BUREAUCRACY OR HIGH PERFORMANCE: WHAT ARE
THEY AND WHO ADOPT THEM*
Song Yang
Assistant Professor
Department of Sociology and Criminal Justice
211 Old Main
University of Arkansas
Fayetteville, AR 72701
June 2003
*Direct all correspondence to Song Yang via email [email protected] or
write to 211 Old Main, Department of Sociology and Criminal Justice, University
of Arkansas, Fayetteville, Arkansas 72701
BUREAUCRACY OR HIGH PERFORMANCE: WHAT ARE
THEY AND WHO ADOPT THEM*
Abstract
At the time of increased work reorganization, I investigate the components of bureaucratic and
high performance work systems, and source of variation in organization adoption of bureaucratic
and high performance practices. Findings from the 1996 National Organizations Survey suggest
that while degree of workplace formalization, level of hierarchy, and number of
departmentalization strongly indicate bureaucratic organizations, the alleged high performance
indicators of teamwork, skill enhancement, job autonomy, and innovative pay structures do not
glue together to identify high performance organizations. Multivariate analyses reveal that size,
domestic and foreign market competition, and organization performance pressures are related to
work reorganization. In general, large organizations under low market competition and low
performance pressure tend to stick with bureaucratic systems, whereas large size, along with
high level of domestic and international market competition plus tremendous performance
pressures, leads to adoption of many high performance practices. The implications for future
studies on work organization are also discussed.
1
BUREAUCRACY OR HIGH PERFORMANCE: WHAT ARE
THEY AND WHO ADOPT THEM*
Despite decades long studies on work structure transformation, researchers lack
consensus on what constitute the transformed organizations, which are often labeled as high
performance work systems (Osterman 1994; Kalleberg and Moody 1996; Cappelli and Neumark
2001; Lawler, Mohrman, and Ledford 1995; Keller 1978; Godard and Delaney 2000; Huselid
1995; Cappelli et al. 1997: chpt. 3; Appelbaum and Batt 1994; Doeringer et al. 1991: chpt. 9). A
study on American workplace human resource practices using the 1991 National Organizations
Survey measured high performance work systems with four indicators of decentralization, job
training, performance-based compensation, and firm internal labor market (Kalleberg and Moody
1996), whereas another study using two waves of survey of American manufacturing
establishments measured high performance work systems with five categories: self-directed work
team, job rotation, quality control, total quality management, statistic process control (Osterman
1994; Osterman 2000). Indeed, a recent review of the new paradigm of human resource
management practices identified several shortcomings in this line work, one of which is that high
performance work systems are measured based on researchers’ judgment or preference (Godard
and Delaney 2000).
Although researchers discuss that high performance work systems accomplish labor
control and coordination in contrast to the methods in bureaucratic work systems (Cappelli et al.
chpt. 3), few researchers empirically investigate the changing workplace structures in relation to
the traditional bureaucratic work systems. Research studies suggest that high performance work
systems contradict bureaucratic work management in human resources practices. Bureaucratic
2
organizations adopt rigid work structures and task assignments to establish a assembly line
model and facilitate mass production (Scott 1998; Appelbaum and Batt 1994, chpt 1). In
contrast, the new economy is characterized by product customerization and constant innovation
(Knoke 2001; Appelbaum and Batt 1994; chpt 8), which provides fertile ground for high
performance work systems with flattened hierarchy, democratic work structures, and equal
participation in the decision-making from all levels of workers (Osterman 1994; Kalleberg and
Moody 1996; Lawler, Mohrman, and Ledford 1995). However in recent empirical research on
work systems, a set of essential indicators for bureaucratic work systems, such as personnel
selection, performance evaluation, job design, grievance procedures, and promotion criteria
(Marsden, Cook, and Kalleberg 1996), are also used to measure high performance work systems
(Huselid 1995). Despite the empirical evidence that bureaucratic and high performance practices
may simultaneously present in one work organization (Yang 2003), a clarification to demarcate
the difference between bureaucracy and high performance work systems and to define their
respective components is crucial to advance scholarly undertake in this vein of work (Godard
and Delaney 2000).
In this paper, I first review previous literature on organizational transformation, focusing
on those defining characteristics for traditional bureaucratic work organizations and transformed
high performance work systems, and the driving factors for organizations to adopt the two
distinctive work structures. I then conducted the confirmatory factor analyses and multivariate
regression to address the two issues with a national representative sample of U.S. work
establishments. Respectively the two issues are 1) the relationship between bureaucratic and high
performance work systems and their respective components; 2) the explanatory factors for the
3
levels of bureaucratization and high performance practices adopted in American work
organizations.
Bureaucracy and High Performance Work Systems: What are They?
Witnessing the changing landscape of world economy with increasingly globalized
competition (Knoke 2001), a recent surge of research studies on work organizations focus on
how organizations reconstruct their business conducts and human resources management in
respond to the new economic challenges (Osterman 1994, 2000; Cappelli et al. 1997; Doeringer
et al. 1991; Appelbaum and Batt 1994; Kalleberg et al. 1996). In a striking consistency,
researchers noted that transformed high performance work systems contradict bureaucratic
organizations various in aspects of human resources management (see Appelbaum and Batt
1994: chpts 1, 2 and 3; Cappelli et al. chpt. 3).
