Cohre Home   

Reinventing the Workplace:
How Business and Employees Can Both Win:  Employee Support
by Dr. David I. Levine

Employee Empowerment
Direct Participation
Representative Participation
Employee Motivation
Sharing Gains
Pay for Knowledge
Group Cohesiveness
Employment Security
Guaranteed Individual Rights
Employee Capability
High Levels of Training
Sharing Information
Reinventing The American Workplace Home


Q: What did Pope John XXIII tell the journalist who asked him how many people work at the Vatican?
A: "About half."

More than two centuries ago, Scottish economist Adam Smith outlined the efficiency benefits of a detailed division of labor, in which each worker performed the same task repeatedly. Implicit in his model of pin manufacturing was a hierarchy the same hierarchy that evolved into the almost complete division between thinking and doing at GM-Fremont.

Smith also outlined the problems with a detailed division of labor; for example, a person who again and again executes "a few very simple operations . . . generally becomes as stupid and ignorant as it is possible for a human creature to become."(1) Debate continued for many generations about the conditions under which managers can afford to hire only employees' hands, not their heads as well.

The fundamental problem of management is to convince employees that they will benefit by working hard and sharing ideas. If workers share their ideas with a noncooperative management, management will merely raise the expected level of output; the result is that workers will end up working harder for lower wages. Conversely, if managers give autonomy and rewards to workers who restrict output, profits will plummet. One likely outcome is that managers will not trust workers, will not expect workers to contribute ideas, and will not permit workers to innovate when they have ideas. The dilemma can be resolved only if managers and workers learn to trust each other. Workers will only work hard and share ideas if managers follow through on their promises of rewards for good ideas and efforts.

Elaborate theoretical explanations have been put forth for the virtues and weaknesses of participation.(2) Although employee involvement can raise productivity, satisfaction, and product quality, it may not always work. Given the many potential costs of participation, the evidence of participation's effects on productivity is, unsurprisingly, mixed. Each cost corresponds to a management practice that must be in place for involvement to succeed (see boxes on advantages and disadvantages of employee involvement and on difficulties in measuring the relationship between participation and performance).(3)

No magic formula exists for creating successful employee involvement. Nevertheless, organizations striving to tap into the resources of their work force must build support from employees, managers, unions (when present), and business partners. Three preconditions are required to build employee support and capability for involvement to succeed.(4) First, workers need to be involved and given some responsibility for organizational change and their daily work. Employees need mechanisms through which their suggestions can be evaluated and implemented. If jobs are narrowly defined and rigid, workers have less opportunity to make suggestions. The methods used for creating employee involvement range from active suggestion systems to self-directed work teams. What they have in common is that they give employees the ability to propose changes in work methods, the product, the choice of technology,other matters that affect performance, and the quality of worklife. Second, employees must be motivated to make improvements. While the incentive mechanisms in high-involvement organizations vary widely, the objective is to share the rewards gained from new ideas. Many organizations link pay and rewards to performance and new skills. In addition, most high-involvement organizations find it important to ensure that ideas generated do not result in adverse actions such as layoffs or dismissals. Third, workers must be capable of making improvements. Their knowledge and skills are developed through training and continuous learning, and their contributions are made possible through information sharing.

back to top

Employee Empowerment

For employees to be empowered, the company must provide means for employees to participate both at the shop or office floor and at higher levels in the organization.

back to top

Direct Participation

Successful employee involvement requires that workers have substantive involvement, not the superficial ability to make suggestions (but not be listened to) or to make decisions (but not about anything that matters). Employee involvement may pertain to many topics, including how employees do their job, how safe the jobs are, and how pleasant the working conditions are. What these topics have in common is that they focus on matters of importance to the work force. This simple lesson is frequently ignored in American efforts to implement programs calling themselves employee involvement. Not surprisingly, lack of employee empowerment is associated with high failure rates among such programs.

Consultative Participation. Consultative participation involves direct participatory arrangements with little or no formal influence. Typically, employees are allowed to give their opinions, but final decisions are made by management. The focus most often is on workplace organization and other shop-floor and personnel issues important to workers and about which they often have significant information not readily available to management. While worker suggestions are solicited, workers are not permitted to decide how to solve problems, and often they do not even implement the suggestions that are accepted by management.

In the United States, the quality circles (QCs) fad of the early 1980s provided strong evidence on the effectiveness of purely consultative participation. Quality circles at their peak were implemented by perhaps half of all large U.S. employers, although typically only in a minority of workplaces and including only 25 percent of the work force at each workplace. QCs usually consist of small voluntary groups of employees from the same work area who meet on a fairly regular basis to identify and solve quality, productivity, and other problems. Frequently, members of QCs receive special training in such subjects as group dynamics and problem solving. Despite their name, QCs often deal with subjects other than quality for example, work flow, productivity, safety, and employee welfare.

