Since the early 1980s, the organization of
work and the structure of firms have been
undergoing a process of change that is not yet
completed. New technologies, the
intensifi-cation of domestic and international
com-petition, and various political changes that
have affected many countries, have forced
firms to change their business strategy in order
to survive, and have allowed them to adopt
policies that make them more profitable
(Osterman, 1994; Ichniowski et al., 1996;
Baker, 1999; Cappelli and Neumark, 2001).
Nien-Chi Liu, and Yong-Seung Park*
The Organization of Work:
Changes and Their
Consequences
* Avner Ben-Ner is Director of the Industrial Relations Center and Chair of the Department of Human Resources
and Industrial Relations at the University of Minnesota’s Carlson School of Management. abenner@csom.umn.edu
Fanmin Kong is Associate Professor of Human Resources and Industrial Relations at the Guanghua School of
Management at Peking University in Beijing, China. fkong@gsm.pku.edu.cn
Tzu-Shian Han is Associate Professor in the Department of Business Administration at National Chengchi
Uni-versity in Taipei, Taiwan. than@nccu.edu.tw
Nien-Chi Liu is Assistant Professor of Human Resources Management at the National Central University’s
Institute of Human Resources Management in Taipei, Taiwan. nliu@cc.ncu.edu.tw
Yong-Seung Park is Assistant Professor of Human Resources Management and Industrial Relations in the
Department of Business Administration at Kyung Hee University, South Korea. yspark@nms.kyunghee.ac.kr The paper documents the changes in human resource practices from the early 1980s to the
middle of the 1990s, using a unique and comprehensive data set concerning a sample of about 800 firms from a wide range of industries in the state of Minnesota. A major aspect in these changes concerns the widespread adoption of practices that emphasize employee involvement in decision-making and in firm financial returns. Focusing on cross-sectional data the paper examines the determinants of human resource practices and finds that differences in the complexity of tasks and the interdependence among the tasks of core employees explain well the prevalence of employee participation in decision-making both individually and in groups. The association between human resource practices and outcomes of interest to employees and shareholders is complicated; the empirical findings do not support strong statements concerning the effect of employee participation in decision-making and in financial returns on outcomes. JEL codes M12, L23, J53
support these rights, and whether these rights are afforded to individual employees or groups of employees.3
Programs fostering employee participation
in decision-making include, at the individual
level, latitude for employees to make discretionary decisions; at the group level, various plans that involve employees in making some decisions (such as quality circles, self-managed work teams, and joint labor-management committees); and, at the firm level, employee representation on the board of directors.
Plans that provide for employee
participa-tion in the financial performance of the firm
tie employees’ compensation to some measure of performance. Individual financial perfor-mance plans include commissions and performance-based pay. Group bonuses and gain-sharing programs are examples of group-level financial performance plans, while stock purchases, cash profit sharing, deferred profit sharing, and employee stock ownership plans (ESOPs) are firm-level financial performance plans. Finally, s complemen example, sel live up to t receive tra Individual-include skill and trainin based prac redesign, a skills.4 In order human reso the 1990s Relations Minnesota h from comp ment (state a aspects of present arti 1994–96 su information over time.5 Figure 1 with at leas emphasizes h decision-urns, with inds. Today, h computers ted by com-workers are am; in some le decision-heir work, ame human machine. A es receives a ts as part of or own stock work. Many r financial not to their that of the However, in k in the more supervisors’ ges. The first the changes 1980 to the es? In many ess strategies g goods and emands on workers are plexity than y must work technology d. In other ed much, or rategy made The second some factors
that affect the contemporary organization of work, focusing on the tasks of core employees. What are the consequences of the organization of work? Are workplaces that rely on more employee participation safer than other workplaces? Do shareholders or workers benefit more financially in more participatory firms or in workplaces organized in the more traditional way? The third objective of the paper is to offer a preliminary analysis of the consequences of the organization of work for workers and shareholders.