Central to the bureaucratic management is the use of elaborated rules, policy, and
regulations to govern worker-employer relations, to classify job at hierarchical and lateral levels,
and to formally define job responsibilities. A study documented how a bureaucratic workplace
used formal rules to create a fine-graded division and stratification among workers, to evaluate
job performance, determine promotion, and spell out firing or hiring procedures (Edwards 1979:
131-162). The bureaucratic labor control and coordination fits nicely with mass production,
which needs an efficient and formalized method to organize a large number of workforce
(Marsden, Cook, and Kalleberg 1996). However, since 1970s, work organizations operating on
mass production have received increasing challenges in their core product market or service
domain from those operating on transformed work structures (Appelbaum and Batt 1994:1-25).
Globally, US employees become the most expensive labor force that even with the mass
production that intends to minimize labor cost by maximizing output per worker, wages for
4
workers in many other countries are only a fraction of those of American workers. Moreover, the
proliferation of microprocessor-based technology not only increased workers’ productivity
worldwide, but also facilitated a fast product delivery and a quick response in the product or
service design to accommodate customers’ idiosyncratic needs (Appelbaum and Batt 1994:1-25).
For many American plants that were benefiting from mass production operations, the needs to
look for the alternatives to mass production and associated bureaucratic labor control are
pressing.
American employers striving to reform the mass product methods and bureaucratic
control often embarked on many different routes. Earlier, researchers documented that American
human resources model evolved in U.S. organizations since 1950s to facilitate employeremployee communications and employee involvement in corporate business (Kochan, Katz, and
McKersie 1986). Later, a new labor management called flexible mass production aims at
reducing costs by cutting full time labor force, making great use of micro-processor technology,
and increasing low-paid, temporary, and part-time workers. However, neither of these two
approaches fundamentally changed the bureaucratic management in many mass production
plants, which continue to use the fine-grained hierarchical and lateral labor divisions, and
formalized rules and regulations to manage its labor force (Appelbaum and Batt 1994).
The intensified global market competition accelerates American workplace restructuring
and exposes several failures of the bureaucratic work system in response to the new market
conditions. To improve quality and respond to customers’ needs more effectively, frontline
workers need more power to make quick decisions. To accomplish this power relegation
employers eliminate many intermediate managers, which also serves to cut much labor cost. To
facilitate smooth operations for workforce down in the hierarchy, their job definitions need to be
5
flexible, job discretion needs to be expanded, skills need to be constantly upgraded, and their
high performance needs to be rewarded (Cappelli et al. 1997:90-91). Organizations adopting a
cluster of innovative work practices such as teamwork, job rotation, performance-based salary
system, quality circle through statistical control programs, power relegation, and total quality
management are often called high performance work systems (Osterman 1994; Kalleberg and
Moody 1996; Cappelli and Neumark 2001; Lawler, Mohrman, and Ledford 1995; Keller 1978;
Godard and Delaney 2000; Huselid 1995; Cappelli et al. 1997: chpt. 3; Appelbaum and Batt
1994; Doeringer et al. 1991: chpt. 9). However, as previously discussed, high performance work
system is empirical measured with a set of indicators overlapping but different from each other.
In this study, I formally define the concept of bureaucratic and high performance work
systems with two respective sets of indicators. Central to the bureaucratic labor control is a set of
formal rules and corporate policy that created a fine-grained classification based on workers’
specialty and managerial responsibilities (Edwards 1979; Marsden, Cook, and Kalleberg. 1996).
In contrast, high performance work organizations de-formalize the governing structure and declassify workforce by extensive use of teamwork to integrate workers with different specialty
and responsibilities, empower workers down in the ranks with greater job discretion and
autonomy, reward workers by linking their performance to their pay scale, and upgrade and
expand their skills via job rotation, work transfer, and cross training programs. The following
three hypotheses summarize the above discussion.
H1: Bureaucratic organization is indicated by the level of hierarchy, the lateral
departmentalization, and workplace formalization.
H2: High performance work system is indicated by teamwork, skill upgrading, worker
empowerment programs, and performance based pay structures.
6
Bureaucracy or High Performance: Who Adopt Them?
The following sections are devoted to uncover why some organizations stick to the
traditional bureaucratic work structures while others adopted innovative high performance
practices.
Size and workplace structures
Over decades researchers found a number of important characteristics that induce
workplaces to adopt different work structures. Early work in this vein found that size matters in
organizational structuring. Large organizations have great needs to specialize and subdivide
work to maximize economies of scale (Blau 1970, 1972). Growing number of subdivisions
within large organizations also require corresponding increase in managerial personnel to
coordinate and integrate efforts from each individual subunit (Lawrence and Lorsch 1967).
Compelling evidence from studies during that time suggested that size is the most important
predictor for organizational development of a large number of specialized departments, tall
hierarchy, and formalized governing structures with written rules and documents (Blau and
Schoenherr 1971; Meyer 1972; Child 1973; Mansfield 1973).