Numerous studies have examined consultative participation, in which workers have little power. M. L. Marks and others, for example, examined the effects of QCs in a manufacturing plant. The study compared the productivity of employees who volunteered to participate in the QC program with the productivity of nonparticipants over a twenty-four-month period. While the two groups had similar demographic characteristics and almost identical initial work records, the productivity of the nonparticipants increased by less than 10 percent, while that of the participants increased by more than 20 percent during the course of the study.(5)

Although a number of studies indicated that QCs had positive impacts, the preponderance of evidence does not suggest positive long-term effects from consultative participation. Maryellen Kelley, for example, found that productivity was lower at nonunion metal-working establishments (plants or offices) with worker-management problem-solving committees (typically quality circles) than at similar establishments without committees. William N. Cooke reached similar conclusions. In his study of Michigan manufacturing: Introducing teams in nonunion settings (typically with only consultative power) had minimal impact on productivity.(6) Both studies indicated that in union settings, when workers in teams had rights to due process, implicit gainsharing via the bargaining process, and representative participation, productivity was higher in factories with teams.

The most important evidence on the long-run effects of QCs comes from studies that do not directly measure their performance, but measure their longevity. Most studies revealed that the half-life of QCs was less than three years.(7) Anecdotal evidence suggests that virtually all QCs of the mid-1980s are no longer functioning; most have closed down, a minority have evolved into more substantive forms of involvement.

Evidence repeatedly showed that managers liked QCs because they gave workers the impression that someone was listening to them. In one survey, the most common benefit of QCs reported by managers was that they permitted the supervisors to explain managerial actions to the workers.(8) Until managers believe their subordinates have valuable ideas and are willing to substantially empower workers, few improvements in quality will result, and few employee involvement programs will flourish.

The studies of consultative participation, in short, support a general consensus that purely advisory shop-floor arrangements are not likely to achieve sustainable improvements in productivity. Such improvements require work reorganization and a broadening of employee participation in decisionmaking.(9)

Substantive Participation. Substantive participation in work and workplace decisions includes formal, direct participation schemes, such as work teams. Usually substantive participatory arrangements concentrate on the same kinds of issues as consultative arrangements, but workers have more influence. (Intermediate levels of power also exist, such as where workers' suggestions are implemented unless management specifically justifies its negative response. Sometimes, management is required to explain what changes would make the suggestions acceptable.)

Several early studies examined short-term empowerment, often as experiments. These results were typically positive.(10) Several other studies suggested that more participation in the form of more channels for participation and coverage of a broader range of issues leads to higher productivity. A study by Daniel Mitchell and others constructed an index of the number of issues over which participation was involved. Regression results indicated that the participation index was positively related to the productivity of both clerical and production workers.(11)

Two 1980 studies by John R. Cable and Felix R. FitzRoy examined a group of West German "industrial partnership" firms that provided both profit sharing and participatory rights for their workers. Cable and FitzRoy constructed an index of participation based on managers' perceptions of worker participation in various areas. The index rises as the extent of worker participation increases from no participation to purely advisory participation to active influence and as the number of issues on which workers participate expands. The regression results indicated that as the participation index rises, value-added rises (holding constant the amounts of labor and capital). Cable and FitzRoy also split their sample into companies with above- and below-average levels of employee involvement. On average, the high-involvement companies outperformed the low-participation firms by 15 percent, 177 percent, and 33 percent in output per worker, output per unit of capital, and profitability (rate of return on capital employed), respectively.(12)

More recent research has often focused on work teams. Although the role of teams varies, in many cases workers in work teams are given wide discretion to organize their tasks and operate with very little supervision. Typically, these groups make their own work assignments and determine their own work methods, subject to overall workflow requirements.

Work teams have been given responsibility for developing relations with vendors, determining which operations can be handled individually and which by the group as a whole, setting work pace (perhaps fast in the morning and slow in the afternoon), training new employees, and keeping financial records. Sometimes team members serve in roles normally reserved for staff personnel or supervisors: chairing the plant safety committee, redesigning work equipment, or troubleshooting customers' problems. At some locations, the job of supervisor is rotated among members of the group. At GM's Saturn plant, "councilors" (first-line supervisors)are elected by their subordinates.

Even in companies with substantive participation, empirical research indicated that productivity increases do not follow automatically. For example, a 1989 study by John F. Krafcik on the automobile industry suggested that team production techniques, which are an integral part of the Japanese automobile production system, have a positive effect on productivity in Japanese plants. But these plants also have several other unique features of Japanese industrial relations, and the studies did not distinguish the effects of work teams from the effects of the other factors on productivity performance.(13)

In contrast, econometric work on a U.S. auto plant by Harry C. Katz, Thomas A. Kochan, and Jeffrey Keefe in 1988 distinguished the productivity effects of work teams from the productivity effects of other forms of substantive and consultative participation in technology and work group decisions. The results suggested that team production techniques, by themselves, lower labor productivity and quality, while substantive participation in work groups when coupled with technology-related decisions increases labor productivity and product quality. The statistical association between the measures of participation and plant performance, however, are weak (many of the estimated coefficients are not statistically significant at even the 10 percent level).(14)

These results and the findings of other empirical studies by Katz and his colleagues suggested that team production techniques, like QCs, accomplish little unless they are integrated with broader changes in the company's industrial relations environment, including changes that facilitate greater employee participation in higher-level decisionmaking. This conclusion is buttressed by qualitative case study evidence indicating that team production produces sizable improvements in both organizational performance and the quality of working life when accompanied by such organizational changes as increased worker participation in business strategy decisions, job-security programs, and group pay reward structures.(15) Broader organizational changes appear to be essential to the long-term viability and performance of substantive employee participation.