The paper summarizes the results of an ongoing project conducted at the Industrial Relations Center at the University of Minnesota.1It is written in a fashion aimed to
inform readers of the main qualitative results of the project, and since it covers a very large territory, the relevant literature and theoretical background are quite large, the data sources and issues that are related to them are numerous, and the econometric techniques employed in the analyses summarized in the paper are diverse. In order to maintain coherence of the argument and to keep the paper at a reasonable length, the paper is brief on these matters (but points to papers that contain more detailed information).
Changes in the Organization of
Work Since the Early 1980s
The organization of work consists of the practices that guide and direct the work of employees. These human resource practices can be grouped according to various criteria. We focus on the allocation of the key rights in an organization, the rights to decision-making and to financial returns,2and practices that
3. Not all practices that we consider in this paper as human resource pr tioners, and not all the practices that practitioners term human resourc of this paper. For example, we consider employee ownership to guide resource practice, but practitioners regard it as a financial plan. And pr whereby employees can arrive and leave work at individual times, as a nore it for our purposes.
4. In the remainder of this paper we ignore the distinction between grou in the text are the actual practices that are included in our empirical w 5. The Minnesota Human Resources Practices Survey focused on 2,051 p with at least 20 employees representing a broad spectrum of industrie response rate was 42% (861 surveys); in many analyses, the number o cause incomplete responses to some items or missing observations in survey data. The survey questionnaire asked respondents about plans decision-making (for example, through self-managed work teams) or for which they work (for example, through individual or group incen human resources practices (such as training, employment security, and ticipation in decision making; the degree of information sharing with out by different groups of employees; the company’s reliance on comp of firm organization. The survey also asked when various programs a tinued. The dataset used for the study included information supplie strument, data made available by Minnesota state agencies, federal age mation. For further information about the study, see Ben-Ner et al. (2 rted by grants from the Alfred P. Sloan Foundation, and the University of
Regional Affairs and the Retail Food Industry Center. n-Ner and Jones (1995).
The Choice of Human Resource
Practices
Theoretical considerations
Much has been written in recent years about the choice of the human resource practices. The theoretical perspectives employed by different authors vary, and so do their empirical frameworks. As a result, the findings are not directly and strictly comparable.7In
this article we proceed with our theoretical perspective, which is derived primarily from organizational economics and from the sociology of organizations.8 Management
chooses practices that guide the work of employees to promote the firm’s objectives. These practices address agency-managerial and technical-administrative problems.
Agency-managerial problems stem from the
fact that workers prefer, when they can, to pursue their own interests rather than the objectives established by management. For example, workers may prefer to emphasize different projects than those that maximize profit for the firm, or to work less intensely than how management would like to work. The consequence of such potential behavior may be loss of productivity, and certainly loss of profits. To combat this, when workers’ efforts or their results are easily observable, management requires and enforces work standards and guides directly employees’ behavior, and pays fixed wages. (We shall refer to this as the ‘old’ or ‘traditional’ organization of work). When workers’ efforts and direct results are not easily observable and when their behavior does not offer good clues about their productivity, management relinquishes direct control and grants workers some
discretion i same time i link worke monitors so keeps recor workers’ att over time, a of where the to this as the of work). Th resource p managerial productivit profitability for manage problems, combating likely to be profit gener Technica
from the fac and abilitie are not per firm in wh problems r profits, and providing ed (and suppl acquire skil ameliorate knowledge for learning from other managemen up to a poin the elimina problems, in practices’ co The hum assified into ature of the the practice he growth in ctices during 6). Clearly, te employee and financial lly since the milar growth th has been d half of the perhaps even
reached a plateau, by the middle of the 1990s. The most prevalent type of plan throughout the period was group-based participation in financial performance, with nearly two-thirds of respondents using this approach by 1996. The proportion of firms providing group-based participation in decision-making increased more than fourfold, from 11% to 48%. Although growth has occurred in all industries, it was most pronounced in the manufacturing sector and was least significant among firms engaged in commerce (the industry distribu-tion is not shown in Figure 1).
decision-making programs, which includes only employee discretion, is n about this category only for the year of the survey.