Less clear is how organization size affects workplace adoption of high performance
practices. On the one hand, large organizations face great need to adopt high performance work
practices. Common to large organizations is motivation problem, which results from workers’
frustration over impersonal management and insignificant work, as they often cannot see the
impact of their work on the customer and on company performance (Lawler et al.1995: 97-129).
Moreover large organizations commonly have great resources to advocate a work structural
transformation and to sustain short term financial loss in hope of obtaining long term benefits,
7
which is crucial for implementing high performance practices (Lawler et al.1995: 97-129;
Osterman 1994; Appelbaum and Batt 1994: 8). On the other hand, large organizations also
embrace very rigid structures that impede the dissemination of successful restructuring
throughout. Often times restructuring in large organizations take place in different divisions
without a central coordination overall, which results in that workers ignore or resist change
efforts (Lawler et al.1995: 97-129). In sum, although size is related to organizational adoption of
bureaucratic and high performance practices, the direction of the effect from size to bureaucracy
is much clear than the effect from size to high performance.
H3: The larger the organization, the greater the organizational bureaucratization.
Market and Workplace Restructuring
Increased competition in domestic and foreign market drives U.S. organizations to
adopted high performance work practices via several different mechanisms. First, organizations
operating under tremendous competition are despaired to outperform their competitors by
improving product quality and cutting cost. To accomplish these, organizations indeed need to
adopt innovative high performance work practices (Osterman 1994). Second, American
companies competing with foreign firms are exposed to various innovative work practices
adopted by their international competitors. Indeed, a study documented four innovative work
systems prevalent in foreign firms: Swedish sociotechnical systems, Japanese lean production,
Italian flexible specialization, and German diversified quality process (Appelbaum and Batt
1994: Chpt. 3 and 4). It is highly plausible to speculate that American establishments operating
in international market will try to emulate their international competitors in adopting those high
8
performance practices. Empirical work supplied compelling evidence that increase in market
competition stimulates organizational adoption of high performance practices (Osterman 1994;
Lawler et al. 1995:117-123).
H4: The higher the domestic market competition, the greater use of high performance
organizational practices.
H5: The higher the international market competition, the greater use of high performance
organizational practices.
Although very little empirical research is done on how market competition affects
organizational bureaucratization, much theoretical discussion suggests that market competition
decreases bureaucratization (Appelbaum and Batt 1994: chpt. 2; Cappelli et al.1997: chpt. 3;
Kalleberg and Moody 1996). Increased market competition requires that organizations swiftly
respond to customers’ idiosyncratic need, constantly improve product quality, all of which run
counter to bureaucratic work structures that facilitate mass production of large sum of
standardized goods. International exposures to many innovative work practices vividly illustrate
efficient alternatives to bureaucratic work models, thus providing a learning opportunity for U.S.
firms to change the bureaucratic systems.
H6: The higher the domestic market competition, the lower the organizational
bureaucratization.
H7: The higher the international market competition, the lower the organizational
bureaucratization.
9
Performance Pressure and Organization Restructuring
Scarce research is done on how performance pressure affects work reorganization. But
the effect of performance pressure on work restructuring may not as straightforward. On the one
hand, organizations encountering performance problems, financial losses, or economic crisis
may be despaired to look for alternatives to overcome those problems. On the other hand,
performance problems can also cause financial drainage that organizations lack sufficient
resources to sponsor a fundamental work restructuring. A study succinctly summarizes that
successful implementation of high performance practices involves companies that experienced
performance problems, and sufficient resources to support extensive investment in work
reorganization and to incur possible short-term financial loss in order to gain long term gains
(Appelbaum and Batt 1994: 8). To exam this effect, I include in the equations predicting work
restructuring the performance problems with quality of product or services, with new product or
services, with retaining essential employees, and with customers’ satisfactions.
Data
To test my hypotheses, I analyze data from a national survey of diverse work
establishments, the 1996-97 National Organizations Study (NOS) (Kalleberg, Knoke, and
Marsden 1999). The NOS was a computer-assisted telephone interview (CATI) survey
conducted in 1996-1997 by interviewers from the Minnesota Center for Survey Research
(MCSR) at the University of Minnesota. The sample was drawn from the Dun and Bradstreet
(D&B) Market Identifiers list of approximately 15 million establishments. D&B stratified the
sample into 40 categories of increasingly broader employee-size ranges, from 1-4 to 4,000-4,999
and 5,000 or more. Within each stratum, D&B selected cases for MCSR to interview that were
10
sampled randomly with probability proportional to size (PPS). This procedure was designed to
obtain a final sample whose size distribution would reflect the establishments experienced by
typical employees. MCSR conducted most of the CATI interviews with the human resources
manager or a functionally equivalent informant from each establishment. The survey completion
rate was 54.6% of eligible establishments, of which 86% were done by CATI and 14% by mail
questionnaire. The final sample contains data from 1,002 establishments.
Measures
This section describes measures of variables in my analyses. Table 1 discusses in detail
the construction of all the variables.
