Content uploaded by Hui Liao
Author content
All content in this area was uploaded by Hui Liao
Content may be subject to copyright.
Do They See Eye to Eye? Management and Employee Perspectives of
High-Performance Work Systems and Influence Processes on
Service Quality
Hui Liao
University of Maryland, College Park
Keiko Toya
Doshisha University and Marketing Excellence
David P. Lepak and Ying Hong
Rutgers, The State University of New Jersey
Extant research on high-performance work systems (HPWSs) has primarily examined the effects of
HPWSs on establishment or firm-level performance from a management perspective in manufacturing
settings. The current study extends this literature by differentiating management and employee perspec-
tives of HPWSs and examining how the two perspectives relate to employee individual performance in
the service context. Data collected in three phases from multiple sources involving 292 managers, 830
employees, and 1,772 customers of 91 bank branches revealed significant differences between manage-
ment and employee perspectives of HPWSs. There were also significant differences in employee
perspectives of HPWSs among employees of different employment statuses and among employees of the
same status. Further, employee perspective of HPWSs was positively related to individual general service
performance through the mediation of employee human capital and perceived organizational support and
was positively related to individual knowledge-intensive service performance through the mediation of
employee human capital and psychological empowerment. At the same time, management perspective of
HPWSs was related to employee human capital and both types of service performance. Finally, a
branch’s overall knowledge-intensive service performance was positively associated with customer
overall satisfaction with the branch’s service.
Keywords: strategic human resource management, high-performance work systems for service quality,
human capital and motivation, employee performance, customer satisfaction
A large body of strategic human resource management (HRM)
research suggests that the use of high-performance work systems
(HPWSs), or systems of human resource (HR) practices designed to
enhance employees’ competencies, motivation, and performance, is
associated with lower employee turnover rates (e.g., Huselid, 1995),
higher labor productivity (Datta, Guthrie, & Wright, 2005), lower
injury rates and better safety performance (Zacharatos, Barling, &
Iverson, 2005), and better company performance (e.g., Huselid,
1995). At the same time, however, several authors have raised con-
cerns about some aspects of strategic HRM research. Prior research
has primarily focused on managerial reports of the use of HPWS,
ignoring the role of individual employees’ actual experiences with
these systems (Lepak, Liao, Chung, & Harden, 2006). Further, extant
strategic HRM research has predominantly taken a macro-level ap-
proach and focused on establishment or firm-level outcomes; to date,
there is a “dearth of research aimed at understanding how multiple (or
systems of) HR practices impact individuals [italics added]” (Wright
& Boswell, 2002, p. 262). Moreover, the majority of the strategic
HRM research has been conducted in manufacturing environments,
neglecting the considerable presence of service factors, which now
account for 60% of world gross domestic product (GDP) and domi-
nate economies in most nations (e.g., 71% of the GDP in Canada,
73% in the United Kingdom, 74% in Japan, and 78% in the United
States; The World Factbook, 2007).
The primary objective of this study was to address these issues.
First, given that there may be a disconnection between what
managers and companies say they do as formal practices of the
HPWS and what individual employees actually experience, in this
study we examined employee perspective with the HPWS in
addition to the management perspective. Second, we integrated
macro- and micro-level HRM research to examine the influence of
Hui Liao, Management and Organization Department, Robert H. Smith
School of Business, University of Maryland, College Park; Keiko Toya,
Graduate School of Business Sciences, Doshisha University, and Market-
ing Excellence, Tokyo, Japan; David P. Lepak and Ying Hong, Human
Resource Management Department, Rutgers, The State University of New
Jersey.
This research was supported in part by a research grant to Hui Liao from
the Rutgers University School of Management and Labor Relations. An
earlier version of this article was presented at the Academy of Management
Meeting, Philadelphia, August 2007.
We thank Fumeo Maeda and Yasuhiro Kurita for their invaluable help.
We thank seminar participants at the Hong Kong University of Science and
Technology; Ohio State University; Rutgers, The State University of New
Jersey; University of Kansas; and University of Maryland, College Park for
helpful comments and suggestions.
Correspondence concerning this article should be addressed to Hui Liao,
Robert H. Smith School of Business, University of Maryland, 4506 Van
Munching Hall, College Park, MD 20742. E-mail: hliao@rhsmith.umd.edu
Journal of Applied Psychology © 2009 American Psychological Association
2009, Vol. 94, No. 2, 371–391 0021-9010/09/$12.00 DOI: 10.1037/a0013504
371
HPWS on individual performance and to understand the psycho-
logical processes through which the influence materializes. Third,
we examined whether unit-level employee overall service perfor-
mance translates into an important performance metric for service
organizations, specifically, customer satisfaction.
In what follows, we first integrate strategic HRM and the service
management literature to discuss what an HPWS entails in the
service context and then discuss the importance of understanding
the system from the employees’ perspective as well as the psy-
chological processes through which such a system operates to
influence individual employees’ service performance and, ulti-
mately, customer satisfaction. Figure 1 outlines the theoretical
framework of the influence of a service-quality– oriented HPWS
on employee service performance.
High-Performance Work System for Service Quality
Bowen and Ostroff (2004) noted that the content of work
systems “should be largely driven by the strategic goals and values
of the organization” and that “the foci of the human resource
management practices must be designed around a particular stra-
tegic focus, such as service or innovation” (p. 206). The crux of
this statement is that to be effective, work systems must reflect
how employees add value, and this is achieved by linking the
practices within a system toward some strategic anchor. Without
an objective, work systems lack a clear direction for employees.
Therefore, all components of the HPWS should be chosen and
designed to achieve a specific organizational objective. This stra-
tegically focused approach is consistent with the argument that to
be effective, work systems should achieve horizontal fit among
various HR practices, such that these practices complement and are
aligned with each other and achieve vertical fit, such that the work
system is aligned with the organization’s strategy (Becker &
Gerhart, 1996; Huselid, 1995; Wright & McMahan, 1992).
As noted by management theorists, there are two basic strategies
for service organizations. The first is to focus on minimizing costs
and use a mass production approach in accordance with scientific
Taylorism (Porter, 1980). Although reliance on technology may
help reduce cost and improve efficiency of service (Levitt, 1972),
it may not be a sustainable advantage, as it is easily imitable and,
more importantly, the low-cost pressure may create a vicious circle
where employees and customers are increasingly dissatisfied
(Schlesinger & Heskitt, 1991). The other strategy focuses on
providing high-quality service in order to enhance customer satis-
faction and build a long-term relationship with customers (Gutek,
1995; Porter, 1980). Happy, long-term customers “buy more, take
less of a company’s time, are less sensitive to price, and bring in
new customers” (Reichheld, 1996, p. 57). Keltner (1995), for
example, found that a service-quality–focused strategy contributed
significantly to the fact that German banks performed better than
U.S. banks in the 1980s. Other research has also shown a service
differentiation strategy to be associated with higher performance
of service firms (e.g., O’Farrell, Hitchens, & Moffat, 1993). Sim-
ilarly, Rust, Moorman, and Dickson (2002) compared the financial
returns for three types of strategies, including (a) cost cutting, (b)
revenue expansion through customer-oriented quality improve-
ments, and (c) dual emphasis on cost cutting and revenue expan-
sion and found that firms adopting the revenue-expansion strategy
performed the best in terms of profitability and stock returns.
The nature of services, including simultaneity of service pro-
duction and consumption, intangibility of service processes and
outcomes, and customer involvement in service production
(Bowen & Schneider, 1988), renders it impossible to do a quality
T1M:
Management-HPWS
T2E:
Employee-HPWS
T3S: Employee
Human Capital
T3S: Employee
Individual
Service Performance
Individual Level
E
m
p
lo
y
ment Grou
p
Leve
l
T3E: Employee
Psychological
Empowerment
T3E: Employee
Perceived
Organizational
Support
Figure 1. Theoretical framework of the influence processes of the high-performance work system (HPWS) for
service quality on employee service performance and data collection schedule and sources. The horizontal
dashed line separates employment group-level and individual-level constructs; the arrow crossing a dashed line
from upper level to lower level represents cross-level, top-down effect on the lower level outcome variables.
Management-HPWS: Management’s perspective of the HPWS practices generally in use for an employee group
of certain employment status. Employee-HPWS: Employee individually experienced HPWS. T1M ⫽Time 1
management survey; T2E ⫽Time 2 employee survey; T3E ⫽Time 3 employee survey; T3S ⫽Time 3
supervisor evaluations.
372 LIAO, TOYA, LEPAK, AND HONG
control check after production to ensure quality as in the manu-
facturing setting (Schneider, White, & Paul, 1998). Therefore, the
performance of front-line employees, or their behaviors of helping
and serving customers to address customer needs (Liao & Chuang,
2004), directly influences customer satisfaction with the service
quality. In order for front-line employees to provide high-quality
service, organizations need to design a work system that ensures
that employees have the knowledge, skills, and abilities, as well as
the motivation, to meet customer needs. A few studies have
explicitly examined the linkage between HRM practices and ser-
vice quality. Schneider et al. (1998) proposed that service quality
rests on a set of organizational “foundation issues” that support
and facilitate front-line employee service delivery, which include
internal service provided by support staff, efforts to remove ob-
stacles to work, and employee participation and training. Using
data from bank branches, Schneider et al. found that the foundation
issues were positively associated with branch service climate,
which was positively associated with customer evaluations of
service quality. Batt (2002) found that high-involvement practices
characterized by high skills, discretion, and incentives were asso-
ciated with lower quit rates and subsequently higher sales growth
of call centers. Liao and Chuang (2004) examined three HRM
practices and found that employee involvement in decision making
and service training were positively related to restaurant employ-
ees’ service performance, which in turn was positively related to
customer satisfaction and loyalty.
Building on these studies as well as the frameworks of HPWSs
described by Pfeffer (1998) and Zacharatos et al. (2005), we
propose an HPWS for service quality and define it as a system of
HR practices designed to enhance employees’ competencies, mo-
tivation, and performance in providing high-quality service to
external customers. In this view, HPWS includes practices of
extensive service training, information sharing, self-management
service teams and participation, compensation contingent on ser-
vice quality, job design for quality work, service-quality– based
performance appraisal, internal service, service discretion, selec-
tive hiring, employment security, and reduced status differentia-
tion.
1
This conceptualization of the HPWS for service quality
captures the foundation HRM issues deemed important for service
delivery in Schneider et al.’s (1998) framework and includes the
HR practice dimensions examined in prior strategic HRM studies
in the service settings (Batt, 2002; Delery & Doty, 1996). In the
current study, these practices are anchored with a goal to promote
service quality. For example, extensive training emphasizes edu-
cating employees on how to provide quality service; performance
appraisal uses service criteria; and contingent compensation links
pay to service quality. These work practices, taken together, pro-
vide employees with the knowledge, skills, and abilities; re-
sources; information; and discretion they need to meet customer
demands, as well as the motivation to provide high-quality service.
Understanding the Employee Perspective
of the High-Performance Work System
Focusing on employees as providers of services also dictates that
particular attention be paid to employees’ actual experience of the
HPWS for service quality. Wright and Boswell (2002) noted that
“much of the macro HRM research empirically assumes invari-
ability in HR practices across large groups of jobs within organi-
zations” (p. 264) and has focused primarily on management per-
spective of the HR practices generally implemented for all of the
employees in an organization. Few studies have examined the
HPWS targeted to different employment groups, and even fewer
have examined the HPWS actually experienced by individual
employees. We argue that this is an oversight for several reasons.
First, different employee groups may not have identical expe-
riences of HR practices. There is a long-standing tradition of
variance in exposure to HR practices, such as compensation,
training, and promotion opportunities across administrative bound-
aries such as exempt versus nonexempt status or managerial versus
nonmanagerial status (Huselid, 1995). Companies may also use
different HR practices to match the requirements of particular
employee groups (Miles & Snow, 1984; Osterman, 1987). For
example, Lepak and Snell (2002) showed that core employees
received significantly greater exposure to a commitment-oriented
work system and that noncore employee groups tended to be
managed by work systems that convey lower levels of investment
in employees. Relatedly, Melian-Gonzalez and Verano-Tacoronte
(2006) found that the work systems used for core employees were
more sophisticated than those used for other employee groups.
Similarly, Lepak, Taylor, Tekleab, Marrone, and Cohen (2007)
demonstrated that establishments provided higher levels of expo-
sure to high-investment work systems for core employees, as
compared with support employees, and that the differences were
magnified in nonmanufacturing environments. The logic for differ-
entiating work systems across employee groups is that work systems
are used to match the level of investment needed to reflect the
relative status of different employees groups and maximize their
potential contributions to competitive success. Building on this
logic and existing research, we anticipate that employees in groups
of different employment status will experience differences in their
exposure to the HPWS practices. Such between-group difference is
expected to contribute to the variability in employees’ experiences
with the HPWS.
Second, even within the same employee group, members who
theoretically should share the same HPWS practices may be
treated differently or have different perceptions or experiences of
the practices in place. This within-group difference is another
source of the variability in employees’ experiences with the
HPWS. For instance, the organizational diversity literature has
revealed earning differences between women and men and be-
tween White persons and people of color who are similarly qual-
ified and working in the same job within the same organization,
suggesting the existence of differential treatment toward different
employees in management practices (e.g., Joshi, Liao, & Jackson,
2006; Pfeffer & Davis-Blake, 1987). The organizational justice
literature has suggested that employees often form different eval-
uations on the allocations of resources, such as pay and promotion,
the procedures used to arrive at these allocations, and the inter-
personal treatment and information given to the employees as
these procedures are carried out (e.g., see Colquitt, Conlon,
Wesson, Porter, & Yee, 2001, for a review). Further, the leader–
1
We excluded the practices of selective hiring, employment security,
and reduced status differentiation in our empirical testing of this study due
to the nature of the sample. See the Measures section for a detailed
explanation.