7. For a survey of the literature on the choice of human resource practice Ner, Kong and Liu (2001).
8. See Ben-Ner, Montias and Neuberger (1993) and Liu (1998).
n of Decision-Making (DM), Financial Returns (FR), & Supporting (SUP) Practices
ortion of Firms Having Each Type of Plan
5 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 individual-oriented FR group-oriented FR
SUP group-oriented SUP
ng (DM), financial returns (FR), & supporting (SUP)
developer’s tasks are significantly more complex and less predictable. Almost by definition, this situation demands that the employee exercise judgment and individual discretion rather than follow a simple set of rules or procedures handed down by a supervisor. Furthermore, such complex tasks will likely require consultation and cooperation with other software developers, as well as joint decision-making with co-workers in the context of a work team. Generally speaking, the greater the complexity of tasks and the higher the uncertainty of the outcome, the more useful it becomes to involve employees in decision-making.
If uncertainty influences the amount of employee involvement, the level of
inter-dependence between employees helps
deter-mine the scope – that is, whether these programs should operate at the individual or group level. In general, the more employees
rely on each useful group
Empirical re
For the sake focus here o returns pr practices, in information companies) the inciden same propo group-level 42% and 4 third of fir whereas alm level incent tion). A majori resource pra of horizonta both decisio nd agency-erse. Some ne problem, ddressed by ample, team anagement’s employees’ or delegating ployees (an from the to deal with n their work onsultation ministrative ent practices of a package actices may result. Two ombination e first is the which refers among the n resource hat decision-pported by ome form of , and that s should be for making al returns in making and have to be ng practices. ns vertical relationship evels of the mplies that should not up decision-in fdecision-inancial , then there ance in the ed practices,
and vice versa. A firm that has only decision-making participation practices violates the principle of horizontal consistency, whereas a firm that combines individual decision-making discretion with group financial returns, but lacks individual financial returns and group decision-making is both horizontally and vertically inconsistent. A firm that has all four practices may be inconsistent, unless the practices are very carefully designed so as not to contravene each other.
The intensity of the technical-administra-tive and agency-managerial problems varies with several contingencies. These might include such things as a firm’s business strategy; the technology of production; market forces; and the firm’s size, age, and ownership structure. For example, technology may affect the complexity of tasks, and the ability of supervisors to observe workers’ actions; firm size may affect the effectiveness of incentives that are subject to free-ridership problems, and so on. Consequently, the choice of human resource practices will vary among different firms because they generally face different contingencies.
Two major dimensions of task
environ-ment constitute the most important contingencies: the uncertainty of outcome of the employees’ efforts, and the interdepen-dence among employees’ work. Uncertainty
of outcome stems primarily from the
complexity of an employee’s tasks. Com-plexity prevents both the employee and his or her supervisor from being able to predict the exact outcome of a given task. To illustrate, let us compare a situation where the outcome is quite easy to predict – such as collecting tolls at a toll booth – with one where it is not – say, developing a new software application. The toll collector’s tasks are fairly simple, and the work can be governed by simple rules and monitoring. In contrast, the software
Table 1.
Consistency in the Choice of Decision-Making and Financ Human Resource Practices
Individual-level plans Decision-making but no financial return Financial returns but no decision-making Decision-making and financial returns p Neither decision-making nor financial re
Decision-making participation Financial returns participation
Group-level plans Decision-making but no financial return Financial returns but no decision-making Decision-making and financial returns p Neither decision-making nor financial re
Decision-making participation Financial returns participation
and in a preliminary fashion the consequences of these practices for their putative intended beneficiaries, the owners, as well as for workers.10
Theoretical considerations
All stakeholders in a firm, primarily owners and their employees, draw their benefits from
the same eco of the firm. in maximizi given size o increasing i the other pa their benefi actions that vel, whereas d 33% have up level, and nconsistency t in 44% of ve decision-ncial returns ave financial ng). At the ge of firms A significant h individual r or not this consistency g of different ithout more briefly the of human only on the ssed in the ble 2, the es providing cipation in returns. The riptors of a rtainty and nd transfer-xecute these ed on Likert ittle’ and 5 nce of tasks,
out the tasks, uncertainty he sum of ks, the range eral control age of firm, ). Given the endent
vari-ables (coded 0 if no practice of a given type is employed in a firm, 1 if at least one practice of that type is employed), our estimation method is logistic regression.
