Academy of Management The Academy of Management Journal
Academy of Management The Academy of Management Journal
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms
Academy of Management is collaborating with JSTOR to digitize, preserve and extend access
to The Academy of Management Journal
This content downloaded from 152.118.24.10 on Wed, 07 Nov 2018 05:11:02 UTC
All use subject to https://about.jstor.org/terms
® Academy of Management Journal
2012, Vol. 55, No. 6, 1264-1294.
http://dx.doi.org/10.5465/amj.2011.0088
KAIFENG JIANG
DAVID P. LEPAK
JIAHU
University of Notre Dame
JUDITH C. BAER
Rutgers, the State University of New Jersey
In the past two decades, researchers in strategic 2002), higher productivity and quality (MacDuffie,
human resource management (HRM) have exam 1995), better service performance (Chuang & Liao,
ined why and how organizations achieve their 2010), enhanced safety performance (Zacharatos,
goals through the use of human resource (HR) prac Barling, & Iverson, 2005), and better financial per
tices. Although traditional HRM research has fo formance (Huselid, 1995).
cused on the impact of individual HR practices, the Despite the robust evidence for the positive rela
strategic perspective on HRM research emphasizes tionships between HRM and various organizational
bundles of HR practices, often referred to as high outcomes (Combs, Liu, Hall, & Ketchen, 2006), im
performance work systems (HPWS), high-involve portant issues remain regarding the mechanisms
ment work systems, and high-commitment work through which HRM is associated with different
systems, in examinations of the effects of HRM on organizational outcomes. First, the theoretical logic
employee and organizational outcomes (Wright & underlying the mechanisms linking HRM and
McMahan, 1992). A burgeoning body of strategic organizational outcomes remains fragmented
HRM research has shown that the use of systems
(Huselid & Becker, 2011; Wright & Gardner, 2003).
of HR practices intended to enhance employees'
Specifically, some researchers have adopted a be
knowledge, skills, and abilities, motivation, and
havioral perspective to suggest that HR practices
opportunity to contribute is associated with posi
affect organizational outcomes by influencing em
tive outcomes such as greater commitment (Gong,
ployee role behaviors; if employees act in ways that
Law, Chang, & Xin, 2009), lower turnover (Batt,
are consistent with company goals, performance
should improve. Other researchers have adopted
We thank the action editor for this article, Jason Shaw, more of a human capital and resource-based per
and three anonymous reviewers, Patrick McKay, Rebecca
spective, focusing on the potential contributions of
Kehoe, and Mark Huselid for helpful comments and sug
employees' competencies—that is, their knowl
gestions. We acknowledge financial support from the
SHRM Foundation (Project No. 143). The interpretations, edge, skills, and abilities. Interestingly, although
conclusions, and recommendations are those of the au employees contribute through both their competen
thors and do not necessarily represent those of the SHRM cies and their actions, researchers have typically
Foundation. focused on one perspective to understand how HR
1264
Copyright of the Academy of Management, all rights reserved. Contents may not be copied, emailed, posted to a listserv, or otherwise transmitted without the copyright holder's
written permission. Users may print, download, or email articles for individual use only.
This content downloaded from 152.118.24.10 on Wed, 07 Nov 2018 05:11:02 UTC
All use subject to https://about.jstor.org/terms
2012 Jiang, Lepak, Hu, and Baer 1265
systems impact organizational outcomes (excep standing goals of macro HRM research. Inde
Becker and Huselid (1998) considered this relation
tions include Takeuchi, Lepak, Wang, and Takeu
chi [2007]). Considering multiple perspectives
ship as one of the essential pursuits of strategic
HRM research. This stream of research has several
simultaneously provides a broader and more com
plete picture of the relationship between HRM andkey components. First, organizational outcomes are
organizational outcomes. viewed as multidimensional. Drawing on Dyer and
Second, although prior research has demonstrated Reeves's (1995) work, researchers in strategic HRM
the mechanism through which HRM relates to some have categorized organizational outcomes into
organizational outcomes, it remains unclear as to how three primary groups related to HRM: HR out
HRM relates to different organizational outcomes that comes, operational outcomes, and financial out
range from very proximal (i.e., HR outcomes) to more comes. HR outcomes refer to those most directly
distal (i.e., financial outcomes). This lack of integrarelated to HRM in an organization, such as em
tion is problematic given the different perspectives ployee skills and abilities, employee attitudes and
adopted in the literature, perspectives that might
behaviors, and turnover. Operational outcomes are
those related to the goals of an organizational op
highlight the importance of different but potentially
related outcomes. Exploring the possible paths be eration, including productivity, product quality,
tween HRM and financial outcomes will likely pro quality of service, and innovation. Financial out
vide a more integrative model of how HR systems comes reflect the fulfillment of the economic goals
operate to impact a multitude of related and imporof organizations. Typical financial outcomes in
tant outcomes (e.g., Becker & Huselid, 1998; Delery clude
& sales growth, return on invested capital, and
Shaw, 2001; Guest, 1997). return on assets. In this study, we use "organiza
Third, it is assumed in existing research that thetional outcomes" to refer to all three categories of
components of HR systems have identical impacts outcomes at the organizational level.