Dependent variables
Organizational bureaucratization consists of three items: (1) formalization, which is
indicated by five questionnaire items such as written job description, written record of job
performance, employment contracts, written documents of personnel evaluation, documents of
hiring/firing procedures; (2) levels, indicated by responses to the question “how many levels are
there between a first-line supervisor and the top official in your organizations?” (3)
departmentalization, indicated by questions asking whether organizations have separate
departments for personnel and labor relations, health and safety departments, and equal
opportunity or affirmative action.
High performance organization consists of four items: (1) teamwork is measured by a
question “does your organization have established committees made up of workers and managers
who meet regularly to deal with issues such as new technology, quality improvement, production
or service delivery, health and safety, or do workers meet by themselves without management to
11
discuss issues related to production or service delivery?” (2) skill is measured by four questions
“are core workers cross-trained?” “are core workers involved in job rotation?” “are core worker
involved in statistical process control?” and “how often core workers transfer to another job
family?” (3) autonomy is measured by two questions “how much choice do core workers have
concerning the best way to accomplish their assignments?” and “which best describes how
closely core workers are supervised as they do their work?” (4) compensation is measured by
two questions “are core workers paid using group incentives such as gain-sharing?” and “do core
workers participate in a profit-sharing or bonus program?”
Independent variables
Size is measured by taking natural log of the total number of full time and part time
employees. Market competition is measured by a single questionnaire item “how much
competition would you say there is in your main market or service area (none=1, very little =2, a
moderate amount = 3, a great deal = 4)?” Foreign market competition is indicated by “how much
competition would you say there is in your main market or service area from foreign
organizations (none=1, very little = 2, a moderate amount = 3, a great deal = 4)?” Performance
problem with the quality of product or services is measured with a single item “how would you
compare organization’s performance in the quality of the product or service to that of other
organizations that do the same kind of work? (much better = 1; somewhat better = 2; about the
same = 3; somewhat worse = 4; much worse = 5)” Performance problems with new product or
service, with retaining essential workers, and with customer satisfactions are measured with
similar questionnaire items with the same coding scale. To eliminate the possibility of
spuriousness between the dependent variables and the key independent variables under
investigation, I included in the multivariate equation the controlled variables such as
12
organization age, organization status (organizations with parents = 0, complete independent
organization = 1), and organizational market area (regional/national/international =1;
neighborhood/city/state = 0), and organizational industry (manufacturing = 1; others = 0).
Findings
To empirically exam components of the latent factors bureaucratic and high performance
organizations, I conducted the confirmatory factor analysis (CFA) with the two latent factors and
seven indicators (for more on CFA, see Knoke, Bohrnstedt and Mee 2003: Chpt. 12). Figure 1
lists the AMOS estimates of standardized confirmatory factor loadings on the two latent factors
(Byrne 2000). Various indicators such as GFI (0.996), AGFI (0.967), and RMSEA (0.053)
suggest a close fit between the theoretical model and the sample data. The negative covariance of
–0.27 between the bureaucracy and high performance shows that organizations score high on
bureaucratization tends to score low on high performance adoption. This result suggests that
bureaucracy and high performance are exclusive to each other in the process of work
reorganizations. Instead of using a combination of bureaucratic and high performance practices,
organizations tend to adopt either one of them in their workforce management. The standardized
factor loadings of formalization (0.74), hierarchy (0.54), and departmentalization (0.72) on the
latent factor “bureaucracy” are fairly high, suggesting a strong indication of bureaucracy from
the three indicators, thus lending empirical support to H1. In contrast, only teamwork (0.93)
loads highly on the latent construct high performance organization. Skill enhancement (0.19),
performance-based compensation (0.18), and job autonomy (0.08) have low factor loadings,
which do not support the H2 that the latent construct of high performance work system is
indicated by all four indicators. This result suggests that while bureaucracy has a set of well-
13
defined indicators, high performance does not have a set of strong indicators well clustered under
the popular label. Hence, instead of analyzing the variation in adopting the latent factor high
performance work system as a single construct, I regress its four components, teamwork, skill,
compensation, and job discretion, on the set of explanatory variables under investigation in the
regression analyses section.
Table 2 lists descriptive statistics of the variables in the regression equation. The results
show that American work establishments are fairly formalized, indicated by an average score of
3.25 in the formalization measure. The mean scores for the other two bureaucratic indicators,
hierarchy and departmentalization are 2.45 and 1.83 respectively. In the four high performance
measures, the average establishments in the NOS sample have approximate 3 teamwork
committees to deal with corporate issues regarding technology, product quality, delivery, health
and safety. They also have more than 1 skill enhancement programs, provide workers with “more
than moderate amount” of job autonomy, and use no innovative pay programs. On the average,
the sampled NOS work establishment hired 1,000 workers (Exp.(6.90)=993). Organizations
reported to have “more than moderate amount of competition” in their operating market, as
indicated in the average score of 3.28 in the market competition measure. The level of
international market competition is lower then the main market competition as the NOS work
establishments are facing less than moderate level (average of 1.77 in the measurement) in
foreign market competition. The NOS organizations are not under tremendous performance
pressure as the average scores on product/service quality (1.69), new product (2.06), customer
service (1.83), and retaining essential workers (2.18) all suggest that organizations perform
approximately “somewhat better” than their competitors in those dimensions. Space restraint
14
prevents full documentation of all variables in the analysis, please see table 2 for the detailed
description.