The results highlight the importance of task uncertainty in shaping human resource practices, as indicated by the results for the individual and group-level decision-making involvement practices. Interdependence among the tasks of employees is also impor-tant (but statistically somewhat less signi-ficant); it has the expected positive effect on the presence of group decision-making practices, and surprisingly, an even more significant effect on individual involvement in decision-making. However, neither uncertainty nor interdependence have apparently direct bearing on firms’ adoption of financial returns plans. The level of skills is, as expected, positively associated with individual discretion in decision-making, but has no relationship with other dependent variables.
Outcomes for Owners and Workers
The operation of any firm is directed by the desire to attain certain objectives. Typically, for-profit firms are operated on behalf of their owners to maximize profits (or the value of their firms). This is done under a variety of constraints, including the need to hire managers and workers and the resulting agency-managerial and technical-administra-tive problems. The previous section has dealt with human resource practices that are instituted in order to ameliorate these problems, as well as with the underlying factors that affect the reliance on various practices. In this section we explore brieflyn-making nor financial returns plans are considered to have consistent
hor-those companies that have both types of plans. 10. This important question is the focus of our present research efforts unControl and Prevention and the National Institute for Occupational S
Table 2.
The Choice of Human Resource Practices. Logistic Regressions
Decision making Decision making Fin induvidual level group level ind
Task environment uncertainty 0.43*** 0.31*** 0 (0.12) (0.11) (0 interdependence 0.18** 0.14* 0 (0.09) (0.09) (0 Skills transferability 0.05 0.04 0 (0.08) (0.08) (0 level 0.26** 0.07 0 (0.11) (0.10) (0 Control variables number of -0.00 0.16** 0 employees (log) (0.00) (.07) (0 unionization -0.08 0.26 0 (dummy) (0.22) (0.22) (0 firm age 0.00 0.00 -0 (0.00) (0.01) (0 firm age 0.00 0.00 -0 squared (0.00) (0.00) (0 Industry commerce -0.20 -0.07 0 (dummy) (0.19) (0.17) (0 services -0.32 -0.32 0 (dummy) (0.23) (0.22) (0 Number of obs. 749 749 749 Prob. > chi2 0.00 0.00 0 Log-likelihood -448.4 -494.6 -475 Notes:
1. Estimated standard errors are in parentheses.
ell-being of dimensions, occupational m work, and ay seek to ng wherever ven wages, their work es that were of the pie as employees. hese effects theoretical ces; here we ons.11Recall are selected t and other mal set of w of owners) l in another cies. And the me may not adverse to, what might e adverse to come will be r choice of ctors; these mpensating d gain from y associated ce of safety ted loss of rker absence kers, on the n the choice if they have
decision-making discretion, and they are likely to emphasize safety over productivity efforts if they do not have a share in firm financial returns.
Empirical results
Table 3 summarizes the results of regressions of outcomes for firms, owners, and workers, using the same cross-sectional dataset that we used in the regressions underlying Table 2. In order to present the results succinctly as possible, a similar estimating framework is maintained across different outcome measures. The regressions include almost the same the set of independent variables: measures of human resource practices, controls for task environment and skills, and controls for firm size, industry, and union status; in the regressions represented by the first two columns we also included firm assets.12We
present estimates only for decision-making and financial returns practices. The table includes two sets of regressions for each dependent variable. In one regression we include variables that identify the presence of decision-making and financial returns practices at the individual and group levels (the variable names, parameter estimates, F and R-squared statistics are italicized and appear on the right side of each column). In the other regression we included also interaction effects, written out in detail, with the omitted category being “no decision-making or financial returns participation plans.” The parameter estimates on the variables not included in Table 3 are very similar across the two regressions. All regressions were estimated by OLS.
n (1995), Park (1997), and Kong (2001) for more elaborate treatments of
for presentational purposes, but at the same time it acts as a straightjacket rmined in somewhat different processes which entail different estimating methods.