373
HIGH-PERFORMANCE WORK SYSTEMS
member exchange (LMX; Graen & Uhl-Bien, 1995) literature has
proposed that leaders establish different social exchange relation-
ships with different subordinates; employees who have a high-
quality LMX with their supervisor have the advantages of ample
resources, more training opportunities, premier assignments, emo-
tional support, decision-making responsibilities, and cooperative
interactions with the supervisor (Liden & Graen, 1980). Consistent
with this view, Rousseau (2005) noted that individual employees
negotiate their idiosyncratic employment arrangement, or
“I-deals,” with employers. In addition, Bowen and Ostroff (2004)
noted that
HRM practices can be viewed as a symbolic or signaling function by
sending messages that employees use to make sense of and to define
the psychological meaning of their work situation (e.g., Rousseau,
1995). All HRM practices communicate messages constantly and in
unintended ways, and messages can be understood idiosyncratically,
whereby two employees interpret the same practices differently
(Guzzo & Noonan, 1994). (p. 206)
Taken together, these theoretical perspectives and empirical
findings imply that HR practices may be implemented differently
for employees, or at least that employees may perceive or experi-
ence differences in exposure to work practices. The lack of uni-
formity across employees in their experiential-based perceptions
about the HPWS practices suggests that there will exist a discon-
nection between what management says about the HPWS practices
generally implemented for a particular employee group (hereinaf-
ter referred to as management-HPWS) and the HPWS practices
actually experienced by the individual employees in that group
(hereinafter referred to as employee-HPWS). Therefore, it is nec-
essary to examine employee-HPWS in addition to management-
HPWS to understand the psychological process through which formal
practices of HPWS influence individual employees’ human capital,
motivation, and behaviors in service delivery.
Influence of Employee Perspective of the
High-Performance Work System on Individual
Service Performance
Next, we examined the influence process of employee-HPWS
on employee performance. We argue that the employee-HPWS is
positively related to employee service performance by enhancing
employee human capital and motivation required for service de-
livery. Figure 1 depicts the proposed mediation relationships. Our
focus on these mediating factors is in line with the prominent
theory of job performance (Campbell, McCloy, Oppler, & Sager,
1993), which suggests that the direct determinants of individual
performance are individuals’ human capital and motivation and
that other individual differences, such as experiences and person-
ality, and contextual factors such as HR interventions, indirectly
influence job performance through their impact on the individuals’
human capital and motivation.
Human Capital
Human capital refers to employee knowledge, skills, and abili-
ties that are valuable for the firm (e.g., Subramaniam & Youndt,
2005). Strategic HRM research has used human capital as a the-
oretical underpinning (Jackson & Schuler, 1995), with the argu-
ment that one important function of HRM is its “buying” and
“making” of desirable employee knowledge, skills, and abilities,
which can in turn be used to create value for the firm (e.g., Becker
& Huselid, 1998; Delery & Shaw, 2001; Lado & Wilson, 1994;
Snell, Youndt, & Wright, 1996). Despite its wide mention in
empirical strategic HRM studies (e.g., Batt, 2002; Huselid, 1995),
the role of human capital has rarely been examined explicitly as a
mediator between work systems and performance outcomes.
In a service context, employees need to have a good knowledge
about the services, products, and customer needs and to have the
abilities and skills to meet customer needs. We argue that
employee-HPWS, a key mission of which is to select and develop
service talents through practices such as service-quality–focused
hiring, training, information sharing, performance feedback, and
so forth may enhance employee human capital for service delivery
and subsequently service performance. Therefore, we propose the
following:
Hypothesis 1: The positive relationship between employee-
HPWS and employee service performance is mediated by em-
ployee human capital.
Motivation
Motivation refers to an individual’s direction, intensity, and
duration of effort (Campbell et al., 1993). Whereas human capital
provides the capabilities for employees to contribute, motivation
deals with the extent to which employees are willing to utilize
these capabilities. Jackson and Schuler (1995) pointed out that “the
potential value of human capital can be fully realized only with the
cooperation of the person” (p. 241). HRM practices need to effec-
tively align the interests of employees and employers so that
employees are willing to exert their effort (Delery & Doty, 1996).
Yet, motivation has seldom been measured explicitly or tested in
strategic HRM studies. In contrast, the importance of motivation
has long captured psychology researchers’ attention. Two motiva-
tion constructs that are particularly related to individual employee
job performance are psychological empowerment and perceived
organizational support.
Psychological empowerment. Psychological empowerment re-
fers to individuals’ self-motivating mechanisms and consists of
meaning, competency, self-determination, and impact (Spreitzer,
1995; Thomas & Velthouse, 1990). Meaning is an individual’s
perceived value of work compared with the individual’s personal
goals and standards; competence is an individual’s confidence in
his or her ability to perform the work satisfactorily; self-
determination is an individual’s sense of control in initiating
and changing actions; and impact is an individual’s perceived
influence over important strategic, administrative, or operating
outcomes.
Although psychological empowerment reflects individuals’ in-
nate intrinsic task motivation, it can be influenced by external
practices (Seibert, Silver, & Randolph, 2004; Spreitzer, 1995). We
argue that HPWS may represent such empowering work practices.
For examples, service-quality–focused performance feedback and
information sharing may help employees perceive the service tasks
to be meaningful and important; extensive service training and
delegation of decision-making power may enhance employees’
confidence in their competence in service delivery; and increased
374 LIAO, TOYA, LEPAK, AND HONG
discretion in customizing service delivery and handling customer
complaints may increase employees’ perceived self-determination
and impact.
We further argue that psychological empowerment will lead to
better service performance. Research shows that psychological
empowerment leads to employee commitment to and internaliza-
tion of task goals (Kanter, 1983), persistence of effort in nonrou-
tine situations and resilience in adversary situations (Bandura,
1977), initiative (Thomas & Velthouse, 1990), and innovative
behaviors (Spreitzer, 1995). All of these characteristics may facili-
tate service performance, which requires employees to learn about the
diverse needs of customers, handle the uncertainty customers intro-
duce when participating in service, and adapt their interpersonal
style and service offering to customer needs. Indeed, Gwinner,
Bitner, Brown, and Kumar (2005) found intrinsic motivation to be
positively related to employee adaptive service behaviors.
Whereas prior studies have demonstrated the positive impact of
intrinsic motivation on performance criteria, the evidence concern-
ing the role of extrinsic motivation is less convincing. Cognitive
evaluation theory (Deci, 1975) suggests that extrinsic motivation
regulates a behavior by way of external contingencies, diminishes
feelings of autonomy, and activates a change in perceived locus of
causality from internal to external, hence undermining intrinsic
motivation. Ryan, Mims, and Koestner (1983), on the contrary,
found that extrinsic rewards contingent on high-quality perfor-
mance given in a supportive interpersonal context actually en-
hanced intrinsic motivation. However, Gwinner et al. (2005) found
extrinsic motivation to be unrelated to employee customized ser-
vice delivery. Given these inconclusive arguments and findings,
we focused on intrinsic motivation in our hypothesis development
and controlled for extrinsic motivation in our analyses.
Hypothesis 2: The positive relationship between employee-
HPWS and employee service performance is mediated by
employee psychological empowerment.
Perceived organizational support. Beyond psychological em-
powerment, we propose that another reason for employees to be
motivated by the HPWS is a favorable social exchange with the
organization. On the basis of the social exchange theory (Blau,
1964), Settoon, Bennett, and Liden (1996) argued that “positive,
beneficial actions directed at employees by the organization and/or
its representatives contribute to the establishment of high-quality
exchange relationships that create obligations for employees to
reciprocate in positive, beneficial ways” (p. 219). Perceived orga-
nizational support, or employees’ perceptions of the extent to
which organizations value employees and care about their well-
being (Eisenberger, Fasolo, & Davis-Mastro, 1990), is essential in
forming such obligations (Shore & Wayne, 1993). Researchers
have argued that when employees perceive a high level of orga-
nizational support, they may use behaviors valued by the organi-
zation as currency to reciprocate the benevolent treatment from the
organization (Lambert, 2000; Shore et al., 2004). Similarly,
Schneider and Bowen (1985) argued that service employees are
likely to treat customers well if the organization treats the employ-
ees well.
Numerous efforts have examined the antecedents and outcomes
of perceived organizational support. Such support has been shown
to be influenced by organizations’ investment in employees
through HR practices, such as training and development, and
organizations’ recognition of individual achievement through
practices such as promotions and salary increases (e.g., Wayne,
Shore, Bommer, & Tetrick, 2002; Wayne, Shore, & Liden, 1997).
In addition, perceived organizational support is found to play a key
role in relating to employees’ commitment to the organization
(Eisenberger et al., 1990; Settoon et al., 1996; Wayne et al., 1997),
satisfaction (Shore & Tetrick, 1991), conscientiousness in per-
forming tasks and innovation (Eisenberger et al., 1990), loyalty
(Tetrick, Shore, Newton, & Vandenberg, 2007), and organizational
citizenship behavior and lower intention to quit (Wayne et al.,
1997).
Despite these efforts, to our knowledge, few strategic HRM
studies have integrated perceived organizational support as a link-
age between HPWS and employee performance. One exception
was the study by Allen, Shore, and Griffeth (2003), which showed
that service employees’ perceptions of participation in decision
making, fairness of rewards, and growth opportunities were posi-
tively associated with their development of perceived organiza-
tional support, which, in turn, was positively associated with their
job satisfaction and organizational commitment, as well as lower
turnover.
Building on these rationales and findings, we argue that per-
ceived organizational support may be a potential path by which
HPWS influences employee performance in a service context.
Practices such as extensive training reflect companies’ investment
in employees; practices such as internal career opportunities, ser-
vice discretion, and self-management service teams demonstrate
management’s respect of employees’ opinions and initiatives; and
practices such as service-quality– based performance appraisal and
compensation show the recognition of employee service excel-
lence. Therefore, we propose the following:
Hypothesis 3: The positive relationship between employee-
HPWS and employee service performance is mediated by
employee perceived organizational support.
Influence of Management Perspective
of High-Performance Work Systems on
Individual Outcomes
Thus far we have proposed hypotheses regarding the influences
of employee-HPWS on individual outcomes. As noted above,
employee-HPWS may differ from management-HPWS. However,
this is not to suggest that there is no relationship between these two
perspectives. We argue that as management-HPWS represents the
HPWS practices generally implemented for a particular group of
employees, to a certain extent it reflects the objective environment
shaped by formal management practices. As a result, management-
HPWS provides a contextual cue for employees to form their
perceptions and experience of the work system. Therefore, we
expect management and employee perspectives of the HPWS to be
positively related.
At the same time, however, employee-HPWS may have a more
proximal relationship with employee individual outcomes, because it
is the employees’ actual experiences and perception of the context,
not the context itself nor the cues obtained from the context, that
directly determine their reactions (James, James, & Ashe, 1990;
Schneider, 1990). Therefore, management-HPWS affects an em-
375
HIGH-PERFORMANCE WORK SYSTEMS
ployee to the extent that it affects the employee’s experience and
perception of the HPWS. As a result, we propose that employee-
HPWS mediates the effects of management-HPWS on employee
human capital and motivation.
Hypothesis 4: The positive relationship between management-
HPWS and individual employee human capital, psychological
empowerment, and perceived organizational support is medi-
ated by employee-HPWS.
Taken together, Hypotheses 1 through 4 suggest that
management-HPWS influences employee-HPWS, which, in turn,
impacts individual employees’ human capital, psychological em-
powerment, and perceived organizational support, and these fac-
tors further influence individual employee service performance.
Linking Employee Service Performance
to Customer Satisfaction
Next, we propose that employee service performance may fur-
ther influence customer satisfaction on two levels. First, at the
individual level, employee service performance to a particular
customer, or the one-on-one interaction between an employee and
a customer, may directly affect the customer’s satisfaction with
this particular service encounter and the customer’s decision re-
garding whether to continue to use the service of this particular
employee (Gutek, 1995; Gutek, Bhappu, Liao-Troth, & Cherry,
1999). For example, by observing 191 bank tellers, Pugh (2001)
found that individual tellers’ display of positive emotions was
directly associated with their customers’ positive affect and sub-
sequently with positive evaluations of service quality. Liao and
Chuang (2007) found that a hairdresser’s individual-level service
performance directly determines customers’ willingness to have a
long-term service relationship with the hairdresser and the number
of long-term customers the hairdresser maintains. Therefore, an
individual employee’s service performance may directly affect
customer satisfaction.
Second, a customer’s overall experience with a service unit may
be shaped by the customer’s multiple encounters with multiple
service employees in the unit (Liao & Chuang, 2004). For exam-
ple, a banking customer may likely interact with different front-
line employees on different visits and for different banking or
investment needs. Thus, at the business-unit level of analysis, the
overall service performance across the front-line employees may
affect customer satisfaction with the unit’s overall service delivery.
Whereas theoretically both individual-level and unit-level em-
ployee service performance may affect customer satisfaction, test-
ing the individual-level effect requires an exact match between the
employee and the customers he or she serves (see Liao & Chuang,
2007, for such a design and test). The current study did not include
such matched data but instead assessed the individual customers’
satisfaction with the unit’s overall service. Therefore, with full
acknowledgment of the theoretical and business relevance for the
effect of individual-level employee service performance, in the
current study we focused on the effect of unit-level employee
service performance and proposed the following:
Hypothesis 5: A unit’s overall employee service performance
is positively related to customer satisfaction with the unit’s
overall service quality.
Method
Study Design and Participants
We tested the proposed theoretical framework using data from a
national bank in Japan. All 92 branches of the bank participated in
the study. In order to link information from multiple stakeholders
involved in the service profit chain, test the temporal linkages
among the study variables, and reduce common method bias, we
collected information from five sources (headquarters, branch se-
nior managers, employee supervisors, customer-contact employ-
ees, and branch customers), in two formats (surveys and archival
data), and in three phases.
Specifically, at Time 1, branch management (including branch
senior managers and employee supervisors) filled out a survey
about the HPWS practices implemented in their branch for three
employee groups of different employment status (group differ-
ences are explained in the Measures section). At Time 2, individ-
ual employees filled out a survey about their personal experiences
of the work system. At Time 3, employees filled out another
survey on variables including their psychological empowerment,
extrinsic motivation, and perceived organizational support. We
separated employee measures into these two surveys at different
times to reduce common-method bias. In addition, at Time 3,
employee supervisors provided evaluations of each individual em-
ployee’s human capital and service performance. The time interval
between two adjacent phases ranged from 2 to 4 weeks. Concur-
rent with the data collection within the branches, customer satis-
faction measures were collected from the branch’s external cus-
tomers, and employee demographic information and branch
characteristics, such as size, age, and local competition, were
obtained from the bank headquarters’ archival data.