Maryellen Kelley's study of machine tools operators provided evidence that substantive involvement is more important than consultative involvement.(16) She found that workers who program their own machine tools have substantially higher productivity, even though in nonunion settings having a quality circle or similar committee has a negative effect on productivity.(17)

Anthony Carnevale and his colleagues studied 239 organizations that identified themselves as restructuring. They found that more extensive use of teams and higher skill variety are correlated with higher quality and lower absenteeism.(18)

Other studies focused not on work practices, but on employees' perceptions of empowerment and commitment. For example, Daniel Denison and Gary S. Hansen and Birger Wenerfelt examined companies that had employees, within a single section or division, fill out a questionnaire known as the "Survey of Organizations." They found that companies whose workers reported more participative cultures had a higher return on investment and sales and a higher return on assets than the industries as a whole.(19)

back to top

Representative Participation

Successful cases of employee involvement often provide representative participation to complement direct participation. Typical institutions include union-management committees, workers' representatives on the board of directors, and works councils.

Representative forms of participation are important for resolving disputes that occur at a lower level; for example, when a work group feels a suggestion will increase safety, but the supervisor disagrees. At NUMMI, for example, a union-management safety committee meets to discuss problems that are not settled on the shop floor. Representative forms of participation are also important when issues span multiple work groups. Finally, these forms of participation can provide benefits to all workers that no single worker has an incentive to provide. For example, on average no single worker will find it worth his or her time to ensure that management is telling the truth in its financial statement, even if that statement is the basis of gainsharing. An elected representative, however, will be able to check the numbers and provide reassurance to all employees.

Overall, the available empirical literature suggests that various forms of representative participation can improve performance when they are part of a package of participatory policies. Taken alone, they may improve labor-management relations, but they have little effect on productivity. This section focuses on codetermination and works councils; union labor-management committees, the most common form of representative participation in the United States, are discussed in the following chapter.

Membership on Company Boards of Directors. Union representatives have served on the boards of Chrysler and several financially troubled airline, steel, and trucking companies. In addition, employee directors (not always selected by the union) have been involved in a considerable number of employee stock ownership plans (ESOPs) and worker buyouts. As has been the experience in other countries, employee directors in the United States have not been very important.(20) They have been handicapped by rules keeping board deliberations confidential, thus restricting communication with constituents. Directors from the shop floor lack the technical expertise to make contributions in areas such as finance. Regardless of the employee director's skill, management can usually keep issues that matter the most to employees off the board's agenda. In any case, boards, which meet as infrequently as once a quarter, usually exert little influence.

Workers on the board may be more important when board membership serves as only one of several channels for employee participation and only one of several distinctive features of the company's industrial relations system. For example, in a 1987 study on British retail cooperatives, Derek C. Jones found that the presence of worker representatives on the board of directors had a positive but modest effect on productivity.(21) The ineffectiveness of codetermination has been verified in empirical studies of West German codetermination laws requiring worker representatives on company boards of directors. Jan Svejnar concluded that the introduction of codetermination had either no significant productivity effect or a mildly negative one. In 1976 Germany passed a law requiring most large employers to increase from one-third employee representation to equal numbers of employee and stockholder representatives. Felix R. FitzRoy and Kornelius Kraft found that between 1976 and 1983 companies subject to this law enjoyed significantly less productivity growth than did smaller companies not subject to it. They pointed out that part of the result was plausibly because of the influence of employee directors in slowing layoffs. If this was the case, then the slower measured productivity growth represented a private cost to the owners, but not a social cost, because the workers were more productive at the companies than they would have been if unemployed.(22)

Work Councils. All continental nations of western Europe require large employers to have employee representation councils. The councils typically have the right to be informed about the financial situation of the company and to discuss certain employment-related matters such as training and layoffs. In some countries, works councils have the authority to veto some employment-related decisions. In Sweden, for example, no layoffs can occur without the concurrence of the works council.

Most researchers have found minimal effects on productivity for mandated works councils.(23) At the same time, case study evidence suggests that works councils can often serve a useful role in moving information up and down the organization and in improving the quality of decisionmaking. In addition, works councils create a problem-solving atmosphere, and managers prefer dealing with works councils to unions for problem solving.