Table 3 displays unstandardized coefficients from OLS regression of organizational
adoption of bureaucratic work structures and high performance job autonomy dimension, and
Negative Binomial regression of high performance teamwork, skills enhancement, and
performance-based pay programs (For details on Negative Binomial Regression, see Paul Allison
1991: Chpt. 9). In explaining level of bureaucratization, table 3 shows that each unit of increase
in organizational size increases organizational bureaucratization by 0.266 units, which supports
H3 that large organizations are more bureaucratized than do small establishments. In support of
H6 that the higher the market competition, the lower the organizational bureaucratization, table 3
shows that each unit of increase in market competition reduces organization bureaucratization by
0.093 units. In contrast, empirical results show that foreign market competition is not related to
organizational bureaucratization, which did not support H7 that predicts a positive relationship
between the two variables. In partial support of H4 that market competition stimulates
organizations to adopt more high performance practices, market competition is only related to
high performance job autonomy dimensions. Table 3 shows that each unit of increase in market
competition increases workers’ discretion by 0.214 units. Also in partial support of H5 that
foreign market competition compels organizations to adopt more high performance practices,
foreign market competition in only related to skills enhancement. Each unit of increase in foreign
market competition increases skill enhancement programs by 6.6% (100 * (e0.064 -1) = 6.6%).
The results also indicate that performance pressures in producing new product and retaining
essential worker are related to workplace structuring. Each unit of increase in performance
pressure on new product reduces organizational bureaucratization by 0.128 units. In contrast, it
15
increases organizational adoption of high performance teamwork and compensation programs
respectively by 9.4% (100 * (e0.090 -1) = 9.4%) and 4.9% (100 * (e0.048 -1) = 4.9%). Performance
pressure on retaining essential workers also impels organizations to implement more
performance-based salary programs. Each unit of increase in pressure to retain essential workers
increases the performance-pay programs by 5.5% (100 * (e0.054 -1) = 5.5%).
Several other findings are worth noting. Size is linked to high performance teamwork and
skill enhancement. Large organizations tend to have more teamwork and skill increasing
programs than do small workplaces. Manufacturing companies are much less bureaucratized but
embracing more skill enhancement programs than do establishments operating in other
industries. Independent organizations are much more bureaucratized than dependent
organizations with parents. Organizations with market area at regional, national, or international
levels relegate much power to give their workers great discretion than do organizations operating
with a local, city, or state level market. With only 11 predictors, model 1 explains a large portion
of 51.9% variation in organizational bureaucratization.
Overall the results show that bureaucratization is indicated by a set of distinctively
defined indicators such as formalization, vertical hierarchical layers, and lateral
departmentalization, lending support to previous study that bureaucratic work systems use formal
policies and fine-grained classification at lateral and vertical dimensions to manage its workforce
(Edwards 1979). However, confirmatory factor analyses of indicators drawn from previous
research on high performance work systems show that the four high performance indicators of
teamwork, skills enhancement, job autonomy, and performance pay programs do not glue
together to define a high performance organization. This result is in consistency with previous
empirical findings (Osterman 1994) and theoretical discussions that American employers are
16
implementing high performance practices on a gradual piecemeal basis, instead of deploying a
lump-sum batch of innovative programs simultaneously (Appelbaum and Batt 1994). Moreover,
highly bureaucratized establishments tend to have lesser high performance practices, and vice-avise.
Evidences also show that size, main and foreign market competition, performance
pressures from various domains are related to the structuring of work establishments, net of
effects from other controlling variables such as organization market area, age, industry, and
independency. Large sizable workplaces are more bureaucratized than are small establishments,
whereas organizations facing great market competition and performance pressure on new
product are less bureaucratized than the other organizations. The effects of market and foreign
market competition, and performance pressures on high performance programs are less
consistent and depend on which high performance practice entailed. Market competition compels
employers to relegate more power to their workers; employers facing foreign competition
increase use of skill enhancement programs; performance pressures in new product stimulate
employers to use more teamwork and performance-based pay; employers with problems
retaining workers use more performance-based pay programs than do their peers.
Discussions
Using the 1996 national organizations survey of diverse work establishments in U.S., this
study made considerable contributions to the literature of workplace reorganization. First,
although theoretical discussions on workplace transformation from bureaucracy to high
performance practices are abundant, empirical investigation on their respective components and
correlation is scarce. Filling this gap, this study showed that organizations with high level of
17
bureaucratic controls tend to adopt less of high performance practices. Vis-à-vis, organizations
adopting many high performance practices tend to discard traditional bureaucratic controls. This
finding essentially agrees with previous literature that the growth of high performance practices
is associated with the decline of bureaucratic work management. In particular, high performance
teamwork replaces bureaucratic hierarchy and departmentalization to integrate, rather than to
classify and categorize, workers in corporate functions such as quality control, problem-solving,
and decision-making. High performance power relegation, cross training and job rotation also
break many bureaucratic policies and regulatory codes to facilitate workers’ response to
idiosyncratic occasions that defy easy predictions.