To solicit honest responses, we arranged for the marketing
consulting company run by Keiko Toya to administer the entire
data collection. All branch senior managers, supervisors, and em-
ployees of the bank were invited to participate in the study, and a
stratified random sample of each branch’s external customers was
selected to represent customers of different asset levels. The mar-
keting consulting company precoded the surveys with branch and
individual identification numbers and sent all of the surveys di-
rectly to the participants’ home together with a postage-paid return
envelope addressed directly to Keiko Toya. The respondents were
assured of confidentiality and that nobody from the bank would
have access to their individual responses. To further reduce po-
tential psychological stress, we did not include any question in the
employee surveys pertaining to individuals’ names, locations, or
personal demographic information and obtained these data directly
from the bank’s headquarters. Upon returning their completed
survey, the customers each received a payment in Japanese yen
equivalent to about 10 U.S. dollars, while the bank’s managers,
supervisors, and employees did not receive any payment.
With endorsement from the bank’s top management, we
achieved high response rates for the branch manager and employee
samples across the branches. A total of 292 (91%) branch manag-
ers filled out the Time 1 management survey, 1,149 (92%) em-
ployees filled out both of the two employee surveys, and 223
supervisors provided performance evaluations for 959 (77%) of
the employees. We had a final usable sample with complete
matched management–supervisor– employee information for 830
employees from 91 branches. We compared the 830 employees
376 LIAO, TOYA, LEPAK, AND HONG
included in the final sample with the excluded employees having
incomplete information and found that the final sample had higher
female representation and were younger than the unused sample.
To account for the potential effects of gender and age, we con-
trolled for these variables in all analyses. In addition, we obtained
1,772 usable customer responses from 75 branches, representing
an overall response rate of 27%, which is comparable to the survey
response rate from random survey sampling reported in the mar-
keting literature (e.g., Seiders, Voss, Grewal, & Godfrey, 2005).
We conducted ttests of two-sample means and found that, in
comparison with these 75 branches, the 16 branches with no
customer data had fewer female employees and more competing
banks in the neighborhood but no statistically significant differ-
ence for any of the other study variables. We controlled for these
variables in all analyses.
Survey Translation Procedures
For measures that were originally in English, we followed an
iterative translation procedure. First, Hui Liao worked closely with
a Japanese linguist who teaches Japanese at a U.S. university to
translate the surveys from English to Japanese. Second, Keiko
Toya, who is a Japanese native and proficient in English, teaches
marketing at a university in Japan, and has extensive consulting
experience with the Japanese banking industry, checked the trans-
lation for accuracy, discussed the relevance and appropriateness of
the questions in depth with focus groups of managers and employ-
ees from the bank, identified areas of concerns, and suggested
questions to be deleted, added, and modified. Third, the Japanese
linguist, in consultation with Hui Liao and Keiko Toya, worked to
resolve these issues. Fourth, Keiko Toya’s business partner, who is
a Japanese native with a U.S. master’s in business administration
degree and consults with the Japanese banking industry, improved
the readability of the questions through discussions with Keiko
Toya.
Next, we describe the measures for the main study variables
involved in the analyses of employee service performance and
customer satisfaction, respectively, by each level of analysis. We
then describe the measures for the control variables by each level
of analysis.
Measures of Individual-Level Variables Relevant for
Predicting Employee Service Performance
Employee-experienced high-performance work system. The
employee measure assessed employees’ individual experiences of
the work system. Employees answered questions on the basis of
their personal experience and understanding of the HR practices on
a 5-point Likert scale ranging from 1 (strongly disagree)to5
(strongly agree). Our measure of HPWS included eight practice
dimensions that aim to enhance employee human capital, psycho-
logical empowerment, and obligation in delivering high-quality
service. Table 1 reports the practice dimensions, scale sources,
means, standard deviations, and coefficient alphas.
The first six practice dimensions were based on Zacharatos et
al.’s (2005) framework, and the measures were primarily derived
from the employee version of the HPWS used in Study 2 of
Zacharatos et al. and adapted in the current study to have a
service-quality focus. The dimensions include extensive service
Table 1
High-Performance Work Systems (HPWSs): Subscales, Employee Perspective (Employee-HPWS), and Management Perspective (Management-HPWS) for Different
Employee Groups
Practice dimension
Employee-HPWS
(␣⫽.82)
Management-HPWS for
employee group 1
(␣⫽.86)
Management-HPWS for
employee group 2
(␣⫽.86)
Management-HPWS for
employee group 3
(␣⫽.84)
Scale sourceMSD␣Loading MSD␣Loading MSD␣Loading MSD␣Loading
Extensive service training 3.13 0.61 .76 .51 3.52 0.58 .80 .54 3.51 0.55 .81 .52 2.97 0.71 .84 .46 Delery & Doty (1996);
Zacharotos et al. (2005)
Information sharing 3.53 0.63 .84 .79 3.96 0.50 .83 .68 3.85 0.51 .82 .63 3.12 0.68 .85 .65 Zacharotos et al. (2005)
Self-managed teams and
participation 3.37 0.61 .80 .55 3.70 0.55 .74 .70 3.68 0.50 .72 .73 3.31 0.61 .77 .80 Zacharotos et al. (2005)
Compensation contingent
on service performance 2.96 0.61 .78 .81 3.19 0.64 .64 .51 3.13 0.61 .66 .41 2.73 0.61 .78 .47 Zacharotos et al. (2005)
Job design 3.63 0.58 .72 .51 3.77 0.52 .51 .58 3.79 0.51 .58 .65 2.95 0.69 .67 .69 Zacharotos et al. (2005)
Service-quality-based
performance appraisal 3.09 0.68 .87 .52 3.82 0.70 .87 .52 3.85 0.58 .81 .51 3.61 0.70 .84 .41 Delery & Doty (1996);
Zacharotos et al. (2005)
Internal service 3.46 0.77 .73 .56 3.82 0.59 .81 .52 3.80 0.57 .82 .52 3.66 0.61 .83 .58 Schneider et al. (1998)
Service discretion 2.83 0.67 .86 .59 3.15 0.63 .76 .59 2.95 0.66 .81 .66 2.72 0.72 .85 .73 Developed for present study
377
HIGH-PERFORMANCE WORK SYSTEMS
training (6 items, e.g., “The branch supports me to join the cus-
tomer service training program provided by the headquarters”),
information sharing (8 items, e.g., “Customers’ suggestions on
how to improve service quality is shared with me”),self-
management service teams and participation (6 items, e.g., “Sug-
gestions for improving customer service from employees like me
are usually implemented in full or in part within this branch”),
compensation contingent on service quality (8 items, e.g., “My pay
is tied to the quality of service I deliver to the customers”), job
design for quality work (5 items, e.g., “My job is designed to be
simple and repetitive”; reverse-coded),and service-quality– based
performance appraisal (4 items, e.g., “To what extent does your
branch evaluate your performance based on a track record of your
courteous service to customers”). Several items deemed irrelevant
for the Japanese banking industry were dropped from Zacharatos
et al.’s (2005) original scale (e.g., “Such things as my previous
injuries and my alcohol and substance use were assessed before I
was hired to work here”). We also supplemented Zacharatos et
al.’s measures of the training and performance appraisal practices
with items adapted from Delery and Doty’s (1996) measure of
work systems for banking industries (e.g., “The training programs
I went through in this branch prepared me to provide high-quality
service”).
In addition to these six dimensions, we included two practices
deemed important for the service context. Internal service refers to
the interdepartmental support provided to service employees and
has been argued to influence front-line employees’ service deliv-
ery to external customers (Schneider et al., 1998). We used the two
items reported in Schneider et al. (1998) to assess internal service
(e.g., “Employees in the other departments of this branch cooper-
ate well with me to get my job done”). Service discretion refers to
the level of authority employees have in resolving customer com-
plaints and customizing service offering and has been argued to
influence employee prompt and responsive handling of customer
complaints and employee adaptive service behaviors (Gwinner et
al., 2005). We developed five items to assess service discretion
(e.g., “I have the discretion to customize the service offering to
meet customer needs”).
We dropped selective hiring, which was included in Zacharatos
et al. (2005), because the bank has centralized staffing at the
headquarters and thus the branches do not hire their own employ-
ees. We also dropped employment security and reduced status
differentiation dimensions, included in Zacharatos et al., because
as we explain later, as a standard practice across branches, the
bank has clear hierarchical status differentiation for different
groups of employees, and whether an employee has lifetime em-
ployment security depends on the group to which the employee
belongs. Therefore, there would be no between-branches variance
in formal management practices in terms of these three dimen-
sions. In addition, because a goal of the current study was to
examine the same set of HR practices from both the employee and
management perspectives, we excluded these three dimensions
from the measures of the employee perspective.
We then calculated a unitary index of employee-HPWS. This
index approach has been recommended and widely used in stra-
tegic HRM research, as it is consistent with one of the fundamental
principles of strategic HRM research, which argues that the impact
of HR practices is best understood by examining the system of HR
practices in place instead of examining HR practices individually
(Becker & Huselid, 1998; Delery, 1998; Guthrie, 2001; Huselid,
1995; Lepak et al., 2006; Ostroff & Bowen, 2000; Wright &
Boswell, 2002). We followed the subscale aggregation approach,
as supported by Drasgow and Kanfer (1985) and used in prior
strategic HRM studies such as Zacharatos et al. (2005). We first
calculated the subscale scores, averaging across items of the same
practice dimension (e.g., selective hiring), which was justified by
the high internal consistency of the subscales; we then created the
index of employee-HPWS, averaging across the eight practice
dimensions, which again was justified by an alpha of .89 across the
HR practices. Further, we conducted a principal factor analysis for
the practices included in employee-HPWS. Only one factor had an
eigenvalue of greater than 1, providing additional support for the
unitary-index approach. Table 1 reports the factor loadings.
Employee psychological empowerment. In the second survey
sent to employees, employees reported their individual psycholog-
ical empowerment using Spreitzer’s (1995) 12-item scale adapted
to the service context. The 5-point Likert scale (1 ⫽strongly
disagree;5⫽strongly agree) assessed employee perceived mean-
ing (e.g., “The service work I do is meaningful to me”), compe-
tency (e.g., “I am self-assured about my capabilities to perform my
activities in customer service”), self-determination (e.g., “I have
considerable opportunity for independence and am free in how I do
my job in serving customers”), and impact (e.g., “I have significant
influence over what happens in my branch”) in service delivery.
Alpha was .89.
Employee perceived organizational support. In the second
employee survey, employees evaluated their perceived organiza-
tional support from the branch, using the eight-item perceived
organizational support scale of Eisenberger, Cummings, Armeli,
and Lynch (1997). Respondents indicated their agreement with
each of the statements using a 7-point scale ranging from 1
(strongly disagree)to7(strongly agree) (e.g., “My branch cares
about my well-being”). Alpha was .89.
Employee human capital. An employee’s human capital was
evaluated by his or her direct supervisor using a five-item human
capital scale by Subramaniam and Youndt (2005) and Youndt,
Subramaniam, and Snell (2004). Participants responded on a
7-point scale ranging from 1 (strongly disagree)to7(strongly
agree). The items were adapted to describe service related knowl-
edge, skills, and abilities (e.g., “This employee is highly skilled in
serving customers”). Alpha was .94.
Employee service performance. Employees’ direct supervisors
evaluated two types of service performance for each employee on
a 7-point scale (1 ⫽highly unsatisfactory;7⫽highly satisfac-
tory): general service performance and knowledge-intensive finan-
cial service performance. General service performance refers to
the overall professional appearance and the reliability, responsive-
ness, assurance, and empathy displayed by employees in serving
customers. It was assessed by 23 items adapted from the widely
used service quality scale of Parasuraman, Zeithaml, and Berry
(1994; e.g., “how satisfactory is this employee’s performance
regarding providing services as promised?”). General service per-
formance is applicable for all service sectors, and these service
quality criteria have been used in numerous marketing studies (see
Asubonteng & McCleary, 1996, for a review). Knowledge-
intensive service performance refers to the financial service that
requires specialized professional knowledge and skills and is only
relevant for the financial industry. We developed an 18-item scale
378 LIAO, TOYA, LEPAK, AND HONG
to assess employee performance in recommending, designing, and
selling appropriate investment products, flexible/fixed-return
products, pension plan products, loan products and mortgage
products, managing customer tangible and intangible assets,
handling irregular cases, and so forth. Alpha was .97 for general
service performance and .96 for knowledge-intensive service
performance.
Because employee supervisors provided the ratings for individ-
ual employees’ human capital, general service performance, and
knowledge-intensive service performance in one survey and be-
cause knowledge-intensive service performance is a new scale, we
conducted a series of confirmatory factor analyses to examine the
discriminant validity of these three constructs. We first created
item parcels for each scale in order to yield more stable parameter
estimates (Kishton & Widaman, 1994; Yuan, Bentler, & Kano,
1997). For example, the 23 items of the general service perfor-
mance scale were randomly assigned to five parcels, and the
average item scores of the five parcels were used as indicators for
general service performance. We found that the one-factor model
fit the data poorly,
2
(54) ⫽4,218.44, non-normed fit index
(NNFI)⫽.77, comparative fit index ( CFI)⫽.82, incremental fit
index ( IFI)⫽.82, and root mean square residual ( RMSR)⫽.19.
In contrast, the hypothesized three-factor model fit the data well,
2
(51) ⫽757.28, NNFI ⫽.96, CFI ⫽.97, IFI ⫽.97, RMSR ⫽
.045, and significantly better than the one-factor model, ⌬
2
(3) ⫽
3,461.15, p⬍.001; better than an alternative two-factor model
combining human capital and general service performance into
one factor, ⌬
2
(2) ⫽315.39, p⬍.001; better than an alternative
two-factor model combining human capital and knowledge-
intensive service performance into one factor, ⌬
2
(2) ⫽3,284.2,
p⬍.001; and better than an alternative two-factor model combin-
ing general service performance and knowledge-intensive service
performance into one factor, ⌬
2
(2) ⫽3,130.62, p⬍.001. Fur-
ther, we constrained the estimated correlation parameter (
ij
)
between two of the three variables to 1.0 at a time and then
performed chi-square difference tests on the values obtained for
the constrained and unconstrained models (Anderson & Gerbing,
1988). All of the three chi-square difference tests revealed that the
unconstrained models fit the data significantly better than did the
constrained model, suggesting that the imposed constraint was
unrealistic. In addition, we used the test recommended by Ander-
son and Gerbing (1988) as well as Bagozzi, Yi, and Phillips (1991)
to examine whether the 95% confidence interval around the cor-
relation of each pair of the factors contained the value of one (⫹1
or ⫺1). We found that none of the 95% confidence intervals
contained the value of one. Taken together, these results provided
discriminant validity evidence that the three factors are distinct
from one another.