As with codetermination, works councils may be more important when coupled with other forms of involvement. A 1991 econometric study by Motohiro Morishima examined the productivity effects of joint consultation committees, a form of representative participation common in Japan. These committees are corporate- and plant-level bodies that deal with business strategies and plans pertinent to the entire organization. The committees serve two major functions: sharing of business information with employees and prior consultation by the management with employees on upcoming business decisions. Morishima's econometric results indicated a strong and sizable positive association between an index of information sharing and firm profitability, employee productivity, and labor costs. An important point about the Japanese case is that "information sharing complements other aspects of the Japanese enterprise-level industrial relations system."(24)

back to top

Employee Motivation

What benefit has man from his toil under the sun? -ECCLESIASTES

Traditional reward systems focus on individual achievement and the skills used at the current job. New forms of work require new forms of rewarding the cooperation and learning that underlie high-involvement workplaces. In addition, employees must be reassured that their ideas will not lead to layoffs or punishments such as dismissals.

back to top

Sharing Gains

Some kind of sharing of rewards from involvement is a key element of almost all participatory systems. (Profit sharing, which is usually based on the accounting profits of the enterprise, is differentiated here from gainsharing, which is usually based on cost reductions at a particular department, plant, or office.)

Financial sharing can occur without participation and vice versa, but both theory and evidence suggest that the two are likely to go together in successful participatory systems. In the short run, participation may be its own reward for many employees. In the long run, however, sustained, effective participation requires that employees be rewarded for the extra effort that participation entails and that they receive a share of any increased productivity or profits. Workers feel that it is unfair if their ideas generate cost savings and they do not share the benefits. Group-based gainsharing provides workers with incentives to maintain norms of high effort, to monitor each other, and to discipline workers who are shirking. More positively, group-based pay gives workers incentives to cooperate and not to try to advance at the expense of their colleagues.

Just as participation can lead to demands for profit sharing, profit sharing can lead to demands for participation. In a company with profit sharing, workers' incomes depend upon the management's decisions, and workers want to have a say in these decisions.(25) Moreover, growing empirical evidence indicates a positive interaction between profit sharing and participation, implying that the combination is more potent than the sum of its parts.(26) According to a 1991 study, "the available evidence is strongly suggestive that for employee ownership, including profit-sharing and ESOP programs, to have a strong impact on performance, it needs to be accompanied by provisions for worker participation in decision making."(27)

Gainsharing and Total Quality Management. Changes in reward systems have often been neglected by advocates of total quality management (TQM). For example, in many workplaces, the main effect of the quality movement has been that, in addition to their regular work, employees must fill in statistical process control (SPC) charts that track when errors occur. With little employee motivation, few of these charts are completed as they should be. When workers must fill in SPC charts but are not empowered, SPC is little more than a speed-up of the pace of work. Furthermore, without the cooperation of the employees in analyzing the data they collect, SPC loses most of its potential for improvements.

Quality consultants are aware that SPC can be perceived as a speed-up. Quality guru Joseph M. Juran, for example, warned managers that unless workers have substantial training and participation they will view the introduction of control charts as an unofficial tightening of specifications leading to more work.(28) That is, workers will perceive the new quality emphasis as an attempt to make them work harder.

Juran noted that upper management will experience the same problem. With the new emphasis on quality, top managers will need to measure the quality achievements of their divisions, attend more cross-functional meetings, and so forth. Juran advocated that high-level managers' pay should rise substantially when they achieve their quality goals. These bonuses will eliminate high-level managers' concerns about a speed-up of their jobs and motivate them to perform the extra quality-oriented tasks. Juran ignored the possibility of (or need for) rewarding workers for doing extra work to achieve their quality goals, however.

In contrast to many total quality consultants' recommendations, successful TQM efforts, such as at NUMMI, provide a number of gainsharing mechanisms. Most directly, NUMMI's workers' pay is tied to increases in productivity and to performance based on productivity and indicators of the quality of their cars. In addition, individuals and groups receive modest rewards for their suggestions. Finally, the presence of a union implies gainsharing exists over the long term; if NUMMI prospers, the union will bargain for a share of the profits. All of these incentives help motivate employees to share their ideas and enforce norms of high effort within the factory.

Several studies have found that high-involvement companies are more likely to have gainsharing. For example, in the 1987 survey of the Fortune 1000, gainsharing, where pay is based on plant or departmental performance, is roughly twice as prevalent in companies with above-average employee involvement than in companies with below-average involvement.(29) A 1991 survey of steel plants revealed a wider gap. More than half of the companies with work teams had multiple incentives based on a department's performance, while only 11 percent of those without work teams had nontraditional incentives.(30)

The gap is smaller for profit sharing, a company wide incentive. In the Fortune 1000, profit sharing is almost half again more likely in companies with above-average levels of employee involvement than in companies with below-average levels (36 percent versus 26 percent).(31) In a 1977 survey of the U.S. population, employees who reported having above-average autonomy at work also were about half again more likely to have profit sharing than those with below-average autonomy (23 percent versus 15 percent).(32) In a 1994 steel industry survey, profit sharing was only slightly more prevalent in plants with work teams (68 percent versus 64 percent).(33)

In his 1990 survey of publicly traded companies, Mark Huselid found that high-involvement companies also tended to have more use of profit sharing and incentive plans. Similarly, as a result of his 1990 survey, Paul Osterman reported that establishments with at least half the core work force involved in high-involvement practices were more likely to have profit sharing (48 percent versus 39 percent). Differences in incidence of gainsharing were not statistically significant.(34) Canadian establishments with semiautonomous work teams were more likely to have both profit sharing and gainsharing.(35)

In the 1987 survey of the Fortune 1000, companies that reported above-average success from employee involvement provided substantially higher levels of gainsharing and slightly higher profit sharing. Consistent with their emphasis on group-based rewards, they provided similar levels of individual incentives and merit pay. In the 1990 survey, companies that reported above-average success from employee involvement provided higher levels of profit sharing, gainsharing, and nonmonetary awards. They had similar levels of individual incentives, team incentives, and employee stock ownership.(36)

Employee Ownership. Employee ownership occurs in two main forms in the United States: thousands of employee stock ownership plans, in which workers own a minority of the company's shares, and a much smaller number of companies that are majority owned by the work force. Although some of the employees at the majority-owned companies own their stock via an ESOP, the companies are referred to here as worker cooperatives.