Table 3.
Outcomes for Workers and Owners: OLS Regressions. Dep
Productivity Profitability Wages
Individual-level plans
Decision-making but -.439* -4.74 .052 no financial returns (.262) (4.80) (.049)
Financial returns but .023 2.13 .094*
no decision-making (.264) (4.78) (.053) Decision-making and -.159 -15.28*** .204* financial returns (.258) (4.99) (.060) Decision-making -.312* -10.48*** participation (.187) (3.70) (. Financial returns .147 -2.53 participation (.193) (3.67) (.
Group -level plans
Decision-making but -.026 -9.57 -.189*
no financial returns (.497) (8.49) (.069)
Financial returns but .169 -2.85 .071 no decision-making (.290) (5.48) (.052) Decision-making and .194 -2.29 .010 financial returns (.300) (5.67) (.056) Decision-making .023 -1.31 -participation (.186) (3.44) (. Financial returns .237 -1.38 participation (.238) (4.52) (. Number of observations 98 91 606 Mean of dependent variable 17.67 6.65 10.21 Prob>F 0.000 0.000 0.000 0.000 0.00000 0 R squared .908 .516 .369 .908 .483 Notes:
1. The dependent variable in the productivity regression is log sales; in profitability it is afte extraordinary items divided by common equity); in wages it is log of average wage; in injur employment variability it is the coefficient of monthly employment variation during the ye 2. All regressions include the following control variables: employment (log), union (dummy), - individual level (dummy), support practices - group level (dummy), skill complexity, and and transferability (1-5 Likert scales). The productivity and shareholder economic returns 3. The number of observations varies because of missing observations. The small number of ob
that dependent-variable and asset information is available only for public firms. 4. Two regressions were run for each dependent variable. In one, variables that identify the
practices at the individual and group levels were included; the independent variables, param regressions are italicized and right-adjusted in each column. In the other regression interac is “no decision-making or financial returns participation plans.”
flexibility in production. Although the changes have been sweeping, they have not affected all firms equally. Many occupations – such as those of line workers in manufacturing, services, and commerce – have become more complex and inter-dependent. In addition, the task of organizing workers has become much more complex than in the times of relative stability in the economic and technological environments. The flexibility that is being demanded of most employees is also required of those in positions of responsibility.
In view of the fact that managers choose human resource practices on behalf of shareholders to optimize their returns, it is surprising that the majority of the firms in our sample do not seem to benefit from their human resource practices neither in productivity nor in profitability. Individual discretion in decision-making appears to be particularly deleterious to both these impor-tant outcomes. Group-oriented practices have small negative – but statistically insignificant – effects on profitability and moderate to large effects, though again but statistically insignificant, on productivity (a 19.4% return on a combination of group decision-making and financial returns participation plans). These results must be qualified severely, given the low number of observations, the fact that the human resource practices have been summarized here rather coarsely, the great number of additional factors affecting such gross firm-level aggregates as productivity and profitability, and possible econometric
specification do have a tr Why do fir practices? P right answe work, they or they mak or implicit resource pra Workers by the hum adverse to individual e with high incentives, profitability another var Group dec returns is no seems to do either, and with greater impact on e This pap attempt to several imp organizatio Many more a richer and not be accom paper. The s is that more this data se firmly what practices on xied by the y, sales. The bb-Douglas ted by the nd control he decision-plans are he exception on-making, ive and large clusion here ccompanied is associated ing perhaps n is used for rs to be little ain outcome by after-tax s divided by eneral con-at individual mpanied by ociated with is important uctivity and ry, and that all. re captured rates, and ual decision-with signi-ially when ncentives.14 ut decision-with higher
wages. Participation in group financial returns is also associated with higher wages, but not so with group decision-making. In fact, group decision-making without group incentives is associated with lower wages than no decision-making and no financial returns plans at all.