Measures of Group-Level Variables Relevant for
Predicting Employee Service Performance
To measure management’s assessment of the HPWS practices
targeted to different employee groups at each branch, we asked all
managers to fill out a survey consisting of three sections; the HR
practice questions repeated in each section for three different
groups of employees. In the bank evaluated in this study, employ-
ees are categorized into three groups with different official status,
employment security, and advancement opportunities. Group 1, or
“Sougoshoku” employees, is the group with the highest status;
employees in this group enjoy a lifetime contract, have unlimited
transferability to branches of their choice and promotion opportu-
nities to management, and are mostly men. Members of Group 2,
or “Ippanshoku” employees, have lower status, maintain a lifetime
contract but limited transferability and fewer promotion opportu-
nities to management, and are mostly women. Those in Group 3,
or part-time employees, have the lowest status, no lifetime con-
tract, and no transfer or promotion opportunities and are predom-
inantly women. All three groups of employees have direct contact
with customers. Because management may have different practices
targeted to different groups, we measured management perspec-
tives of HPWS for each group separately.
We conceptualize management-HPWS as an eight-dimension
system corresponding to the employee-HPWS measures. The
items were primarily adapted from the management version used
in Study 1 of Zacharatos et al. (2005) and supplemented with items
from Delery and Doty (1996) and Schneider et al. (1998), except
for the service discretion dimension, which, as noted above, was
mostly created for this study. On a 5-point scale ranging from 1
(strongly disagree)to5(strongly agree), branch managers an-
swered questions about the formal HR practices for each of the
three employee groups.
We then calculated the index scores of management-HPWS
for the three groups of employees following a similar procedure
used in calculating the employee-HPWS index score. The high
internal consistency both within each practice dimension and
across the eight dimensions, as well as the overall one-factor
structure of the practices, empirically justified the index ap-
proach (see Table 1). In each branch, an average of three
managers reported management-HPWS targeted toward the em-
ployee groups.
2
We aggregated the index scores across the
managers of the same branch to form the measure of manager-
HPWS for each employee group. Aggregation is justified by a
high interrater agreement (r
wg(j)
; James, Demaree, and Wolf,
1984) that adjusted for a slight negative skew in the expected
variance; the median r
wg(j)
value was .97 for management-
HPWS targeted toward Group 1 employees, .98 targeted toward
Group 2 employees, and .97 targeted toward Group 3 employ-
ees. In addition, ICC
1
(intraclass correlation), or the proportion
of between-branch variance in the total variance, was signifi-
cant for all three types of measures, with the values of .16, .11,
and .17, respectively, which are comparable to the inflated
median ICC
1
value of .12 reported in the organizational liter-
ature (see Bliese, 2000). Further, ICC
2
, or the reliability of
branch mean, was .38, .28, and .34, respectively. The relatively
low ICC
2
values are a direct result of the low ICC
1
values and
the very small number of managers from each branch. Low
ICC
2
values suggest that it may be difficult to detect emergent
relationships using branch means (Bliese, 2000); however, they
should not prevent aggregation if aggregation is justified by
theory and supported by high r
wg(j)
(Chen & Bliese, 2002;
2
In our sample, 9 out of the 91 branches had only one manager and
represented branches of smaller sizes. We tested the hypotheses with and
without the data from these 9 branches and found the pattern of the results
to be highly consistent. We retained these branches in our formal analyses
in order to have a more representative sample of the population.
379
HIGH-PERFORMANCE WORK SYSTEMS
Kozlowski & Hattrup, 1992). Therefore, we proceeded with
aggregation, acknowledging that the relationships between the
management-HPWS and the other study variables may be un-
derestimated.
Measures of Variables Relevant for Predicting
Customer Satisfaction
Individual-level customer overall satisfaction with branch ser-
vice. Branch customers provided evaluations of their satisfaction
with the branch’s overall service using eight items to assess the key
indicators of service quality for the Japanese banking industry (e.g.,
teller performance, visiting service provided to the customer’s
home or workplace). Alpha for this scale was .88.
Branch-level average employee service performance. In order
to examine how customer satisfaction is influenced by the overall
service performance of the branch employees, we aggregated
employee ratings of the branch’s overall general service perfor-
mance and knowledge-intensive financial service performance to
the branch level to represent the branch’s employee overall general
service performance and knowledge-intensive service perfor-
mance. Alpha was .96 for the former and .96 for the latter. The
aggregation was justified by a high median r
wg(j)
of .89 and .92,
respectively, for the two types of service performance.
Control Variables
Individual-level control variables. We controlled for employ-
ees’ age and gender at the individual employee level of analysis;
information on these variables was obtained from the headquar-
ters’ archival data. In addition, in assessing the effect of employee
psychological empowerment or intrinsic motivation on employee
service performance, we controlled for the influence of extrinsic
motivation. We measured employee-perceived extrinsic motiva-
tion following the procedure described in Tyagi (1985): First,
performance expectancy (E; 6 items; alpha ⫽.92), as well as the
instrumentality (I; 5 items; alpha ⫽.84) and valence (V; 5 items;
alpha ⫽.83) of extrinsic rewards for performance, were measured
with items answered by employees adapted from Gwinner et al.
(2005); then, the extrinsic motivation score (EM) was calculated
by standardizing the score of
EM ⫽
冋
E⫻
冉
冘
k⫽1
n
Ik⫻Vk
冊册
.
When predicting a customer’s satisfaction with the branch’s
overall service, we controlled for the customer’s gender, age, and
total household asset as reported by the responding customer at the
individual customer level of analysis.
Group-level control variables. We controlled for employees’
affiliation with a specific employment group at the group level,
using the archival data from the bank’s headquarters (Group 1 ⫽
1 if the employee belonged to employee Group 1; Group 2 ⫽1if
the employee belonged to employee Group 2; and Group 3 was
used as the comparison group).
In addition, we controlled for the average ratings of employee-
HPWS at the group level, which represents the employees’ shared
experience of the HPWS within the same group in a branch.
Through social interactions, employees of the same group may
come to form certain common perceptions or psychological states,
which may further influence employee attitudes and behavior
above and beyond employee individual experience with the work
system (e.g., Liao & Chuang, 2007; Ostroff & Bowen, 2000).
Therefore, we aggregated employee-HPWS to the group level as a
control variable in order to show that, after accounting for
employee-shared experience of HPWS, an individual employee’s
idiosyncratic experience with the HPWS (measured as a Level 1
variable, employee-HPWS) was significantly related to employee
human capital, psychological empowerment, perceived organiza-
tional support, and performance. Aggregation is justified by a high
r
wg(j)
adjusted for a small negative skew in the expected variance
(James et al., 1984); the median r
wg(j)
value was .96.
Branch-level control variables. At the branch level, we con-
trolled for the branch’s age, size, and number of competing banks
in the neighborhood in all of the analyses. Information for these
variables was provided by the bank’s headquarters. In addition,
when testing Hypothesis 5 on the relationship between a branch’s
overall employee service performance and customer satisfaction,
we controlled for branch average management-HPWS (calculated
by first assigning the management-rated HPWS score for a partic-
ular employee group to all of the employees in that group, then
averaging the scores across all the employees in the branch),
branch average employee-HPWS, and branch average employee
human capital, psychological empowerment, extrinsic motivation,
and perceived organizational support. We do not assume a high
agreement across individuals on these variables within a branch;
they are controlled for merely to demonstrate the effect of branch
overall employee service performance on customer satisfaction
beyond the potential effects of these variables on customer satis-
faction at the branch level.
Analytical Strategy
In this study, employees were nested in employment groups, and
employment groups were nested in branches. The theoretical mod-
els involving employee outcomes are also hierarchical, with con-
structs spanning three levels of analysis. Specifically, employee
individual service performance, employee-HPWS, human capital,
psychological empowerment, and perceived organizational sup-
port are conceptualized at the individual employee level of anal-
ysis; management-HPWS targeted at three different groups within
the branch is conceptualized at the group level of analysis; and
branch characteristics, such as size, age, and level of local com-
petition, are conceptualized at the branch-level of analysis. In
addition, customers are nested in branches. The theoretical models
involving customer satisfaction as the outcome also have two
levels, with individual customers’ satisfaction and demographic
characteristics conceptualized at the individual customer level of
analysis and branch collective service performance conceptualized
at the branch level of analysis.
Therefore, we conducted three-level hierarchical linear model-
ing (HLM; Bryk & Raudenbush, 1992) to test Hypotheses 1
through 4, involving employee outcomes, and two-level HLM to
test Hypothesis 5, involving customer outcomes. Before conduct-
ing the HLM analyses, we conducted a series of analyses of
covariance (ANCOVAs), analyses of variance (ANOVAs), and
group mean comparison tests to examine differences in employee
experience with the work system across different groups, the
380 LIAO, TOYA, LEPAK, AND HONG
variance within the same group, and the divergence between
manager and employee perspectives of the work system. These
analyses, although not part of our formal hypothesis testing, serve
to provide the empirical examination for the potential importance
of focusing on employee individually experienced HPWS.
Results
Descriptive statistics, correlations, and alpha values for the
study variables are presented in Table 2.
Employee–High Performance Work System:
Differences Between Employee Groups, Variance
Within a Group, and Divergence From the
Management’s Perspective
Differences between employee groups. In order to understand
whether there are significant differences in employee-HPWS
among the three different employee groups, we conducted
ANCOVA at the individual level, with group membership as the
treatment variable and employee gender and age as the covariates.
The results revealed that group membership was a significant
factor in determining employee-HPWS, F(2, 826) ⫽4.84, p⬍
.0001. We further conducted pairwise mean comparisons for em-
ployee Group 1, Group 2, and Group 3 across branches. The results
revealed significant differences between Group 1 and Group 2 (for
Group 1, M⫽3.49, SD ⫽0.02; for Group 2, M⫽3.19, SD ⫽
0.03; t(87) ⫽10.68, p⬍.001, n⫽88), between Group 2 and
Group 3 (for Group 2, M⫽3.21, SD ⫽0.03; for Group 3, M⫽
3.03, SD ⫽0.04; t(70) ⫽3.80, p⬍.001, n⫽71), and between
Group 1 and Group 3 (for Group 1, M⫽3.49, SD ⫽0.02; for
Group 3, M⫽3.05, SD ⫽0.04; t(73) ⫽10.68, p⬍.001, n⫽74).
3
Across branches, Group 1 generally had the most favorable eval-
uations of the work system, followed by Group 2, and then Group
3, which is consistent with the company’s categorization of em-
ployment status.
Variance within a group. We also examined whether there
were significant between-employee differences in participants’
experience of HPWS within employment status groups. Specifi-
cally, we followed the procedure recommended in Bryk and Rau-
denbush (1992) and performed an ANOVA using a two-level null
HLM analysis (for Level 1, n⫽831; for Level 2, n⫽253), with
employee individual experience of the work system as the outcome
variable. The results showed that the within-group variance esti-
mate was .15 and the between-group variance estimate was .03
(p⬍.001). This indicates that 17% of the variance in employee
experience of HPWS resided between employee groups, whereas
83% of the variance resided within employee groups.
Divergence from the management’s perspective. Our data
also allowed us to examine whether for a given employment
group, there was a discrepancy between employee-HPWS and
management-HPWS. First, we performed mean comparisons
between management and employee perspectives of the work
system. The results revealed that across the 253 employee
groups in different branches, the group mean of management-
HPWS for a certain group was 3.50 (SD ⫽0.02), and the group
mean of employee-HPWS for the same group was 3.25 (SD ⫽
0.02), which was significantly different from, and lower than,
the management-HPWS, t(252) ⫽⫺10.58, p⬍.001, n⫽253.
This suggests that management generally tends to have a more
positive evaluation of the work system than do the employees.
On the other hand, the correlation at the group-level between
management-HPWS for a group and the average employee-
HPWS in that group was .39, and its 95% confidence interval
does not include 0 or 1. These results suggest that overall there
is a divergence between management-HPWS and the group
mean of employee-HPWS, although the two perspectives are
positively related to each other.
Three-Level Hierarchical Linear Modeling Predicting
Individual Employee Outcomes
Hypotheses 1 through 3 predicted that employee-HPWS would
positively influence employee service performance through the
mediation of employee human capital, psychological empower-
ment, and perceived organizational support. We followed the
four-step test procedures for mediation described in Kenny, Kashy,
and Bolger (1998); controlled for the effects of management-
HPWS and a variety of individual, group, and branch characteris-
tics; and report the results in Tables 3 and 4. As a first step,
employee-HPWS needs to be related to employee service perfor-
mance. As shown in Model 2 and Model 5 in Table 4, employee-
HPWS had a significant positive relationship with supervisor-rated
employee general service performance (␥ˆ⫽.48, p⬍.001) and
knowledge-intensive service performance (␥ˆ⫽.24, p⬍.001).
To test Hypothesis 1 regarding the role of human capital as the
mediator, in the second step, we examined whether employee-
HPWS was related to the mediator. As shown in Model 3 in Table
3, employee-HPWS was positively related to human capital (␥ˆ⫽
.51, p⬍.001). Next, in testing the third and fourth steps, we
included both employee-HPWS and the mediator in predicting
general service performance and knowledge-intensive service per-
formance. As reported in Models 3 and 6 in Table 4, employee-
HPWS was no longer significant, yet human capital was signifi-
cantly related to general service performance (␥ˆ⫽.70, p⬍.001)
and knowledge-intensive service performance (␥ˆ⫽.29, p⬍.001).
Therefore, Hypothesis 1 was supported.