Employee stock ownership has long been suspected to lead to higher productivity, as workers act as owners. Jeffrey Pfeffer noted, "It is probably no coincidence that all five of the companies. . .providing the best shareholder returns from 1972 to 1992 appear on the Employee Ownership 1000," a listing of one thousand companies in which employees own more than 4 percent of the stock of a corporation.(37) (The five companies are Southwest Airlines, Wal-Mart, Tyson Foods, Circuit City, and Plenum Publishing.) In spite of those impressive performances, most ESOPs have little or no employee participation and have no measurable effects on productivity. However, ESOP companies that provide employees with additional opportunities for participation in decisionmaking are significantly more likely to outperform conventionally owned companies than ESOP companies that do not. Furthermore, of the various forms of participation, those that reach closest to the shop floor have the biggest productivity effects, while stock voting rights or employee representation on company boards of directors have insignificant productivity effects.(38) These findings for ESOPs are consistent with the findings for conventionally owned companies.

As with representative participation, the Japanese experience with ESOPs is somewhat different. Perhaps because of extensive shop-floor consultation, in Japanese companies the net effect of introducing an ESOP was an almost 7 percent increase in productivity.(39)

Extensive empirical literature is available on the performance of worker-owned companies (cooperatives) in the United States and abroad.(40) At worker-owned companies, most studies concluded that the extent of employee ownership has a significant positive effect on the firm's productivity. Employee ownership is measured variously as the share of the firm's equity owned by employees or the share of the firm's employees who choose to become firm members by investing in a minimum number of shares. Richard J. Long and Bodil Thordarson found the productivity performance of worker-owned companies superior to that of conventionally owned companies that are otherwise similar. A study of the death rate of French cooperatives yielded indirect evidence of high effectiveness.(41) Death rates of cooperatives were only a fraction of the rate in similar conventional firms.

None of these studies directly measured the productivity effects of the participatory arrangements that usually accompany significant employee ownership or distinguished the effects from the productivity effects of employee ownership per se. Moreover, several of the studies, including the extensive econometric studies of the European producer co-ops, interpreted their ownership variables as proxies for the extent of employee participation in enterprise decisionmaking.(42) Such variables are imprecise and provide no information about the actual form and content of participation in the companies.

Several studies examined the effects of participation on productivity in worker-owned companies. These studies fairly consistently found that participation raises productivity in the worker cooperatives.(43)

Overall, the results of empirical research on employee-owned companies suggested that, within these companies, employee participation is positively associated with productivity. However, the companies almost always have several other characteristics that distinguish them from conventionally owned companies. For example, the European worker coops, which have been the subject of the most extensive empirical inquiry, typically have managements committed to employee ownership and representation, job security, compressed status and compensation differentials, and guaranteed worker rights. As a result, drawing general conclusions about the effects of employee participation on productivity from the experience of these companies is dangerous. The effects are likely much smaller in conventionally owned and organized companies without such characteristics.

back to top

Pay for Knowledge

Traditional workplaces reward people for the particular job they perform. Workers can lose pay or benefits when transferred to lower-skilled jobs, and they have no direct incentive to continuously improve their skills. Given these difficulties, many high-involvement companies have a pay-for-knowledge compensation system that rewards employees for each new skill acquired. These pay systems reward workers for learning multiple skills, which increases flexibility, enhances workers' ability to see the big picture and to suggest improvements, and encourages teamwork.(44)

An increasing proportion of American businesses have introduced knowledge-based pay systems. The proportion of the Fortune 1000 with pay-for-knowledge rose from 40 to 51 percent between 1987 and 1990, although most plans affected less than 20 percent of the labor force.(45) In his 1990 survey, Paul Osterman found that establishments with at least half the core work force involved in high-involvement practices were more likely to have pay for knowledge (36 percent versus 20 percent).(46) Canadian establishments with semi-autonomous work teams were also more likely to have pay for knowledge.(47) Companies in the Fortune 1000 that reported above-average success on employee involvement also have higher pay for knowledge.

back to top

Group Cohesiveness

Most participatory workplaces reduce pay and status differentials among employees, particularly between workers and managers, relative to nonparticipatory workplaces. Smaller differentials are associated with participation for three related reasons.