on outcomes. For example, when scholars adopt an Second, strategic HRM research suggests that dif
additive approach to measure HR systems, eachferent types of outcomes may not necessarily have
equivalent relationships with HR practices (Becker
component of the system is treated as if it exerts an
equal influence on the outcomes under investiga & Huselid, 1998; Delery & Shaw, 2001; Guest, 1997;
tion. Although this is a possible reflection of how Lepak, Liao, Chung, & Harden, 2006; Ostroff & Bo
HR systems operate, scholars have recently chal wen, 2000). Moreover, it is commonly asserted that
lenged this assumption and argued that different HRM may influence the three types of organiza
sets of HR practices may impact the same outcomes tional outcomes in sequence. For example, HR
in a heterogeneous way (e.g., Batt & Colvin, 2011; practices are expected to first influence HR out
Gardner, Wright, & Moynihan, 2011; Gong et al., comes (e.g., employee skills and motivation),
2009; Shaw, Dineen, Fang, & Vellella, 2009; Subra which are proximal and the least likely to be con
mony, 2009). As these studies have suggested, it taminated
is by factors beyond HR practices. HR out
important to explore the differential effects of thecomes, in turn, may mediate the influence of HR
different components of HR systems. practices on productivity, quality, service, safety,
Given these issues, the primary objective of this innovation, and other operational outcomes, which
further affect financial outcomes.
study is to develop an integrative model of the mech
anisms mediating between HRM and organizationalAlthough existing HR research often implies that
outcomes through a meta-analytic approach. Drawing HR outcomes serve as a key mediator between HR
on the behavioral perspective on HRM, human capi systems and key outcomes, the specific natures of
tal theory, and the resource-based view of the firm,
models of this meditation depend on the theoreti
we aim to extend and refine existing HRM-organizacal perspective researchers have adopted when ex
tional outcomes models by exploring multiple mediamining this relationship. On the one hand, several
researchers have adopted the behavioral perspec
ating paths and differentiating among the effects of
subdimensions of HR systems. tive of HRM (Jackson, Schuler, & Rivero, 1989).
According to this perspective, organizations do not
perform themselves, but instead use HR practices
THEORETICAL BACKGROUND AND to encourage productive behaviors from employees
HYPOTHESES and thus to achieve desirable operational and fi
nancial objectives (Becker & Huselid, 1998). If an
Existing Theories and Research on Relationships
organization requires efficient employees, for ex
between HRM and Organizational Outcomes
ample, its chosen HR practices and their effective
Understanding the relationship between ness
HRM would likely differ from those of an organiza
and organizational outcomes is one of the long
tion that requires employees to be cooperative, to
This content downloaded from 152.118.24.10 on Wed, 07 Nov 2018 05:11:02 UTC
All use subject to https://about.jstor.org/terms
1266 Academy of Management Journal December
focus on service, or to engage in some other critical needed to explore how HRM can help organiza
role behavior. The effectiveness of HR practices is tions achieve financial goals through multiple
realized when employees act in ways that are paths (Takeuchi et al., 2007).
needed for implementing strategies and achieving
various business objectives.