We followed the same steps in testing Hypothesis 2 regarding
the role of psychological empowerment as a mediator. As reported
in Model 5 of Table 3, employee-HPWS was positively related to
empowerment (␥ˆ⫽.53, p⬍.001). Models 3 and 6 in Table 4
showed that empowerment was not significantly related to general
service performance but was significantly related to knowledge-
intensive service performance (␥ˆ⫽.14, p⬍.01). Therefore,
empowerment mediated the relationship between employee-
HPWS and knowledge-intensive service performance, partially
supporting Hypothesis 2.
Likewise, Hypothesis 3 was tested by first examining the relation-
ship between employee-HPWS and the mediator, perceived organi-
zational support, which was shown to be significant in Model 7 of
Table 3 (␥ˆ⫽1.43, p⬍.001). Next, in Models 3 and 6 of Table 4,
whereas employee-HPWS became nonsignificant in predicting ser-
vice performance, perceived organizational support was significantly
3
Because not all branches had data for all three groups, the group means
and sample size involved in the mean comparison tests were slightly
different.
381
HIGH-PERFORMANCE WORK SYSTEMS
Table 2
Descriptive Statistics, Intercorrelations, and Internal Consistency of Study Variables
Variable MSD 1 2 3 456789101112
Individual⫺level employee variables (N⫽830)
1. Employee gender
a
0.57 0.50 —
2. Employee age 32.58 8.76 ⫺.37
ⴱ
—
3. Employee⫺HPWS 3.25 0.43 ⫺.31
ⴱ
.05 .82
4. Supervisor⫺rated employee
human capital 4.52 1.20 .11
ⴱ
⫺.04 .15
ⴱ
.94
5. Employee psychological
empowerment 3.22 0.51 ⫺.34
ⴱ
.26
ⴱ
.49
ⴱ
.27
ⴱ
.89
6. Employee extrinsic
motivation ⫺0.02 1.00 ⫺.45
ⴱ
.07
ⴱ
.42
ⴱ
.04 .48
ⴱ
—
7. Employee POS 5.11 0.92 ⫺.12
ⴱ
⫺.08
ⴱ
.65
ⴱ
.18
ⴱ
.27
ⴱ
.24
ⴱ
.89
8. Supervisor⫺rated employee
GSP 4.89 0.96 .04 .03 .20
ⴱ
.89
ⴱ
.29
ⴱ
.06 .23
ⴱ
.97
9. Supervisor⫺rated employee
KSP 3.14 0.93 ⫺.63
ⴱ
.28
ⴱ
.33
ⴱ
.35
ⴱ
.42
ⴱ
.38
ⴱ
.19
ⴱ
.45
ⴱ
.96
Group⫺level employee variables (N⫽252)
1. Employee Group 1 0.36 0.48 —
2. Employee Group 2 0.35 0.48 ⫺.55
ⴱ
—
3. Employee Group 3 0.29 0.46 ⫺.48
ⴱ
⫺.47
ⴱ
—
4. Management⫺HPWS 3.50 0.36 .26
ⴱ
.30
ⴱ
⫺.59
ⴱ
—
5. Group average
employee⫺HPWS 3.25 0.31 .55
ⴱ
⫺.15
ⴱ
⫺.32
ⴱ
.39
ⴱ
—
Individual⫺level customer variables (N⫽1,772)
1. Customer gender
a
0.54 0.50 —
2. Customer age 48.83 13.58 .02 —
3. Customer total household
asset 3.21 2.07 .08
ⴱ
.45
ⴱ
—
4. Customer overall
satisfaction with branch
service 4.45 0.78 .05
ⴱ
.27
ⴱ
.17
ⴱ
.88
Branch⫺level variables (N⫽75)
1. Branch age 40.11 18.39 —
2. Branch size 15.56 7.36 .46
ⴱ
—
3. Number of competing banks
in neighborhood 3.58 2.95 .05 .39
ⴱ
—
4. Branch average
management⫺HPWS 3.55 0.24 .12 ⫺.02 .00 —
5. Branch average
employee⫺HPWS 3.33 0.14 ⫺.02 ⫺.17 .⫺03 .29
ⴱ
—
6. Branch average human
capital 4.48 0.58 .13 .18 ⫺.12 .24
ⴱ
.18 —
7. Branch average
psychological empowerment 3.38 0.22 ⫺.16 ⫺.36
ⴱ
⫺.13 .12 .44
ⴱ
.03 —
8. Branch average extrinsic
motivation 0.05 0.33 .00 ⫺.07 .09 .25
ⴱ
.49
ⴱ
⫺.16 .51
ⴱ
—
9. Branch average POS 5.17 0.39 ⫺.10 ⫺.09 .00 .23
ⴱ
.73
ⴱ
.26
ⴱ
.24
ⴱ
.16 —
10. Branch average overall GSP 5.05 0.30 ⫺.16 ⫺.07 .01 .15 .56
ⴱ
.18 .49
ⴱ
.39
ⴱ
.53
ⴱ
—
11. Branch average overall KSP 3.70 0.23 ⫺.07 ⫺.14 ⫺.05 .04 .26
ⴱ
⫺.13 .39
ⴱ
.33
ⴱ
.16 .51
ⴱ
—
12. Branch average customer
overall satisfaction 4.46 0.25 .12 ⫺.14 ⫺.36
ⴱ
.04 .03 ⫺.02 .19 .10 .06 .18 .37
ⴱ
—
Note. Numbers 1–12 in the top row correspond to the variables in the respective sections of the table. Coefficient alpha values are presented in italics
along the diagonal. Employee-HPWS ⫽Employee individually experienced high performance work system; POS ⫽perceived organizational support;
GSP ⫽general service performance; KSP ⫽knowledge-intensive financial service performance; management-HPWS: Management’s perspective of the
HPWS practices generally in use for an employee group of certain employment status. Branch average management-HPWS was calculated by averaging
across branch employees the management-HPWS score assigned to the employees.
a
Female ⫽1; male ⫽0.
ⴱ
p⬍.05.
382 LIAO, TOYA, LEPAK, AND HONG
related to general service performance (␥ˆ⫽.06, p⬍.05) but was not
significantly related to knowledge-intensive service performance.
Therefore, Hypothesis 3 was partially supported, in that perceived
organizational support mediated the relationship between employee-
HPWS and general service performance.
We conducted Sobel (1982) tests and confirmed that the change in
the significance of employee-HPWS in predicting general service
performance due to the introduction of human capital (z⫽4.09, p⬍
.001) and perceived organizational support (z⫽2.43, p⬍.05) was
significant. Similarly, Sobel tests revealed that human capital (z⫽
3.93, p⬍.001) and psychological empowerment (z⫽2.93, p⬍.01)
significantly reduced the significance of employee-HPWS in predict-
ing knowledge-intensive service performance.
In sum, employee human capital and perceived organizational
support fully mediated the relationship between employee-HPWS and
employee general service performance, whereas human capital and
psychological empowerment fully mediated the relationship between
employee-HPWS and employee knowledge intensive service perfor-
mance, providing some support for Hypotheses 1 through 3.
We followed a similar procedure to test Hypothesis 4, which
proposes that management-HPWS would influence employee human
capital, psychological empowerment, and perceived organizational
support through the mediation of employee-HPWS. As reported in
Model 1 of Table 3, management-HPWS was not significantly related
to employee-HPWS. Therefore Hypothesis 4 was not supported. As
shown in Models 3, 5, and 7 of Table 3, management-HPWS was
positively related to human capital (␥ˆ⫽.56, p⬍.01) but was
insignificantly related to psychological empowerment or perceived
organizational support. The results suggest that management-HPWS
had a direct effect on individual employees’ human capital instead of
a mediated effect through employee-HPWS.
Two-Level Hierarchical Linear Modeling Predicting
Customer Satisfaction
Hypothesis 5 predicted that branch overall service performance
would be positively related to customer overall satisfaction with
the branch’s service. We conducted HLM analysis to test this
hypothesis, treating customer satisfaction as an individual cus-
tomer level outcome. This approach allowed us to achieve two
goals, which could not be achieved if we had aggregated customer
satisfaction to the branch level and used collective performance to
predict aggregated customer satisfaction using an ordinary least
squares regression. First, we are able to account for customers’
individual differences in age and gender, which have been shown
to relate to customer satisfaction ratings (e.g., Liao & Chuang,
2004). Second, the number of customer respondents per branch
ranged from 1 to 62, with an average of 24; HLM can account for
this differential precision of information of each branch by using
generalized least squares to estimate the fixed effects, which are
typically reported as the final estimates of the impact of the
predictors on the outcome variables (Hofmann, 1997). In the GLS
procedure, branches with more observations, hence, more reliable
and precise Level 1 estimates, receive more weight (Bryk &
Raudenbush, 1992). Fixed effects can thus be viewed as the
weighted average of the Level 1 coefficients across branches. The
results in Model 4 of Table 5 revealed that branch overall
knowledge-intensive service performance was positively related to
Table 3
Three-Level Hierarchical Linear Modeling Results for Employee High-Performance Work Systems and Mediators
Variable
Employee-
HPWS
Employee knowledge,
skills, and abilities:
Supervisor-rated
employee human capital
Employee motivation:
Employee psychological
empowerment
Employee obligation:
Employee perceived
organization support
Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7
Intercept 3.43
ⴱⴱⴱ
4.13
ⴱⴱⴱ
3.99
ⴱⴱⴱ
3.19
ⴱⴱⴱ
3.04
ⴱⴱⴱ
5.73
ⴱⴱⴱ
5.35
ⴱⴱⴱ
Level 1: Employee level
Employee gender
a
⫺.33
ⴱⴱⴱ
.74
ⴱⴱ
.89
ⴱⴱⴱ
⫺.16 .001 ⫺.36 .04
Employee age ⫺.002 .004 .004 .01
ⴱⴱⴱ
.01
ⴱⴱⴱ
⫺.02
ⴱⴱ
⫺.01
ⴱⴱⴱ
Employee-HPWS .51
ⴱⴱⴱ
.53
ⴱⴱⴱ
1.43
ⴱⴱⴱ
Level 2: Group level
Employee Group 1
b
⫺.02 ⫺.05 .13 .08 .27
ⴱⴱ
⫺.77
ⴱⴱ
⫺.29
Employee Group 2
b
.08 ⫺.27
ⴱ
⫺.29
ⴱ
.19
ⴱⴱ
.17
ⴱ
⫺.30
ⴱⴱ
⫺.35
ⴱⴱⴱ
Management-HPWS .09 .53
ⴱⴱ
.56
ⴱⴱ
⫺.11 ⫺.07 .03 .12
Group average employee-HPWS .74
ⴱⴱ
.25 .49
ⴱⴱⴱ
⫺.04 1.45
ⴱⴱⴱ
.02
Level 3: Branch level
Branch age ⫺.001 .003 .003 ⫺.001 ⫺.001 ⫺.003 ⫺.003
Branch size ⫺.003 .01 .01 .000 .001 ⫺.003 ⫺.004
Number of competing banks in
neighborhood ⫺.001 ⫺.01 ⫺.01 .000 .000 ⫺.001 .000
Pseudo R
2c
0.12 0.07 0.09 0.19 0.32 0.17 0.46
Note. For Level 1, N⫽830; for Level 2, N⫽252; for Level 3, N⫽91. In all models, all variables except for employee gender, employee Group 1, and
employee Group 2 were grand-mean centered. Entries corresponding to the predicting variables are estimations of the fixed effects (␥values) with robust
standard errors. Employee-HPWS ⫽employee individually experienced HPWS; management-HPWS ⫽management’s perspective of the HPWS practices
generally in use for an employee group of certain employment status.
a
Female ⫽1; male ⫽0.
b
The omitted, comparison employee group consisted of part-time employees.
c
Pseudo R
2
values were calculated on the basis
of the formula from Kreft and De Leeuw (1998).
ⴱ
p⬍.05.
ⴱⴱ
p⬍.01.
ⴱⴱⴱ
p⬍.001.
383
HIGH-PERFORMANCE WORK SYSTEMS
customer overall satisfaction (␥ˆ⫽.34, p⬍.05), providing some
support for Hypothesis 5.
Discussion
The primary goal of this study was to examine how employee-
HPWS affected individual performance in the service context.
Although researchers in the area of strategic HRM research have
suggested that management and employee perspectives might di-
verge (e.g., Bowen & Ostroff, 2004; Wright & Boswell, 2002), the
current study is among the first to examine the difference and
relationship between the management and employee perspectives
of HPWS. In addition, in the strategic HRM literature, researchers
have examined the link between the use of HPWS and a myriad of
organizational performance outcomes at the firm level of analysis.
However, organizations do not “perform”; individuals in organi-
zations perform in ways that allow the organizations to achieve
desirable performance outcomes (Kozlowski & Klein, 2000).
Thus, individual performance remains an important performance
criterion for management and psychology research to assess the
effectiveness of work systems. The current study sheds light on the
impact and influence process of a service-quality– oriented HPWS
on individual performance provided to customers. The results
provide several implications for research and practice.
Research Implications
One of the major implications of the findings for this study is
that there is a disconnect between what management says they are
implementing and what employees report they are experiencing in
terms of the HPWS practices. We found that although the corre-
lation between management-HPWS and group average employee-
HPWS was positive overall, the managerial ratings were signifi-
cantly higher than employee ratings of the HPWS. In addition, the
effect of management-HPWS on individual employee-HPWS was
not significant when they were entered into the three-level hierar-
chical linear model. Relatedly, our analyses revealed substantial
variance in employee-HPWS among groups of employees with
different status and among employees within the same group.
These findings demonstrate the variability in HR practices across
employees within the same unit, which has been largely ignored by
prior strategic HRM research (Wright & Boswell, 2002). The
findings thus provide empirical support for the arguments of Lepak
and Snell (1999, 2002) that multiple work systems may exist
within the same establishment and for the arguments of Guzzo and
Noonan (1994) as well as Bowen and Ostroff (2004) that individ-
ual employees may experience and interpret the same set of HR
practices differently.