First, narrow differences in wages and status help develop an atmosphere of trust and confidence between workers and management, reinforcing the atmosphere of participation. Large differences in status can inhibit participation.(48) As employee involvement increases, management relies more on the good will and cooperation of employees. Employees often feel large wage differentials are unfair, and employees who feel disadvantaged are less supportive of the goals of the highly rewarded group.(49)

Second, bonuses based on group output give workers incentives to work for group goals and provide incentives for workers to monitor and discipline free riders. Narrow wage dispersion promotes cooperation, while large wage differences and competition for promotions can reduce cooperation, as workers try to win the bonus or promotion "tournament."(50)

Group-based pay, almost by definition, reduces individual pay differentials. According to Morton Deutsch, who has been studying the relationship between egalitarianism and productivity for more than forty years, when "efficiency requires efficient cooperation, almost any movement towards a democratic egalitarian structure increases effectiveness." Numerous laboratory experiments have found that narrow wage dispersion increases worker cohesiveness and increases productivity.(51)

Third, participation may extend into the realm of compensation. To the extent that the median employee exerts influence on the firm's compensation policy, pressure likely will exist to reduce high-end wages, thereby compressing wage differentials.

Regardless of the theoretical rationale, in practice, most participatory workplaces including worker-owned companies in the United States and abroad, large Japanese companies, and successful participatory companies in the United States tend to pay relatively egalitarian wages and to reduce status differences, largely to induce cohesiveness within the work force. For example, NUMMI has only a single parking lot and cafeteria, and managers are more likely to wear the company uniform than a suit. Personnel research supporting the importance of equality includes works by John F. Witte and Katrina Berman on U.S. participatory companies; Keith Bradley and Alan Gelb on foreign cooperatives; and William Ouchi, Thomas Rohlen, and Ezra Vogel on Japanese companies both at home and abroad. Edward E. Lawler III and Michael Beer and others recommended that participatory employers rely heavily on group-based compensation and narrow wage differentials.(52)

Policies to promote cohesiveness are more common at high-involvement companies. One commonly used mechanism to reduce status differences is to eliminate time clocks and put all workers on salary. Use of all-salary pay is more prevalent at companies with high levels of employee involvement than at companies with low levels of employee involvement, according to the 1987 survey of the Fortune 1000.(53) Similarly, at companies undergoing workplace transformation, extensive use of teams is correlated with an all-salaried pay system.(54) Consistent with a greater emphasis on cohesiveness, in an Indiana sample of manufacturing plants, workers reported a higher average level of cohesiveness at plants where autonomy or quality circle membership was high.(55)

Policies to promote cohesiveness are more common at successful high-involvement companies. In the 1990 survey of the Fortune 1000, companies with high reported success with their employee involvement had substantially higher levels of all-salaried work force.(56)

A sample of business units provides indirect evidence to support this contention. Business units with less inequality (both within management ranks and between management and workers) reported higher quality levels.(57) Product quality is the performance characteristic most responsive to employee effort and initiative as well as the focus of recent TQM-inspired employee involvement efforts.

back to top

Employment Security

Most high-involvement workplaces have implicit or explicit long-term employment contracts with their workers, contracts that stress reciprocal commitments and management's pledge to minimize the need for layoffs. Successful participatory systems usually avoid laying off workers for several related reasons. (Economic crises can prod management, workers, and unions to initiate participatory experiments. Nevertheless, employment stability reduces the costs of maintaining participation.)

Most directly, workers are unlikely to cooperate in increasing efficiency if they fear that by so doing they jeopardize their employment. Guarantees of job security reduce fears that higher productivity will lead to layoffs. Conversely, the fear of layoffs has inhibited the success of participation in several cases.(58)

Workers with job security expect to remain at their jobs for many years and are more likely to forgo short-term gains (for example, shirking) to build a more effective organization. Participation works only if employees share their ideas and if managers reward workers for increased productivity. In the short run, the rational strategy is for workers to hide productivity-improving techniques to enjoy more on-the-job leisure. Similarly, the company's short-run rational strategy is to deny that productivity has increased to avoid raising pay. Only a long perspective can convince both players to cooperate.

To the extent that participation relies upon work group cooperation and employees monitoring one another, long-term employment relations are essential. The longer an employee expects to be in a work group, the more effective are group-based rewards and social approval as motivators.

Participatory employers often make large investments in the selection, socialization, and training of workers. From the firm's point of view, long-term employment relations recover the higher investment in human resources that usually accompanies participation. Conversely, training in a variety of jobs and skills lowers the costs of long-term employment relations, because workers can be shifted to more critical tasks. Moreover, the process of training shows a commitment to workers that may increase their trust in managers' promises to avoid layoffs.

The case study literature on labor-managed companies provides evidence that participatory workplaces rarely lay off worker-members (although temporary workers are laid off). Fairness considerations constrain these companies from reducing the work force via layoffs. More generally, layoffs are inconsistent with the sense of membership and community that participatory companies try to instill.