Second, findings suggest that the alleged high performance practices of teamwork, skills
enhancement, job autonomy, and performance-based pay do not consistently clue together to
define a high performance organization. Workplace transformation takes place piecemeal as
employers adopt a fragment from the menu of a set of high performance practices that are
alternative to bureaucratic controls. The prohibitive cost associated with the comprehensive
adoption of high performance practices may deter organizations from fully implementing the
structural transformation. Pertinent to American companies and economic conditions, there are
some idiosyncratic factors that hamper the work structural transformation in U.S. Top managers
in American companies work to satisfy the demands of portfolio investors, who are less likely to
support a fundamental work reorganization that takes a long time to produce a hard-to-measure
result. Operating in the U.S. economic context, where little guidance is given regarding a
systematic implementation of work reorganizations (Appelbaum and Batt 1994:151), top
managers face tremendous uncertainties and receive little incentives to change the existing
programs. Finally, decades of economic success relying on mass production and bureaucratic
18
controls produce structural inertia for many American companies to stay with bureaucratic
management. For example, a rigidly-defined managerial prerogative stipulates that power,
authority, and decision making is uncharted territory for managers, who are not at ease with
power delegation to low rank workers.
Third, this study identified several sources that explain variation in organization
structural formation. Multivariate regression analyses reveal that large organizations facing low
market competition and low pressure of new product development tend to be highly
bureaucratized, whereas factors including large size, high domestic market and foreign market
competition, and performance pressure from new product development and retaining essential
workers impel organizations to adopt more high performance practices. However, this line of
work needs further research scrutiny. Scholars can collect longitudinal data to better document
the transformation of bureaucracy, the diffusion of innovative high performance practices, and
the causing factors for organizational bureaucracy and high performance adaptation. Also,
managerial orientation is crucial to undertaking innovative work practices (Appelbaum and Batt
1994: 151), future studies can focus on how managerial orientation interact with organization
size, performance pressure, and market competition to affect work organizations. For example,
risk taking managers under tremendous performance pressure and market competition may set
aside sufficient resources to support innovative work reforms in hope of obtaining long-run
benefits, whereas a risk-aversion management under the same level of pressure and competition
may cut budget on work restructuring due to its high cost and less predictable outcomes. Further,
more research are needed to shed light on how institutional environment affect work structural
innovation. For example, an international comparative study suggests that American labor law
rigidly defines management prerogatives to include various decision-making power, which
19
hinders the transition to high performance work systems, as oppose to German, Swedish, and
Japanese employers, who are legally required to share business information and involve
employees in strategic decision-making (Appelbaum and Batt 1994: 155). Empirical studies are
needed to reveal how pertinent legal regulations affect work restructuring in U.S. and other
countries. Another dimension of organization institutional environment is inter-organizational
relations (Galaskiewicz 1985; Knoke 2001: chpt. 4). A research vividly described how
organizations under many uncertainties mimic the behaviors of other organizations in their
environments (Galaskiewicz and Wasserman 1989). While American companies receive little
instructive guidance of systematic implementation of high performance programs, getting to
know other successful models that have prior experience in work reorganization becomes crucial
in waging work restructuring. The other organizations can provide much-needed information on
how to initiate changes, deal with crisis, and solve problems to facilitate transitions for
organizations undertaking high performance programs. This allegation deserves future empirical
studies.
20
REFERENCES:
Allison, Paul. 1999. Logistic Regression: Using the SAS system: Theory and Application. Cary,
NC: SAS Institute
Appelbaum, Eileen and Rosemary Batt. 1994. The New American Workplace: Transforming
Work Systems in the United States. Ithaca, N.Y.: ILR Press
Blau, Peter and Richard Schoenherr. 1971. The Structure of Organizations. New York: Basic
Books
Blau, Peter. 1970. “A Formal Theory of Differentiation in Organizations.” American
Sociological Review 35: 201-18.
Blau, Peter. 1972. “Independence and Hierarchy in Organizations.” Social Science Research 1:124.
Cappelli, Peter and David Neumark 2001. “Do High Performance Work Practices Improve
Establishment-level Outcomes?” Industrial and Labor Relations Review 54/4:737-775
Cappelli, Peter, Laurie Bassi, Harry Katz, David Knoke, Paul Osterman, and Michael Useem.
1997. Change at Work. New York: Oxford University Press
Child, John. 1973. “Predicting and Understanding Organization Structure.” Administrative
Science Quarterly 38: 168-85.
Doeringer, Peter. 1991. Turbulence in the American Workplace. New York: Oxford University
Press
Edwards, Richard. 1979. Contested Terrain: The Transformation of the Workplace in the
Twentieth Century. New York: Basic Books
Galaskiewicz, Joseph. 1985. “Interorganizational Relations.” Annual Review of Sociology
11:281-304
Galaskiewicz, Joseph and Wasserman Stanley. 1989. “Mimetic Processes within an
Interorganizational Field: An Empirical Test.” Administrative Science Quarterly 34/3:454-479
Godard, John and John Delaney. 2000. “Reflections on the ‘High Performance’ Paradigm's
Implications for Industrial Relations as a Field.” Industrial & Labor Relations Review. 53/3:
482-503
Huselid, Mark. 1995. “The impact of human resource management practices on turnover,
productivity, and corporate financial performance.” Academy of Management Journal 38/3: 635673
Kalleberg, Arne and James Moody 1996. “Human Resource Management and Organizational
Performance.” Pp 113-129 in Organizations in America: Analyzing Their Structures and Human
21
Resource Practices, edited by Kalleberg, Arne, David Knoke, Peter Marsden, and Joe Spaeth.