A second major implication is that the results highlight areas of
distinction and overlap between the impact of management perspec-
tive of the HPWS practices generally implemented and employee
individual experience with the HPWS on individual employee out-
comes. From the employee perspective, employee-HPWS had a direct
positive impact on employee human capital, psychological empow-
erment, and perceived organizational support, which were in turn
related to general and knowledge-intensive service performance. Hu-
man capital and perceived organizational support fully mediated the
Table 4
Three-Level Hierarchical Linear Modeling Results Predicting Individual Employees’ Service Performance
Level and variable
Supervisor-rated employee general
service performance
Supervisor-rated employee knowledge-
intensive financial service performance
Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
Intercept 4.59
ⴱⴱⴱ
4.45
ⴱⴱⴱ
4.83
ⴱⴱⴱ
3.10
ⴱⴱⴱ
3.03
ⴱⴱⴱ
3.24
ⴱⴱⴱ
Level 1: Individual level
Employee gender
a
.55
ⴱⴱⴱ
.70
ⴱⴱⴱ
.05 ⫺.38
ⴱⴱ
⫺.31
ⴱ
⫺.57
ⴱⴱⴱ
Employee age .01 .01 .01
ⴱⴱ
.01
ⴱⴱ
.01
ⴱⴱ
.01
ⴱⴱ
Employee-HPWS .48
ⴱⴱⴱ
.06 .24
ⴱⴱⴱ
⫺.01
Supervisor-rated employee human capital .70
ⴱⴱⴱ
.29
ⴱⴱⴱ
Employee psychological empowerment .003 .14
ⴱⴱ
Employee extrinsic motivation ⫺.02 .02
Employee perceived organizational support .06
ⴱ
.01
Level 2: Group level
Employee Group 1
b
⫺.02 .15 .06 .48
ⴱⴱ
.57
ⴱⴱ
.46
ⴱⴱ
Employee Group 2
b
⫺.20 ⫺.22 ⫺.01 ⫺.005 ⫺.01 .03
Management-HPWS .40
ⴱⴱ
.42
ⴱⴱ
.03 .33
ⴱ
.35
ⴱ
.19
Group average employee-HPWS .64
ⴱⴱ
.17 .02 .55
ⴱⴱⴱ
.32
ⴱ
.28
ⴱ
Level 3: Branch level
Branch age .002 .002 .000 .002 .002 .001
Branch size .01 .01 .000 ⫺.002 ⫺.002 ⫺.005
Number of competing banks in neighborhood ⫺.01 ⫺.01 .001 ⫺.01
ⴱⴱ
⫺.01
ⴱⴱ
⫺.01
ⴱ
Pseudo R
2c
0.06 0.09 0.80 0.44 0.45 0.60
Note. For Level 1, N⫽830; for Level 2, N⫽252; and for Level 3, N⫽91. In all models, all variables except for employee gender, employee Group
1, and employee Group 2 were grand-mean centered. Entries corresponding to the predicting variables are estimations of the fixed effects (␥values), with
robust standard errors. HPWS ⫽high-performance work system.
a
Female ⫽1; male ⫽0.
b
The omitted, comparison employee group consisted of Group 3, or part-time, employees.
c
Pseudo R
2
values were calculated
on the basis of the formula from Kreft and De Leeuw (1998).
ⴱ
p⬍.05.
ⴱⴱ
p⬍.01.
ⴱⴱⴱ
p⬍.001.
384 LIAO, TOYA, LEPAK, AND HONG
relationship between employee-HPWS and general service perfor-
mance, whereas human capital and psychological empowerment fully
mediated the relationship between employee-HPWS and knowledge-
intensive service performance. These results provide some support for
the individual-level theory of job performance (Campbell et al.,
1993), suggesting that individuals’ knowledge, skills, and abilities and
motivation are proximal determinants of individual performance and
that the employee experience with HR interventions influences job
performance through its impact on the individuals’ knowledge, skills,
and abilities and motivation. More importantly, the effects of
employee-HPWS on these individual outcomes sustained after we
accounted for the effects of management-HPWS and employee aver-
age perceptions of the HPWS within the group, demonstrating the
utility of considering employees’ idiosyncratic experience with the
HPWS above and beyond management perspective of the HR prac-
tices generally implemented for an employee group and the employ-
ees’ shared perceptions of the HR practices within the group.
Whereas employee-HPWS was directly related to both em-
ployee human capital and motivation (psychological empower-
ment and perceived organizational support), we found that
management-HPWS was directly related only to employee human
capital. One possible explanation of this finding is that some
aspects of employee attributes (i.e., motivation) are more directly
influenced by employee interpretation of the work system than are
other aspects of employee attributes (i.e., human capital), and thus
management-HPWS may have a stronger direct impact on the
latter form of attributes. For example, providing training programs
may positively influence employee human capital independent of
employees’ perceptions of the systems. If the management invests
in the development of employee competencies for service delivery,
their human capital—their knowledge, skills, and abilities to pro-
vide quality service—should be positively affected, even if they
are not aware of or do not perceive those investments as occurring.
In contrast, influencing employee motivation to perform is depen-
dent on employee personal understanding and interpretation of the
HPWS practices.
Taken in combination, these results suggest that we should not
assume homogeneity of employee experience with the HPWS
across employees of different employment groups or even within
the same group, nor should we assume an equivalence of the
employee perspective with the management perspective. Thus, it is
important to directly assess employees’ individual experiences
with the work system in theoretical development and empirical
testing of the effects of the HPWS on individual-level employee
outcomes. Future research may also examine explicitly what gives
rise to employees’ differential experiences by considering the
potential antecedents we mentioned in this study, such as employ-
ment status, organizational justice, and leader–member exchange
relationships. In addition, Bowen and Ostroff (2004) proposed a
set of HR system meta-features that may influence the level to
which employees build shared perceptions about the HR system,
including visibility of the HR practices, understandability of the
HR content, legitimate authority of the HR system, relevance of
the HR system to the strategic goal, instrumentality of the HR
system for employee consequences, validity of the HR practices,
consistency of the HR messages, agreement among principal HR
decision makers, and fairness of the practices. A concerted effort
to assess the effects of these features may represent a promising
avenue for future research.
Beyond examining the divergence between, and the relative
impact of, management-HPWS and employee-HPWS on employee
human capital and motivation and, in turn, on service performance,
Table 5
Three-Level Hierarchical Linear Modeling Results Predicting Individual Customers’ Overall
Satisfaction With Branch Service
Level and variable Model 1 Model 2 Model 3 Model 4
Intercept 4.41
ⴱⴱⴱ
4.41
ⴱⴱⴱ
4.42
ⴱⴱⴱ
4.42
ⴱⴱⴱ
Level 1: Individual customer level
Customer gender
a
.07 .07 .07 .07
Customer age .01
ⴱⴱⴱ
.01
ⴱⴱⴱ
.01
ⴱⴱⴱ
.01
ⴱⴱⴱ
Customer total household asset .02 .02 .02 .02
Level 2: Branch level
Branch age .000 .000 .001 .000
Branch size ⫺.001 ⫺.001 ⫺.001 ⫺.001
Number of Competing Banks in Neighborhood ⫺.01 ⫺.01 ⫺.01 ⫺.01
Branch average management-HPWS ⫺.05 ⫺.06 ⫺.06 ⫺.03
Branch average employee-HPWS .11 ⫺.18 ⫺.27
Branch average human capital ⫺.01 .01
Branch average psychological empowerment .07 .02
Branch average extrinsic motivation .03 ⫺.01
Branch average POS .14 .13
Branch average overall GSP .004
Branch average overall KSP .34
ⴱⴱ
Pseudo R
2
0.071 0.071 0.069 0.074
Note. For Level 1, N⫽1,772; for Level 2, N⫽75. In all models, Level 1 variables except for customer gender were
grand-mean centered. Entries corresponding to the predicting variables are estimations of the fixed effects (␥values),
with robust standard errors. POS ⫽perceived organizational support; GSP ⫽general service performance; KSP ⫽
knowledge-intensive financial service performance.
a
Female ⫽1; male ⫽0.
ⴱⴱ
p⬍.01.
ⴱⴱⴱ
p⬍.001.
385
HIGH-PERFORMANCE WORK SYSTEMS
we also examined the linkages between branch-level service per-
formance and customer satisfaction. Of the two types of service
performance, employees’ overall knowledge-intensive service per-
formance at the branch level was positively related to customer
satisfaction with branch overall service quality, whereas employee
general service performance was not. This result suggests that all
performance metrics might not be created equal in helping firms
gain a competitive service advantage. The resource-based view of
the firm (Barney, 1991) suggests that sustainable competitive
advantage comes from a firm’s resources and capabilities that are
valuable, rare, imperfectly imitable, and not substitutable. We
surmise that the behaviors of general service performance, such as
being courteous, attentive, and reliable to customers and having a
neat, professional appearance may be strictly scripted, or required,
behaviors in banking services and part of customers’ standard
expectation of a service encounter; thus, it may be difficult for
branches to compete for higher customer satisfaction on the basis
of general service performance. Instead, superior knowledge-
intensive service performance may be rarer and less imitable and
hence help differentiate superior banking experiences from others
and create a competitive advantage in service. We encourage
future researchers examining HPWS to keep in mind what perfor-
mance they are referring to in the specific context and to go
beyond general performance metrics by paying attention to the
special metrics that differentiate firms from their competitors.
We found that the HPWS, from both the management and the
employee perspectives, did not have a significant relationship with
customer satisfaction. The lack of a direct link between the two,
however, does not imply that the HPWS is not important in
customer service. The impact of the HPWS on customer satisfac-
tion may be a distal effect, which is likely to be “transmitted
through additional links in a causal chain” (Shrout & Bolger, 2002,
p. 429). Schneider, Ehrhart, Mayer, Saltz, and Niles-Jolly (2005)
also argued that researchers need not necessarily despair when key
organizational variables do not seem to be related to the bottom
line when investigating bivariate relationships. These authors
called for the development of richer models with process variables,
which may reveal that these key variables are indeed important,
although not directly so. The current study does just that. By
introducing employee service performance as the process variable,
we bridge the gap between strategic HRM research and service
management research. The indirect effect of the HPWS on cus-
tomer satisfaction is consistent with the general argument of the
strategic HRM literature that HPWS affects employee behaviors,
which, in turn, affect organizational operational and financial
performance (e.g., see Lepak et al., 2006; Ostroff & Bowen, 2000;
Wright, Gardner, & Moynihan, 2003). This finding is also consis-
tent with the growing evidence from the service linkage research,
which suggests that through front-line employees’ service behav-
iors, internal organizational management transforms into desirable
external customer outcomes (e.g., Heskett, Sasser, & Schlesinger,
1997; Liao, 2007; Liao & Chuang, 2004, 2007; Schneider et al.,
1998; 2005).
Though not the main focus of this study, the results also pro-
vided some interesting implications regarding gender differences
in service performance. We found that women were rated higher
than men in general service performance, whereas the reverse was
true for knowledge-intensive service performance. This finding is
consistent with several lines of gender research. First, gender and
personality theory suggests that men and women are predisposed
to different orientations, with men being more instrumental and
women being more expressive (Evans & Steptoe, 2003). Our
measures of general service performance reflected more social,
relational performance (i.e., reliability, responsiveness, assurance,
and empathy), whereas knowledge-intensive service performance
indicated technical performance. One interpretation of this finding
is that women may excel in social relations, whereas men may
excel in their technical performance. Second, gender role theory
posits that because social norms prescribe women to be communal
and men to be agentic, different genders are “prescribed” with
different behaviors and will try to match themselves with the social
norm (Heilman & Chen, 2005). This reinforces the tendency for
women to perform well in social aspects of service and for men to
perform well in the technical aspect of service. Alternatively, from
the observers’ perspective, gender stereotypes analogous to the
first two points may influence the way supervisors evaluate sub-
ordinates’ performance (Heilman & Haynes, 2005). Thus, men and
women may be evaluated differently by supervisors who are
predisposed to emphasize certain aspects of their performance.
Therefore, the gender differences we observed here might not be a
unique phenomenon in Japan. Nonetheless, research is needed to
examine potential cultural influences in the role of gender on
service performance.
Lastly, we found that employees in Group 1 had better
knowledge-intensive financial service performance than did em-
ployees in Group 3 or Group 2 after a variety of antecedents of
employee performance had been accounted for. We surmise that
because Group 1 has the highest employment status, better per-
formers are likely to be attracted to, selected by, and retained in
this group (attraction–selection–attrition [ASA] theory; Schneider,
1987; Schneider, in press; Schneider, Goldstein, & Smith, 1995).
Therefore, there seem to be other factors (e.g., availability of
resources) that make Group 1 employees better performers beyond
all of the individual, group, and branch factors that we have
specified in this study. In any case, because we explicitly con-
trolled for group membership in our analyses, we partialed out the
effects of the unspecified factors associated with group member-
ship on performance.
Practical Implications
The current study also offers several practical implications.
Marketing research has shown that customer satisfaction is asso-
ciated with customer intention to repurchase and spread positive
word of mouth (e.g., Parasuraman, Zeithaml, & Berry, 1988) and
with financial measures such as return on assets and stock prices
(e.g., Ittner & Larcker, 1996). For example, as Liao and Sub-
ramony (2008) noted, marketing research has shown that a 1%
increase in the American Customer Satisfaction Index (ACSI)
score (Fornell, Johnson, Anderson, Cha, & Bryant, 1996) increases
a medium-sized firm’s (e.g., with $54 billion in assets) future cash
flow by $55 million and reduces the variability of the firm’s cash
flow by more than 4% (Gruca & Rego, 2005). Thus, customer
satisfaction can help firms increase both the volume and the
stability of their future cash flow, hence creating greater share-
holder value. Our findings suggest that a key factor related to
customer satisfaction in the banking service context is branch
employees’ knowledge-intensive service performance; the two
386 LIAO, TOYA, LEPAK, AND HONG
variables had a correlation of .37, with a 95% confidence interval
ranging from .16 to .55, indicating that branch employee
knowledge-intensive service performance accounted for between
3% and 30% of the variance in customer satisfaction. Thus, bank
management may achieve desirable customer and organizational
financial outcomes by providing high-quality professional service.
Our results also provide concrete suggestions on how to improve
knowledge-intensive service performance. By implementing a sys-
tem of service-quality– oriented HPWS practices, employees may
acquire the knowledge, skills, and abilities relevant for the delivery
of superior professional service, become psychologically empow-
ered to appreciate the significant meaning in the tasks, feel the
competence and control they have in performing the tasks, and see
the impact they can make on organizational success.
In addition, we found that branch employees’ general service
performance was not significantly related to customer satisfaction.