Injury rates are slightly – but without statistical significance – lower in firms with individual discretion, and somewhat higher, with low statistical significance, in firms with group decision-making participation.15
Finally, employment stability does not appear to be significantly associated with our measures of decision-making or financial returns participation, with the exception of individual incentives, which are associated with lesser stability (greater variability).16
Conclusions
The organization of work underwent a thorough transformation since the early 1980s. By and large, the typical new workplace relies more heavily on employee involvement in both decision-making and in firm performance, requires greater worker skills, and entails more complex tasks than the prevalent old workplace. A number of changes in the economy are responsible for these transformations, but the single most important proximate factor has been the increased complexity of tasks and the resulting uncertainty of outcomes. The driving forces behind this change are the increased reliance on computer-based technologies, and the change in business strategies towards greater
capital .27, well within the common range.
ty and skill were included, and the estimates on both are positive and
high-otential workers’ exaggerated reporting of claims along with true safety lev-ay be associated with the type of human resource practices in a company.
nt results concerning the effects of human resource practices on outcomes m et al. (2000) and Baker (1999). For a summary of work of the literature
-Ner et al. (1996).
17. Using these and other metrics of company success we found in other w a favorable effect. It is worth noting here that dropping the task envi not change the direction of the results.
References
Appelbaum et al. 2000. Manufacturing Advantage: Why
High-Performance Work Systems Pay Off, Cornell
University Press, Ithaca.
Baker, T. 1999. Doing Well by Doing Good: The Bottom
Line on Workplace Practices, Economic Policy
Institute, Washington D.C.
Ben-Ner, A., J. M. Montias, and E. Neuberger. 1993. «Basic Issues in Organizations: A Comparative Perspective,” Journal of Comparative Economics 17(2): 207-242.
Ben-Ner, A. and D. Jones. 1995. «Employee Partici-pation, Ownership, and Productivity: A Theoreti-cal Framework,» Industrial Relations, 34(4): 532-554.
Ben-Ner, A., T. Han, and D. Jones. 1996. «The Productivity Effects of Employee Participation in Control and in Economic Returns: A Review of Econometric Studies,» in Ugo Pagano and Robert E. Rowthorn (eds.), Democracy and Efficiency in
Economic Enterprises, Routledge, 209-244.
Ben-Ner, A., F. Kong, T. Han, N. Liu, Y. Park, and S. J. Smela, 2001, “What Works at Work? Evidence from the Minnesota Human Resources Management Practices Study,” The CURA Reporter May 2001(May): 9-17, www.cura.umn.edu/reporter/01-May.html
Ben-Ner, A., F. Kong, and N. Liu. 2001. “On The Choice of Human resource practices and Workplace Organization,” Working Paper, Industrial Relations Center, University of Minnesota.
Cappelli, P., and D. Neumark. 2001, “Do High-Performance Work Practices Improve Establish-ment-Level Outcomes?” Industrial and Labor
Relations Review. 54(4):737-775.
Han, T. 1995. «Employee Participation in Decision-Making and Financial Returns: Effects on Firm Performance,» Ph.D. Thesis, Industrial Relations Center, University of Minnesota.
Hart, O. and J. Moore. 1990. “Property rights and the nature of the firm,” Journal of Political Economy, 98: 1119-1158.
Ichniowski, C., T. Kochan, D. Levine, C. Olson, and G. Strauss. 1996. “What Works at Work: Overview and Assessment,” Industrial Relations, 35(2): 299-333.
Kong, F. 2001. «The Effects of Profit Sharing Plans and ESOPs on Firm Employment Fluctuations,» Ph.D. Thesis, Industrial Relations Center, University of Minnesota.
Liu, N. 1998. «Determinants of Innovative Human resource practices and Systems,» Ph.D. Thesis, Industrial Relations Center, University of Minnesota.
Park, Y. 1997. «Occupational Safety: Effects of Employee Participation Plans in Decision-making and Financial Returns,» Ph.D. Thesis, Industrial Relations Center, University of Minnesota.