Thousand Oaks, Calif.: Sage Publications
Kalleberg, Arne., David Knoke, and Peter Marsden. The 1996-1997 National Organizations
Survey [machine readable data file]. University of Minnesota [producer] 2001. Inter-university
Consortium for Political and Social Research (ICPSR) [distributor] 2001.
Knoke, David. 2001. Changing Organizations: Business Networks in the New Political
Economy. Boulder, Colo.: Westview Press
Knoke, David, George Bohrnstedt, and Alisa Potter Mee. 2002. Statistics For Social Data
Analysis. Itasca, Illinois: F.E. Peacock Publishers
Kochan, Thomas, Harry Katz, and Robert McKersie. 1986. The Transformation of American
Industrial Relations. New York: Basic Books
Lawler, Edward, Susan Albers Mohrman, and Gerald Ledford. 1995. Creating High
Performance Organizations: Practices and Results of Employee Involvement and Total Quality
Management in Fortune 1000 Companies. San Francisco, Calif.: Jossey-Bass
Lawrence, Paul and Jay Lorsch. 1967. Organization and Environment: Managing Differentiation
and Integration. Homewood, IL: Irwin
Mansfield, Roger. 1973. “Bureaucracy and Centralization: An Examination of Organizational
Structure.” Administrative Science Quarterly 18: 477-88.
Marsden, Peter, Cynthia Cook, and Arne Kalleberg. 1996. “Bureaucratic Structures for
Coordination and Control.” Pp 69-86 in Organizations in America: Analyzing Their Structures
and Human Resource Practices, edited by Kalleberg, Arne, David Knoke, Peter Marsden, and
Joe Spaeth. Thousand Oaks, Calif.: Sage Publications
Meyer, Marshall W. 1972. “Size and the Structure of Organizations: A Causal Analysis.”
American Sociological Review 37: 434-41.
Osterman, Paul. 1994. “How Common is the Workplace Transformation and Who Adopts It?”
Industrial and Labor Relations Review 47/2:173-188
Osterman, Paul. 2000. “Work Reorganization in an Era of Restructuring: Trends in Diffusion and
Effects on Employee Welfare” Industrial and Labor Relations Review 53/2:179-196
Scott, W. Richard. 1998. Organizations: Rational, Natural, and Open Systems 4th edition. Upper
Saddle River, N.J.: Prentice Hall
Yang, Song. 2003. “A Contextual Analysis of Organizational Commitment.” Sociological Focus
36/1: 49-64
22
Table 1: Variable Construction
Variable
Measuring Items
Names
Bureaucratic Formalization
Organizations
Do each of the following documents exist at org?
Written job description
Written record of job performance
Employment contracts
Documents about doing personnel evaluation
Documents outlining hiring/firing procedure
Levels
How many levels are there between a first-line supervisor
and the top official in your organizations?
Departmentalization
Is there a separate department for the followings?
Personnel and labor relations
Health and safety
Equal opportunity or affirmative action
High
Teamwork
Performance
Does your organization have established committees made
Organizations
up of workers and managers who meet regularly to deal with
New technology?
Quality improvement?
Production or service delivery?
Health and safety?
Do workers meet by themselves without
management to discuss issues related to
production or service delivery?
Skill
Are core workers cross-trained, that is, trained in skills for
more than one job?
Are core workers involved in job rotation?
How often core workers transfer to another job family?
Autonomy
How much choice do core workers have concerning the
best way to accomplish their assignments (no choice =1;
small amount= 2; a moderate amount = 3; a large amount
= 4; complete choice = 5)?
Which best describes how closely core workers are
supervised as they do their work (no supervision = 1;
small supervision = 2; moderate supervision = 3; large
supervision = 4; complete supervision = 5)?
Compensation
Are core workers paid using group incentives such as gainsharing?
Do core workers participate in a profit-sharing or bonus program?
23
Coding Methods
Agreement to each question
in formalization is coded as
1, results summed up to
produce a formalization
scale from 0 to 5
Levels are actual number
reported by respondents
Agreement to each item in
Departmentalization is
coded as 1, results summed
up to produce a
departmentalization scale
from 0 to 3.
Agreement to each item in
Teamwork is coded as 1,
results summed up to
produce a Teamwork
scale.
In skill construction,
agreement to the first two
items is coded as 1, very
often response to the third
item is coded as 1. Results
summed up to produce a
Skill scale.
In job autonomy
construction, I added
supervision scale with the
choice scale, and divide the
results by 2. Thus the large
the number, the greater the
job autonomy.
Agreement to each item in
compensation is coded as 1,
results summed up to
produce a compensation
scale
Table 1 continues
Organization size
Total employment is number of full time and part
time employees in the establishment on June 1,
1996. Full time means 35 or more hours per week.
I take the natural logarithm
of the total employees to
indicate organization size.