However, this does not suggest that general service performance is
unimportant. In fact, a deviation from these typically standard and
expected behaviors will result in customer dissatisfaction
(Solomon, Surprenant, Czepiel, & Gutman, 1985). Thus, compa-
nies should not focus exclusively on employee knowledge-
intensive performance, ignoring the necessity of employee general
service performance. Our results suggest that implementing a
service quality-oriented HPWS will likely help maintain a high
level of general service performance. Under such a system, em-
ployees will acquire the human capital needed for general service
performance and perceive a high level of organizational support
which, in turn, motivates them to provide superior general
service performance to customers as reciprocation to the orga-
nization’s favorable treatment.
These findings also suggest that management should pay par-
ticular attention to employees’ idiosyncratic experiences of the
work system, which directly influence their level of human capital,
psychological empowerment, perceived organizational support,
and service performance. Because HR practices communicate
messages on what management expects, supports, and rewards
(Bowen & Ostroff, 2004; Schneider, 1990), management needs to
improve communications and prevent well-intended HR practices
from being misunderstood by employees. Increased dialogue be-
tween management and employees and regular use of employee
surveys and discussion groups may help management better
understand what employees actually experience in the work-
place and reduce the discrepancy between management and
employee perspectives.
Limitations and Future Research
We acknowledge that the study findings should be considered in
light of several limitations. First, we conducted the analyses using
data from establishments within a Japanese firm. As a result, there
may be some concerns regarding the generalizability of these
findings to other cultural contexts. At the same time, however, the
conceptual arguments used to derive the hypotheses are not cul-
turally bound, and the findings are consistent with the conceptual
arguments developed throughout the strategic HRM literature. In
addition, we explicitly controlled for factors such as employee
employment status, the categorization of which is unique in the
Japanese culture context, making the results about the key study
variables interpretable in other culture contexts. Relatedly, this
sample was based on 91 establishments from a single banking
organization. It is possible that there may be a corporate culture
that constrains the choice of HR practices used in the establish-
ments and employees’ reactions to those practices. Thus, range
restriction might have rendered the effects of the work system on
employee and customer outcomes to be underestimated. At the
same time, however, the compatibility across branches in terms of
products, services, pricing, marketing strategies, and so forth helps
reduce concerns about potential confounding effects due to differ-
ences on these factors. Nevertheless, future research needs to test
the relationships across a wider array of organizations.
A second potential limitation of this study is that we focused
only on an HPWS targeted toward service quality. In reality,
organizations need to juggle multiple demands and achieve mul-
tiple goals simultaneously. For example, in the banking industry,
service quality is a critical component of organizations’ success;
meanwhile, organizations need to improve safety, reduce work-
place discrimination and inequity, increase shareholder value, and
act as responsible member of the community. Future theoretical
and empirical work is needed to provide guidance on how to
balance the needs and align the interests of different organizational
stakeholders in designing the work system.
A third limitation is that although we obtained data from five
rating sources in three time periods, because of practical con-
straints, we were not able to completely eliminate common method
bias. In particular, employees provided evaluations of both their
own experiences with the work system and their psychological
empowerment and perceived organizational support, rendering the
observed relationships among these variables subject to common
source bias. However, conceptually it was necessary to have
employees self-report these measures because it was the individ-
uals’ idiosyncratic experience and perceptions that were of con-
cern. In addition, we measured employee-HPWS and the other
employee-rated variables at two separate times in two surveys,
thus reducing the common method bias.
In addition, supervisors evaluated individual employees’ human
capital as well as their service performance. On the one hand, the
performance evaluation literature suggests that supervisory ratings
of the different aspects of a subordinate often are subject to halo
bias, which may explain the high correlation between the em-
ployee human capital and general service performance measures.
On the other hand, the high correlation may suggest that human
capital is a key driver of employee service performance (account-
ing for 80% of the variance in employee general service perfor-
mance). This is reasonable, as an employee with high levels of
service-related knowledge, skills, and abilities should be in a better
position to provide more reliable, responsive, and confident ser-
vice. Nonetheless, we call for future research to examine the
relationship using ratings from different sources. We should note
that even given the high correlation between human capital and
employee general service performance, employee perceived orga-
nizational support still had a significant positive relationship with
employee general service performance beyond the effect of human
capital, providing convincing evidence for the importance of em-
ployee perceived organizational support, thus lending support to
the social exchange theory.
A fourth limitation is the short time period under which this
research took place. Although the data were collected at different
points in time, it is possible that the effects of the work system on
387
HIGH-PERFORMANCE WORK SYSTEMS
individual employees may take longer to materialize. Longitudinal
research that examines the relationship between HPWSs and the
important outcomes over different lengths of time may provide
unique insights into not only the nature of the relationships but also
the time lag necessary to realize the benefits of the work systems.
In addition, future research that adopts the field experiment meth-
odology is especially needed to test the causal influences of the
work systems on employee and organizational outcomes.
Lastly, we took a cross-level approach in the current study to fill the
void in strategic HRM research by examining the effects and influ-
ence process of the macrolevel HPWS on microlevel employee
performance outcomes. Future research, similar to that reported in
Chen, Thomas, and Wallace (2005); DeShon, Kozlowski, Schmidt,
Milner, and Weichmann (2004); and Ployhart, Weekley, and
Baughman (2006), may model the exemplar multilevel research to
examine simultaneously the impact and influence processes of
HPWSs on performance outcomes at both the individual and unit
levels of analysis and to test the homology (e.g., Chen, Bliese, &
Mathieu, 2005) for the effects of such systems at multiple levels.
Concluding Remarks
Despite its limitations, the current study has a number of
strengths. The key theoretical advancements include the following:
(a) examining the differences and relationship between manage-
ment and employee perspectives of the HPWS for service quality,
(b) integrating various theoretical perspectives to propose a cross-
level framework for the effects of HPWS on individual-level
service performance, and (c) directly testing, rather than assuming,
the intermediate mechanisms. The key empirical and methodolog-
ical strength lies in the use of HLM to account for the hierarchical
nature of our models and data and in the collection of data in two
formats from three phases and five rating sources. Our study
design reduced common-method bias and allowed us to examine
simultaneously the interfaces between management, front-line em-
ployees, and customers and to examine how internal management
practices and employee psychological processes influence impor-
tant external performance criteria. The key practical implications
relate to showing the greater importance of branch overall
knowledge-intensive service performance over general service
performance in influencing customer satisfaction and to dem-
onstrating that the HPWS for service quality was directly, as
well as indirectly, related to employee knowledge-intensive
service performance.
In conclusion, this study bridges the gap between macrolevel
and microlevel approaches to HRM and extends strategic HRM
research to the service arena and to individual level processes and
performance outcomes. The findings contribute to our understand-
ing of how a service-quality– oriented HPWS influences individual
service performance, in part, through the mediation of employee
human capital, psychological empowerment, and perceived orga-
nization support, and highlight the importance of incorporating the
employee perspective into the examination of HPWS.
References
Allen, D., Shore, L. M., & Griffeth, R. W. (2003). The role of perceived
organizational support and supportive human resource practices in the
turnover process. Journal of Management, 29, 99 –118.
Anderson, J. C., & Gerbing, D. W. (1988). Structural equation modeling in
practice: A review and recommended two-step approach. Psychological
Bulletin, 103, 411– 423.
Asubonteng, P., & McCleary, K. J. (1996). SERVQUAL revisited: A
critical review of service quality. Journal of Services Marketing, 10,
62– 81.
Bagozzi, R. P., Yi, Y., & Phillips, L. W. (1991). Assessing construct
validity in organizational research. Administrative Science Quarterly,
36, 421– 458.
Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral
change. Psychological Review, 84, 191–215.
Barney, J. B. (1991). Firm resources and sustained competitive advantage.
Journal of Management, 17, 99 –120.
Batt, R. (2002). Managing customer services: Human resources practices,
quit rates, and sales growth. Academy of Management Journal, 45,
587–597.
Becker, B., & Gerhart, B. (1996). The impact of human resource manage-
ment on organizational performance: Progress and prospects. Academy
of Management Journal, 39, 779 – 801.
Becker, B. E., & Huselid, M. A. (1998). High performance work systems
and firm performance: A synthesis of research and managerial implica-
tions. In G. R. Ferris (Ed.), Research in personnel and human resources
management (Vol. 16, pp. 53–101). Greenwich, CT: JAI Press.
Blau, P. M. (1964). Exchange and power in social life. New York: Wiley.
Bliese, P. D. (2000). An introduction to multilevel modeling techniques
[Book review]. Personnel Psychology, 53, 1062–1065.
Bowen, D. E., & Ostroff, C. (2004). Understanding HRM-firm perfor-
mance linkages: The role of the “strength” of the HRM system. Academy
of Management Review, 29, 203–221.
Bowen, D. E., & Schneider, B. (1988). Services marketing and manage-
ment: Implications for organizational behavior. Research in Organiza-
tional Behavior, 10, 43– 80.
Bryk, A. S., & Raudenbush, S. W. (1992). Hierarchical linear models.
Newbury Park, CA: Sage.
Campbell, J. P., McCloy, R. A., Oppler, S. H., & Sager, C. E. (1993). A
theory of performance. In N. Schmitt, & W. C. Borman (Eds.), Person-
nel selection in organizations: 35–70. San Francisco: Jossey-Bass.
Chen, G., & Bliese, P. D. (2002). The role of different levels of leadership
in predicting self- and collective efficacy: Evidence for discontinuity.
Journal of Applied Psychology, 87, 549 –556.
Chen, G., Bliese, P. D., & Mathieu, J. E. (2005). Conceptual framework
and statistical procedures for delineating and testing multilevel theories
of homology. Organizational Research Methods, 8, 375– 409.
Chen, G., Thomas, B., & Wallace, J. C. (2005). A multilevel examination
of the relationships among training outcomes, mediating regulatory
processes, and adaptive performance. Journal of Applied Psychology,
90, 827– 841.
Colquitt, J. A., Conlon, D. E., Wesson, M. J., Porter, C. O., & Yee, N. K.
(2001). Justice at the millennium: A meta-analytic review of 25 years of
organizational justice research. Journal of Applied Psychology, 86, 425–
445.
Datta, D. K., Guthrie, J. P., & Wright, P. M. (2005). Human resource
management and labor productivity: Does industry matter? Academy of
Management Journal, 48, 135–145.
Deci, E. L. (1975). Intrinsic motivation. New York: Plenum.
Delery, J. E. (1998). Issues of fit in strategic human resource management:
Implications for research. Human Resource Management Review, 8,
289 –309.
Delery, J. E., & Doty, D. H. (1996). Modes of theorizing in strategic human
resource management: Tests of universalistic, contingency, and config-
urational performance predictions. Academy of Management Journal,
39, 802– 835.
Delery, J. E., & Shaw, J. D. (2001). The strategic management of people
in work organizations: Review, synthesis, and extension. In G. R. Ferris
388 LIAO, TOYA, LEPAK, AND HONG
(Ed.), Research in personnel and human resources management (Vol.
20, 165–197). Stamford, CT: JAI Press.
DeShon, R. P., Kozlowski, S. W. J., Schmidt, A. M., Milner, K. R., &
Weichmann, D. (2004). A multiple goal, multilevel model of feedback
effects on the regulation of individual and team performance in training.
Journal of Applied Psychology, 89, 1035–1056.
Drasgow, F., & Kanfer, R. (1985). Equivalence of psychological measure-
ment in heterogeneous populations. Journal of Applied Psychology, 70,
662– 680.
Eisenberger, R., Cummings, J., Armeli, S., & Lynch, P. (1997). Perceived
organizational support, discretionary treatment, and job satisfaction.
Journal of Applied Psychology, 82, 812– 820.
Eisenberger, R., Fasolo, P., & Davis-LaMastro, V. (1990). Perceived
organizational support and employee diligence, commitment, and inno-
vation. Journal of Applied Psychology, 75, 51–59.
Evans, O., & Steptoe, A. (2003). Gender-related psychological character-
istics and situational determinants of psychophysiological stress reactiv-
ity. Journal of Applied Social Psychology, 33, 756 –774.
Fornell, C., Johnson, M. D., Anderson, E. W., Cha, J., & Bryant, B. E.
(1996). The American Customer Satisfaction Index: Nature, purpose,
and findings. Journal of Marketing, 58, 7–18.
Graen, G. B., & Uhl-Bien, M. (1995). Relationship-based approach to
leadership: Development of leader–member exchange (LMX) theory of
leadership over 25 years: Applying a multi-level multi-domain perspec-
tive. Leadership Quarterly, 6, 219 –247.
Gruca, T. S., & Rego, L. L. (2005). Customer satisfaction, cash flow, and
shareholder value. Journal of Marketing, 69, 115–130.
Gutek, B. A. (1995). The dynamics of service: Reflections on the changing
nature of customer/provider interactions. San Francisco, CA: Jossey-
Bass.
Gutek, B. A., Bhappu, A. D., Liao-Troth, M. A., & Cherry, B. (1999).
Distinguishing between service relationships and encounters. Journal of
Applied Psychology, 84, 218 –233.
Guthrie, J. P. (2001). High-involvement work practices, turnovers, and
productivity: Evidence from New Zealand. Academy of Management
Journal, 44, 180 –190.
Guzzo, R. A., & Noonan, K. A. (1994). Human resource practices as
communications and the psychological contract. Human Resource Man-
agement, 33, 447– 462.
Gwinner, K. P., Bitner, M. J., Brown, S. W., & Kumar, A. (2005). Service
customization through employee adaptiveness. Journal of Service Re-
search, 8, 131–148.
Heilman, M., & Chen, J. (2005). Same behavior, different consequences:
Reactions to men’s and women’s altruistic citizenship behavior. Journal
of Applied Psychology, 90, 431– 441.
Heilman, M. E., & Haynes, M. C. (2005). No credit where credit is due:
Attributional rationalization of women’s success in male–female teams.
Journal of Applied Psychology, 90, 905–916.
Heskett, J., Sasser, W. E. J., & Schlesinger, L. (1997). The service profit
chain: How leading companies link profit and growth to loyalty, satis-
faction, and value. New York: Free Press.
Hofmann, D. A. (1997). An overview of the logic and rationale of hierar-
chical linear models. Journal of Management, 23, 723–744.
Huselid, M. A. (1995). The impact of human resource management prac-
tices on turnover, productivity, and corporate financial performance.