Numerous examples exist of companies that provide high levels of employment security as part of a package of policies to ensure the success of participation.(59) In the United States, Hewlett-Packard and NUMMI have had policies of avoiding layoffs. During the early 1980s Hewlett-Packard adopted hiring freezes, shortened work weeks, and ordered the elimination of perquisites, for example, while other companies were engaging in mass layoffs.(60) During the mid-1980s NUMMI sent its entire work force into retraining during a time when, as one union official put it, General Motors (the previous owners of the plant) "would have closed the second shift."(61)

The relationship between participation and avoiding layoffs appears to be causal, not merely correlational. Evidence comes from IBM, for many decades the preeminent U.S. employer that avoided layoffs. For example, during the 1972 75 slowdown, IBM transferred seventeen thousand workers, completely retraining seven thousand of them.(62) A typical comment by an IBM executive:

"Our people, by using their minds as well as their hands, have cut two-thirds of the hours that go into manufacturing our product. . . . That achievement would have been impossible without productive and committed employees. And much of their commitment stems from the security they know is theirs through our practice of full employment."(63)

Declining market share in the early 1990s induced IBM to begin its first set of layoffs. Morale has plummeted at the company, and anecdotal evidence suggests employee loyalty and participation are substantially lower as well.

No-layoff policies are also common in other successful examples of worker participation, such as the Mondragon worker cooperative network in Spain, as well as in large companies in Japan.(64)

The most direct evidence that downturns in demand hurt participatory efforts is found in the survey of the 1987 Fortune 1000 data survey. "Worsened business conditions" was cited as a barrier to employee involvement efforts by 45 percent of the respondents, with 14 percent claiming that it was a barrier to a "great" or "very great" extent. Furthermore, short-term performance pressures (which worsen during downturns when performance is low) were cited as a barrier by 87 percent of the respondents and as "great" or "very great" by 43 percent.(65)

Policies to avoid layoffs are more common at high-involvement companies. According to a 1982 survey, Indiana manufacturing plants with quality circles and with workers reporting high levels of autonomy also had more workers agreeing that "My job security is good."(66) In his 1990 survey of publicly traded companies, Mark Huselid found that high involvement companies also tended to have more promotion from within.(67)

Companies with above-average levels of employee involvement also had above-average levels of employment security policies in the 1987 survey of the Fortune 1000. This relationship was not replicated in the 1990 survey. In both the 1987 and 1990 surveys, companies that reported above-average success with employee involvement also reported higher proportions of their work forces were covered by policies to provide employment security.(68) In a 1992 survey of U.S. workplaces, companies with work teams were only half as likely to have experienced layoffs as those with no teams (although no difference was found in the incidence of formal policies to avoid layoffs).(69)

The lack of relationship between formal employment security policies and high-involvement practices may be the result of the recent rise of reengineering, in which job redesign is explicitly tied to downsizing. The tensions between asking workers to "work smarter" and laying off their colleagues, when "smarter" workers reduce labor requirements, are not emphasized in the literature prescribing reengineering.(70)

back to top

Guaranteed Individual Rights

Participatory systems usually have rules and procedures to safeguard employee rights. To participate effectively, people need "the assurance that they will not be penalized for their participation. Such acts as criticizing existing procedures or opposing proposed policy changes could invite reprisals from management."(71) Personnel systems governed by the rule of law are perceived as more legitimate and fair than systems in which decisions are at the discretion of managers.(72)

Empirically, the rule of law is found in most high-commitment work organizations, typically implemented with a formal grievance or complaint resolution system.(73) Paul Bernstein found guaranteed rights in essentially all of the successful participatory companies he studied. He concluded that such rights are a necessary component of workplace democratization.

Guaranteed rights increase workers' trust in the company. Several studies indicated how high-trust environments depend on employee perceptions of due process and facilitate employee participation, better performance, creativity, and communication.(74)

Guaranteed individual rights are an important part of long-term employment relations, because workers have an alternative to quitting if they are unhappy about one aspect of their jobs. Richard Freeman and James Medoff surveyed the evidence relating individual rights and performance. They concluded that union workers with guaranteed individual rights have higher productivity and are less likely to quit than are other workers.(75)

The most crucial right for most employees is knowing that they will not be dismissed for their critical ideas but will only be fired when the company can show good cause. In contrast to "just-cause" dismissal policies, the "at-will" dismissal policies at most U.S. companies mean that employment can be terminated at any time "for good cause, for bad cause, or even for cause morally wrong."(76)

When companies guarantee individual rights, traditional motivators such as fear of dismissal become less effective. However, in most successful participatory companies, workers are motivated by group rewards, peer pressure, and so forth not by traditional fear of punishment. The evidence suggests that the gains in perceived fairness and in workers' willingness to participate outweigh the losses for participatory companies. High-commitment companies such as Hewlett-Packard have voluntarily adopted just cause and established many other employee rights, implying that the company anticipates net benefits from constraining managerial actions.

In his 1990 survey of publicly traded companies, Mark Huselid found that high-involvement companies also tended to have more employees with access to a formal grievance or complaint resolution procedure.(77)

back to top

Employee Capability

Even if workers are empowered and motivated, to make improvements they need to have ability. At high-involvement plants, their knowledge and skills are developed through training and continuous learning, and their contributions are made possible through information sharing.

back to top

High Levels of Training

Participatory companies invest much more in their workers than the average U.S. employer. At Saturn, for example, new hires go through 350 to 700 hours of training, even though they are long-time auto workers.(78) Training involves problem solving, quality tools, and understanding the auto industry and labor history. Because Saturn is founded on a philosophy of consensus decisionmaking but just about everyone hates meetings a large portion of the training is focused on effective skills in working in groups. Saturn workers are also trained in a number of different jobs, which improves their understanding of the production process and their ability to discover and implement improvements.

