporation: An Institution in Transition, John R. Meyer and James M. Gustafsom (Cambridge, Mass.: Ballinger, 1988), p. 84.
Belous, Richard. The Contingent Economy: The Growth of the Temporary, Part-Time and Subcontracted Workforce (MacLean, Va.: National Planning Association, 1989).
Bluestone, Barry and Irving Bluestone. Negotiating the Future: A Labor Perspective on American Business (New York: Basic Books, 1992).
Bluestone, Barry and Bennett Harrison. The Deindustrialization of America (New York: Basic Books, 1982).
Bluestone, Barry and Bennett Harrison. The Great American Job Machine: The Proliferation of Low Wage Employment in the U.S. Economy, report prepared for the Joint Economic Committee, November 1986.
Bowles, Samuel, David Gordon, and Thomas Weisskopf, Beyond the Waste Land (New York: Anchor, 1983).
Burtless, Gary. "Earnings Inequality over the Business and Demographic Cycles," in A Future of Lousy Jobs? ed. Burtless (Washington, DC: Brookings Institution, 1990), pp. 77-122.
Christensen, Kathleen. "The Two-Tiered Workforce in U.S. Corporations," in Turbulence in the American Workplace, ed. Peter B. Doeringer with Kathleen Christensen, Patricia M. Flynn, Douglas T. Hall, Harry C. Katz, Jeffrey H. Keefe, Christopher J. Ruhm, Andrew M. Sum, and Michael Useem (New York: Oxford University Press, 1991).
Deakin, S. and Frank Wilkinson. "Labour Law, Social Security, and Economic Inequality" (London: Institute of Employment Rights, February 1989), p. 44.
Donaghu, Michael T. and Richard Barff. "Nike Just Did It: International Subcontracting and Flexibility in Athletic Footwear Production," Regional Studies 24 (December 1990): 537-52.
Freeman, Richard B. "How Much Has De-unionization Contributed to the Rise in Male Earnings Inequality?" in Danziger and Gottschalk, Uneven Tides, p. 159.
Freeman, Richard B. and Lawrence F. Katz. "Rising Wage Inequality: The United States vs. Other Advanced Countries," in Freeman, ed., Working under Different Rules.
Freeman, Richard B. and Joel Rogers. "Who Speaks for Us? Employee Representation in a Non-Union Labor Market," in Bruce E. Kaufman and Morris M. Kleiner, eds. Employee Representation: Alternative and Future DIrections (Madison, Wis.: Industrial Relations Research Association, 1993), pp. 13-79.
Gittleman, Maury B. and David R. Howell. "Job Quality, Labor Market Segmentation, and Earnings Inequality: Effects of Economic Restructuring in the 1980s by Race and Gender" (Graduate School of Management, New School for Social Research, New York City, July 1992, unpublished MS).
Goldthorpe, John. "The End of Convergence: Corporatist and Dualist Tendencies in Modern Western Societies," in Order and Conflict in Contemporary Capitalism, ed. Goldthorpe (Oxford: Clarendon Press, 1984), pp. 315-43.
Karoly, Lynn A. "The Trend in Inequality among Families, Individuals and Workers in the United States: A Twenty-Five Year Prospective" (Santa Monica, Calif.: RAND Corporation, 1990), reprinted in Uneven Tides: Rising Inequality in America, ed. Sheldon Danziger and Peter Gottschalk (New York: Russell Sage, 1993), pp. 19-97.
Katz, Lawrence and Kevin Murphy. "Changes in Relative Wages, 1963-1987: Supply and Demand Factors," Quarterly Journal of Economics 107 (February 1992): 35-78.
Kelley, Maryellen R. and Bennett Harrison. "Unions, Technology, and Labor-Management Cooperation," in Unions and Economic Competitiveness, ed. Lawrence Mishel and Paula B. Voos (Armonk, N.Y.: Sharpe, 1992), pp. 247-86.
Kolata, Gina. "More Children Are Employed, Often Perilously," The New York Times, June 21, 1992, p. 1.
Koretz, Gene. "U.S. Industry May Be Mean, But It Also Looks Too Lean," Business Week, June 14, 1993, p. 22.
Kuttner, Robert. "The Declining Middle," The Atlantic, July 1983, pp. 60-69.
Harrison, Bennett and Barry Bluestone. The Great U-Turn (New York: Basic Books, 1988).
Lardner, James. "The Declining Middle," The New Yorker, May 3, 1993, pp. 108-14.
Levy, Frank and Richard Murnane. "U.S. Earnings Levels and Earnings Inequality: A Review of Recent Trends and Proposed Explanations," Journal of Economic Literature 30 (September 1992): 1,333-81.
Lipietz, Alain. "New Tendencies in the International Division of Labor: Regimes of Accumulation and Modes of Regulation," in Production, Work, Territory, Allen Scott and Michael Storper (London: Allen & Unwin, 1986), pp. 16-40.
Mishel, Lawrence and Jared Bernstein. The State of Working America, 1992 (Armonk, N.Y.: Sharpe, 1993), p. 146.
Nasar, Sylvia. "U.S. Industry's Sluggish Pace: Revision Finds Less Growth of Capacity," The New York Times, May 15, 1993, p. 17.
Pfeffer, Jeffrey and James Baron. "Taking the Workers Back Out: Recent Trends in the Structuring of Employment," in Research in Organizational Behavior, vol. 10, ed. Barry Staw and L.L. Cummings (Greenwich, Conn.: JAI Press), 1988, p. 36.
Polivka, Anne and Thomas Nardone. "On the Definition of Contingent Work," Monthly Labor Review 112 (December 1989): 9-16.
Pollack, Andrew. "Japan's Companies Moving Production to Sites Overseas," The New York Times, Sunday, August 29, 1993, p. 1.
Powell, Walter F. "Neither Markets nor Hierarchies: Network Forms of Organization," in Research in Organizational Behavior 12 (1990): 321.
Regini, Marino. "Industrial Relations in the Phase of Flexibility," (Department of Sociology, University of Milan, Spring 1987, unpublished MS).
Thurow, Lester C. "A Surge in Inequality," Scientific American, September 1987, pp. 30-37.
Tilly, Chris. "Dualism in Part-Time Employment," Industrial Relations 31 (Spring 1992): 331.
U.S. Bureau of the Census. "Workers with Low Earnings: 1964 to 1990," Current Population Reports, Consumer Income, series P-60, no. 178 (Washington, DC: U.S. Government Printing Office, March 1992), pp. 5, 8.
Bennett Harrison is professor of political economy in the H.F. Heinz III School of Public Policy and Management at Carnegie Mellon University. He has written or edited ten books and is a columnist for Technology Review. Harrison also appears regularly as a commentator on PBS's MacNeil-Lehrer Newshour and on the CBS Evening News. Excerpted from Lean and Mean: The Changing Landscape of Corporate Power in the Age of Flexibility. Copyright [C] 1994 by Bennett Harrison. Reprinted by arrangement with Basic Books, a division of HarperCollins Publishers, Inc.