Academy of Management Journal, 38, 635– 672.
Ittner, C., & Larcker, D. F. (1996). Measuring the impact of quality
initiatives on firm financial performance. In S. Ghosh & D. Fedor (Eds.),
Advances in the management of organizational quality (pp. 1–37).
Greenwich, CT: JAI.
Jackson, S. E., & Schuler, R. S. (1995). Understanding human resource
management in the context of organizations and their environments.
Annual Review of Psychology, 46, 237–264.
James, L. R., Demaree, R. G., & Wolf, G. (1984). Estimating within-group
interrater reliability with and without response bias. Journal of Applied
Psychology, 69, 85–98.
James, L. R., James, L. A., & Ashe, D. K. (1990). The measuring of
organizations: The role of cognition and values. In B. Schneider (Ed.),
Organizational climate and culture (pp. 40 –89). San Francisco: Jossey-
Bass.
Joshi, A., Liao, H., & Jackson, S. E. (2006). Cross-level effects of work-
place diversity on sales performance and pay. Academy of Management
Journal, 49, 459 – 481.
Kanter, R. M. (1983). The change masters: Innovation and entrepreneur-
ship in the American corporation. New York: Simon & Schuster.
Keltner, B. (1995). Relationship banking and competitive advantage: Ev-
idence from the U.S. and Germany. California Management Review, 37,
45–73.
Kenny, D. A., Kashy, D. A., & Bolger, N. (1998). Data analysis in social
psychology. In D. Gilbert, S. Fiske & G. Lindzey (Eds.), The handbook
of social psychology (4th ed., Vol. 1, pp. 233–265). Boston: McGraw-
Hill.
Kishton, J. M., & Widaman, K. F. (1994). Unidimensional versus domain
representative parceling of questionnaire items: An empirical example.
Educational and Psychological Measurement, 54, 757–765.
Kozlowski, S. W. J., & Hattrup, K. (1992). A disagreement about within-
group agreement: Disentangling issues of consistency versus consensus.
Journal of Applied Psychology, 77, 161–167.
Kozlowski, S. W. J., & Klein, K. J. (2000). A multilevel approach to theory
and research in organizations: Contextual, temporal, and emergent pro-
cesses. In K. J. Klein, & S. W. J. Kozlowski (Eds.), Multilevel theory,
research, and methods in organizations (pp. 3–90). San Francisco, CA:
Jossey-Bass.
Kreft, I., De Leeuw, J. (1998). Introducing multilevel modeling. London:
Sage.
Lado, A. A., & Wilson, M. C. (1994). Human resource systems and
sustained competitive advantage: A competency-based perspective.
Academy of Management Review, 19, 699 –727.
Lambert, S. J. (2000). Added benefits: The link between work–life benefits
and organizational citizenship behavior. Academy of Management Jour-
nal, 43, 801– 815.
Lepak, D. P., Liao, H., Chung, Y., & Harden, E. (2006). A conceptual
review of human resource management systems in strategic human
resource management research. In J. Martocchio (Ed.), Research in
personnel and human resource management (pp. 217–271).Stamford,
CT: JAI Press.
Lepak, D. P., & Snell, S. A. (1999). The human resource architecture:
Toward a theory of human capital allocation and development. Academy
of Management Review, 24, 31– 48.
Lepak, D. P., & Snell, S. A. (2002). Examining the human resource
architecture: The relationships among human capital, employment, and
human resource configurations. Journal of Management, 28, 517–543.
Lepak, D. P., Taylor, M. S., Tekleab, A. G., Marrone, J. A., & Cohen, D. J.
(2007). An examination of the use of high-investment human resource
systems for core and support employees. Human Resource Management,
46, 223–246.
Levitt, T. (1972). Production line approach to services. Harvard Business
Review, 50, 41–50.
Liao, H. (2007). Do it right this time: The role of employee service
recovery performance in customer perceived justice and customer loy-
alty after service failures. Journal of Applied Psychology, 92, 475– 489.
Liao, H., & Chuang, A. (2004). A multilevel investigation of factors
influencing employee service performance and customer outcomes.
Academy of Management Journal, 47, 41–58.
Liao, H., & Chuang, A. (2007). Transforming service employees and
climate: A multilevel, multisource examination of transformational lead-
ership in building long-term service relationships. Journal of Applied
Psychology, 92, 1006 –1019.
389
HIGH-PERFORMANCE WORK SYSTEMS
Liao, H., & Subramony, M. (2008). Employee customer orientation in
manufacturing organizations: Joint influences of customer proximity and
senior leadership team. Journal of Applied Psychology, 93, 317–328.
Liden, R. C., & Graen, G. B. (1980). Generalizability of the vertical dyad
linkage model of leadership. Academy of Management Journal, 23,
451– 465.
Melian-Gonzalez, S., & Verano-Tacoronte, D. (2006). Is there more than
one way to manage human resources in companies? Personnel Review,
35, 29 –50.
Miles, R. E., & Snow, C. C. (1984). Designing strategic human resources
systems. Organizational Dynamics, (Summer), 36 –52.
O’Farrell, P. N., Hitchens, D. M., & Moffat, L. A. R. (1993). The com-
petitive advantage of business service firms: A matched pairs analysis of
the relationship between generic strategy and performance. Service
Industries Journal, 13, 40 – 64.
Osterman, P. (1987). Choice of employment systems in internal labor
markets. Industrial Relations, 26, 46 – 67.
Ostroff, C., & Bowen, D. E. (2000). Moving human resource to a higher
level: Human resource practices and organizational effectiveness. In
K. J. Klein & S. W. J. Kozlowski (Eds.), Multilevel theory, research, and
methods in organizations (pp. 211–266). San Francisco: Jossey-Bass.
Parasuraman, A., Zeithaml, V. A., & Berry, L. L. (1988). SERVQUAL: A
multiple-item scale for measuring consumer perceptions of service qual-
ity. Journal of Retailing, 64, 12– 40.
Parasuraman, A., Zeithaml, V. A., & Berry, L. L. (1994). Reassessment of
expectations as a comparison standard in measuring service quality:
Implications for further research. Journal of Marketing, 58, 111.
Pfeffer, J. (1998). The human equation: Building profits by putting people
first. Boston: Harvard Business School Press.
Pfeffer, J., & Davis-Blake, A. (1987). The effect of the proportion of
women on salaries: The case of college administrators. Administrative
Science Quarterly, 32, 1–24.
Ployhart, R. E., Weekley, J. A., & Baughman, K. (2006). The structure and
function of human capital emergence: A multilevel examination of the
attraction–selection–attrition model. Academy of Management Journal,
49, 661– 677.
Porter, M. E. (1980). Competitive strategy. New York: Free Press.
Pugh, S. D. (2001). Service with a smile: Emotional contagion in the
service encounter. Academy of Management Journal, 44, 1018 –1027.
Reichheld, F. (1996). The loyalty effect. Boston: Harvard Business School
Press.
Rousseau, D. M. (1995). Psychological contracts in organizations. Thou-
sand Oaks, CA: Sage.
Rousseau, D. M. (2005). I-deals: Idiosyncratic deals employees bargain
for themselves. Armonk, NY: Sharpe.
Rust, R. T., Moorman, C., & Dickson, P. R. (2002). Getting return on
quality: Revenue expansion, cost reduction, or both? Journal of Mar-
keting, 66, 7–24.
Ryan, R. M., Mims, V., & Koestner, R. (1983). Relation of reward
contingency and interpersonal context to intrinsic motivation: A review
and test using cognitive evaluation theory. Journal of Personality and
Social Psychology, 45, 736 –750.
Schlesinger, L. A., & Heskitt, J. L. (1991). Breaking the cycle of failure in
services. Sloan Management Review, 17, 57– 68.
Schneider, B. (1987). The people make the place. Personnel Psychology,
40, 437– 454.
Schneider, B. (1990). The climate for service: An application of the climate
construct. In B. Schneider (Ed.), Organizational climate and culture:
383– 412. San Francisco: Jossey-Bass.
Schneider, B. (in press). The people still make the place. In D. B. Smith
(Ed.), The people make the place: Dynamic linkages between people and
organizations—A festschrift in honor of Benjamin Schneider. Mahwah,
NJ: Erlbaum.
Schneider, B., & Bowen, D. E. (1985). Employee and customer perceptions
of service in banks: Replication and extension. Journal of Applied
Psychology, 70, 423– 433.
Schneider, B., Ehrhart, M. G., Mayer, D. M., Saltz, J. L., & Niles-Jolly, K.
(2005). Understanding organization– customer links in service settings.
Academy of Management Journal, 48, 1017–1032.
Schneider, B., Goldstein, H. W., & Smith, D. B. (1995). The ASA frame-
work: An update. Personnel Psychology, 48, 747–773.
Schneider, B., White, S. S., & Paul, M. C. (1998). Linking service climate
and customer perceptions of service quality: Tests of a causal model.
Journal of Applied Psychology, 83, 150 –163.
Seibert, S. E., Silver, S. R., & Randolph, W. A. (2004). Taking empow-
erment to the next level: A multiple-level model of empowerment,
performance, and satisfaction. Academy of Management Journal, 47,
332–349.
Seiders, K., Voss, G. B., Grewal, D., & Godfrey, A. L. (2005). Do satisfied
customers buy more? Examining moderating influences in a retailing
context. Journal of Marketing, 69, 26 – 43.
Settoon, R. P., Bennett, N., & Liden, R. C. (1996). Social exchange in
organizations: Perceived organizational support, leader–member ex-
change, and employee reciprocity. Journal of Applied Psychology, 81,
219 –227.
Shore, L. M., & Tetrick, L. E. (1991). A construct validity study of the
survey of perceived organizational support. Journal of Applied Psychol-
ogy, 76, 637– 643.
Shore, L. M., Tetrick, L. E., Taylor, M. S., Coyle-Shapiro, J., Liden, R.,
McLean Parks, J., Morrison, E. W., Porter, L. W., Robinson, S., Roeh-
ling, M., Rousseau, D., Schalk, R., Tsui, A., & Van Dyne, L. (2004). The
employee– organization relationship: A timely concept in a period of
transition. In J. Martocchio (Ed.), Research in personnel and human
resources management (Vol. 23, pp. 291–370). Greenwich, CT: JAI
Press.
Shore, L. M., & Wayne, S. J. (1993). Commitment and employee behavior:
Comparison of affective commitment and continuance commitment with
perceived organizational support. Journal of Applied Psychology, 78,
774 –780.
Shrout, P. E., & Bolger, N. (2002). Mediation in experimental and non-
experimental studies: New procedures and recommendations. Psycho-
logical Methods, 74, 422– 445.
Snell, S. A., Youndt, M. A., & Wright, P. M. (1996). Establishing a
framework for research in strategic human resource management: Merg-
ing resource theory and organizational learning. In G. R. Ferris (Ed.),
Research in personnel and human resources management (pp. 61–90).
Greenwich, CT: JAI Press.
Sobel, M. E. (1982). Asymptotic confidence intervals for indirect
effects in structural equation models. In S. Leinhardt (Ed.), Socio-
logical methodology (pp. 290 –312). Washington, DC: American
Sociological Association.
Solomon, M. R., Surprenant, C., Czepiel, J. A., & Gutman, E. G. (1985).
A role theory perspective on dyadic interactions: The service encounter.
Journal of Marketing, 49, 99 –111.
Spreitzer, G. M. (1995). Psychological empowerment in the workplace:
Dimensions, measurement, and validation. Academy of Management
Journal, 38, 1442–1465.
Subramaniam, M., & Youndt, M. A. (2005). The influence of intellectual
capital on the types of innovative capabilities. Academy of Management
Journal, 48, 450 – 463.
Tetrick, L. E., Shore, L. M., Newton, L. A., & Vandenberg, R. J. (2007).
A model of union participation: The impact of perceived union support,
union instrumentality, and union loyalty. Journal of Applied Psychology,
92, 820 – 828.
The world factbook. Washington, DC: Central Intelligence Agency. Re-
trieved October 5, 2007 from https://www.cia.gov/library/publications/
the-world-factbook/index.html
Thomas, K. W., & Velthouse, B. A. (1990). Cognitive elements of em-
390 LIAO, TOYA, LEPAK, AND HONG
powerment: An “interpretive” model of intrinsic task motivation. Acad-
emy of Management Review, 15, 666 – 681.
Tyagi, P. K. (1985). Relative importance of key job dimensions and
leadership behaviors in motivating salesperson work performance. Jour-
nal of Marketing, 86, 76 – 86.
Wayne, S. J., Shore, L. M., Bommer, W. H., & Tetrick, L. E. (2002). The
role of fair treatment and rewards in perceptions of organizational
support and leader–member exchange. Journal of Applied Psychology,
87, 590 –598.
Wayne, S. J., Shore, L. M., & Liden, R. C. (1997). Perceived organiza-
tional support and leader–member exchange: A social exchange per-
spective. Academy of Management Journal, 40, 82–111.
Wright, P. M., & Boswell, W. R. (2002). Desegregating human resource
management: A review and synthesis of micro and macro human re-
source management research. Journal of Management, 24, 247–276.
Wright, P. M., & McMahan, G. C. (1992). Theoretical perspectives for
strategic human resource management. Journal of Management, 18,
295–320.
Youndt, M. A., Subramaniam, O., & Snell, S. A. (2004). Intellectual capital
profiles: An examination of investments and returns. Journal of Man-
agement Studies, 41, 335–361.
Yuan, K. H., Bentler, P. M., & Kano, Y. (1997). On averaging variables in
a confirmatory factor analysis model. Behaviormetrika, 24, 71– 83.
Zacharatos, A., Barling, J., & Iverson, R. D. (2005). High-performance
work systems and occupational safety. Journal of Applied Psychology,
90, 77–93.
Received June 12, 2007
Revision received June 16, 2008
Accepted June 27, 2008 䡲
E-Mail Notification of Your Latest Issue Online!
Would you like to know when the next issue of your favorite APA journal will be available
online? This service is now available to you. Sign up at http://notify.apa.org/ and you will be
notified by e-mail when issues of interest to you become available!
391
HIGH-PERFORMANCE WORK SYSTEMS
A preview of this full-text is provided by American Psychological Association.
Content available from Journal of Applied Psychology
This content is subject to copyright. Terms and conditions apply.