Market competition
How much competition would you say there is in
your main market or service area . . . none, very
little, a moderate amount, or a great deal?
None = 1
Very little = 2
A moderate amount = 3
A great deal = 4
Foreign market competition
How much competition would you say there is in
your main market or service area from foreign
organizations . . . none, very little, a moderate
amount, or a great deal?
None = 1
Very little = 2
A moderate amount = 3
A great deal = 4
Performance pressure on
1. Quality
2. New product
3. Retain essential
workers
4. Customer satisfaction
How would you compare (ORG’s) performance on
the following measures over the past TWO YEARS
to that of other organizations that do the same kind
of work?
1. Quality of products, services, or programs
2. Development of new products, services, or
programs
3. Ability to attract and retain essential
employees
4. Satisfaction of customers or clients
much better = 1
somewhat better = 2
about the same = 3
somewhat worse = 4
much worse = 5
Manufacturing
What is the main product/service in this
establishment?
1980 US Census Industry
Code
Agriculture: 10-60
Manufacture: 100-392
Transportation: 400-472
Whole Sales: 500-571
Retail: 580-691
Finance/insurance: 700-760
Others: 761-932
Manufacturing = 1,
others = 0
Organization Age
Years from start of establishment operations until
1997.
I take the natural logarithm
of the years to indicate
organization age.
Independent organization
Is org in any way part of a larger organization or is
it completely independent?
1 = independent
0 = not independent
Market area
What would you consider to be org’s main
geographic market or service area?
0=Neighborhood/City/State
1=Region/U.S./International
24
Table 2: Descriptive Statistics of Variables
Variable Names
Mean
Median St.d. Data Type
Valid Cases
Formalization
3.25
4.00
1.69
continuous
913
Hierarchy
2.45
2.00
2.09
continuous
907
Departmentalization
1.83
2.00
1.12
continuous
851
Teamwork
2.94
4.00
1.75
continuous
914
Skills
1.39
1.00
0.99
continuous
923
Autonomy
3.13
3.00
0.57
continuous
946
Pay
0.53
0.00
0.68
continuous
948
Organization size
6.90
4.38
8.07
Continuous
981
Market competition
3.28
4.00
.93
Continuous
932
Foreign market competition
1.77
1.00
.99
Continuous
917
Performance on quality
1.69
2.00
.78
Continuous
912
Performance on new product
2.06
2.00
.89
Continuous
872
Retaining essential workers
2.18
2.00
.90
Continuous
899
Performance on satisfaction
1.83
2.00
.82
Continuous
910
Organization age
3.25
3.29
1.02
Continuous
1002
Market area at
region/national/international
.44
--
--
Dichotomous 940
Market area at
neighborhood/city/state
.56
--
--
Dichotomous 940
Manufacturing organizations
.18
--
--
Dichotomous 1002
Independent organization
.32
--
--
Dichotomous 1001
25
Table 3: Unstandardize Coefficients From Regressions of Organization
Adoption of Bureaucratic and High Performance Practices
Independent Variables
Bureaucracy
Skill
(Negative
Binomial)
.537**
(.190)
.032*
(.015)
-.061
(.037)
.064**
(.024)
.065
(.051)
.004
(.040)
.003
(.039)
.045
(.049)
Compensation
(Negative
Binomial)
3.532***
(.130)
-.016
(.010)
.001
(.025)
-.017
(.024)
.016
(.034)
.048*
(.023)
.054*
(.026)
.013
(.032)
Autonomy
(OLS)
-.833***
(.170)
.266***
(.013)
-.093**
(.032)
.043
(.031)
.004
(.044)
-.128***
(.035)
-.003
(.033)
-.013
(.042)
Team
(Negative
Binomial)
.890***
(.132)
.093***
(.011)
-.025
(.024)
-.002
(.024)
.027
(.034)
.090***
(.027)
.027
(.026)
-.043
(.033)
Market area at
region/national/international
.075
(.061)
-.032
(.048)
.033
(.071)
-.056
(.047)
.294**
(.110)
Age
Model χ2 (df=11)
-.030
(.029)
-.187*
(.079)
.231***
(.059)
65.32***
.036
(.023)
.051
(.059)
.061
(.045)
170.84***
-.050
(.032)
.276***
(.082)
.005
(.067)
43.34***
-.010
(.022)
-.096
(.060)
.002
(.045)
12.57***
.014
(.052)
.164
(.125)
-.036
(.106)
51.85***
Model R2
51.9%
5.5%
2.0%
3.5%
3.4%
Number of Establishments
679
763
773
785
785
(OLS)
Constant
Size
Market competition
Foreign market competition
Performance pressure-quality
Performance pressure-new product
Performance pressure- worker
Performance pressure- satisfaction
-1.351***
(.325)
.013
(.024)
.214***
(.066)
.056
(.052)
.009
(.079)
.040
(.063)
.027
(.060)
.104
(.078)
Control variables
Manufacturing
Independent
Numbers in parentheses are standard errors.
*p ≤ .05
** p ≤ .01
*** p ≤ .001
26