Consistent with this high level of training, Saturn workers receive a constant stream of information on how the plant is performing, the status of the automobile industry, and how their department performance compares with its goals. Saturn workers also receive on-the-job training in a variety of jobs as they rotate throughout their work group.

The participatory company's investment in human resources goes far beyond the initial training. Selection costs are typically far higher, because dismissals so rarely occur. Socialization is also a major investment in new hires reaching its ultimate expression in "boot camps" found at some Japanese companies, where new recruits undergo strenuous physical exercises in which they learn to rely on each other.(79)

High level of training is not special to Saturn. As high-involvement employers rely more heavily on decisionmaking by employees, those employees need many skills to have the ability to succeed. Both formal classroom training and training on the job (via job rotation and cross-training) are common at high-involvement enterprises.

Numerous studies found very high rates of return on training. At the same time, evidence is available that training by itself is not effective in raising productivity substantially.(80)

Some of the most convincing studies of the role of training involve detailed cross-national comparative case studies within a particular industry.(81) These studies consistently found that U.S. and British employers train less than their German and Japanese counterparts. For example, newly hired auto assembly workers receive 310 hours of training in Japan and 280 hours of training in Japanese-managed plants located in the United States, but only 48 hours of training at traditional American plants. Moreover, employees at traditional plants have substantially lower autonomy. The net result is substantially lower productivity and quality.

High levels of training are more common at high-involvement companies. In the 1990 Fortune 1000 survey, companies that reported above-average success from employee involvement provided substantially higher levels of training in leadership skills, job skills training, and team-building skills, while they provided marginally higher levels of decisionmaking and problem-solving skills and in quality and statistical analysis and skills for understanding business.(82)

Similarly, in the Indiana survey of manufacturing plants, on-the-job and off-the-job training are both higher at plants where autonomy or quality circle membership are above average.(83) The 1992 survey of steel plants showed that companies with work teams are much more likely to have training in problem solving, high overall training, and sharing of financial information, while the use of teams is correlated with high levels of multilevel training at companies undergoing restructuring.(84)

In his 1990 survey, Paul Osterman found that establishments with at least half the core work force involved in high-involvement practices had higher levels of training (38 percent of workers in off-the-job training versus 22 percent) and cross-training (53 percent versus 31 percent). In his 1990 survey of publicly traded companies, Mark Huselid found that high-involvement companies also tended to have more hours of training per employee. At companies undergoing workplace transformation, higher intensity of training correlated with higher rates of on-time delivery and quality.(85)

High levels of training are more common at successful high-involvement companies. In the 1987 Fortune 1000 survey, companies that reported above-average success from employee involvement provided higher levels of training in decisionmaking and problem-solving skills and in quality and statistical analysis, although they provided only marginally higher levels of training in leadership skills, skills for understanding business, and team-building skills. In the 1990 Fortune 1000 survey, companies that reported above-average success from employee involvement provided substantially higher levels of training in leadership skills, job skills training, and team-building skills. They provided marginally higher levels of training in decisionmaking and problem-solving skills, quality and statistical analysis, and skills for understanding business.(86)

back to top

Sharing Information

The absence of information limits employees from participating in the business's mission and outcomes. High-involvement companies tend to provide their employees with more information on the progress of the company, new technology, and the big picture of the enterprise.

Employees who understand management priorities and budgetary constraints can make better decisions concerning training issues, customer service problems, quality, product design, and work process improvements. For this strategy to be effective, workers should receive training in how to interpret and apply the information they receive. Information that addresses the operating results of the employees' team or unit will be more useful than overall corporate operating results. Although having employees see and understand the big picture is important, information related directly to their work will allow them to make necessary improvement to increase productivity, or it can serve as feedback for improvements already implemented.

The 1987 Fortune 1000 survey revealed that high-involvement employers share more information on the business unit's operating results, fellow employees' pay, competitors' relative performance, planned new technology, and the company's plans, goals, and overall operating results. A 1990 follow-up showed that the gap had widened.(87)

Companies that reported above-average success from employee involvement also provided their employees with more information on the business unit's operating results, fellow employees' pay, competitors' relative performance, and planned new technology. They provided similar levels of information on the company's plans, goals, and overall operating results. In his 1990 survey of publicly traded companies, Mark Huselid found that high-involvement companies also tended to have employee attitude surveys and formal information-sharing programs (for example, a newsletter). Similarly, in Casey Ichniowski, Kathryn Shaw, and Giovanni Prennushi's study of steel lines, companies with teams are more likely to receive financial information on a regular basis.(88)

back to top

© Institute for Research on Labor and Employment, University of California, Berkeley. All rights reserved.
  Last update 2/8/99.