Decomposing HR Systems into
On the other hand, some macro HRM researchers
Three HR Dimensions
have focused less on the behaviors of employees
and more on their competencies within organiza Scholars have recently argued that although em
tions. Researchers taking on this perspective often ployees are exposed to HR systems rather than in
invoke human capital theory and the resource dividual practices, the parts of these systems are
based view of the firm. Human capital theory em not necessarily equivalent in their impact. Most
phasizes that human capital—the composition of research has portrayed an HR system as an additive
employee skills, knowledge, and abilities—is a cen index of a set of individual HR practices (Combs et
tral driver of organizational performance when the al., 2006); there are reasons to believe, however,
return on investment in human capital exceeds la that the highly varied set of HR practices can be
bor costs (Becker, 1964; Lepak & Snell, 1999; Ploy categorized into several subdimensions. Indeed,
hart & Moliterno, 2011). The resource-based view dividing HR systems into subdimensions is not
provides additional insights as to why human cap new in strategic HRM research. For example, draw
ital can help firms to outpace competitors and pro ing on an employee-organization relationship
poses that organizations obtain a competitive ad framework (Tsui, Pearce, Porter, & Tripoli, 1997),
vantage from resources that are rare, valuable, researchers have argued that HR practices may be
inimitable, and nonsubstitutable (Barney, 1991; categorized as falling into HRM inducement and
Mahoney & Pandian, 1992). Researchers have ar investment practices, and HRM expectation
gued that human capital, especially high-quality enhancing practices (e.g., Batt & Colvin, 2011; Gong
and/or organization-specific human capital, has the et al., 2009; Shaw et al., 2009; Shaw, Delery, Jen
potential to serve as a source of competitive advan kins, & Gupta, 1998; Shaw, Gupta, & Delery, 2005).
tage (Wright, McMahan, & McWilliams, 1994). Or The first two types are designed to improve em
ganizations may use HR practices to create and ployees' expected outcomes, whereas the third type
maintain valuable human capital, including both reflects organizations' expectations about employ
generic and organization-specific human capital, ees' contributions.
which in turns drives high operational and finan Taking a different approach, some researchers
cial performance (Becker & Huselid, 1998; Delery & have drawn upon the ability-motivation-opportu
Shaw, 2001; Ployhart & Moliterno, 2011; Snell & nity (AMO) model of HRM and suggested that em
Dean, 1992). ployee performance is a function of three essential
Although the behavioral perspective of HRM, hu components: ability, motivation, and opportunity
man capital theory, and the resource-based view of to perform. Extending this logic, HR systems de
the firm let researchers adopt different angles to signed to maximize employee performance can be
look at the relationships between HR practices and viewed as a composition of three dimensions in
more distal outcomes, under all three perspectives tended to enhance employee skills, motivation, and
HR outcomes are viewed as a critical path from opportunity to contribute, respectively (Appel
HRM to operational and financial outcomes. Even baum, Bailey, Berg, & Kalleberg, 2000; Bailey, 1993;
with this agreement, however, researchers have not Boxall & Purcell, 2008; Delery & Shaw, 2001; Ger
successfully combined multiple approaches to de hart, 2007; Katz, Kochan, & Weber, 1985; Lepak et
lineate an overarching picture of how this path al., 2006). Recently, several empirical studies have
unfolds. For example, most of the extant empirical adopted and validated this conceptual framework
research has examined the influence of HR systems (e.g., Bailey, Berg, & Sandy, 2001; Batt, 2002; Gard
on operational or financial performance either ner et al., 2011; Huselid, 1995; Kehoe & Wright, in
through motivation-related variables (e.g., Chuang press; Liao, Toya, Lepak, & Hong, 2009; MacDuffie,
& Liao, 2010; Collins & Smith, 2006; Gelade & Ivery, 1995; Subramony, 2009).
2003; Gong et al., 2009; McClean & Collins, 2011; In keeping with these studies, Lepak and col
Sun, Aryee, & Law, 2007) or through human capital leagues (2006) suggested that it might be fruitful to
variables (e.g., Cabello-Medina, Lopez-Cabrales, & conceptualize HR practices as falling into one of
Valle-Cabrera, 2011; Yang & Lin, 2009; Youndt & three primary dimensions: skill-enhancing HR
Snell, 2004). Insights into each type of variable are practices, motivation-enhancing HR practices,
important yet insufficient to fully capture the pro and opportunity-enhancing HR practices. Skill
cess linking HRM to outcomes. Thus, research is enhancing HR practices are designed to ensure ap
This content downloaded from 152.118.24.10 on Wed, 07 Nov 2018 05:11:02 UTC
All use subject to https://about.jstor.org/terms
2012 Jiang, Lepak, Hu, and Baer 1267
propriately skilled employees; they include com ployees' skills and abilities. For example, recruit
prehensive recruitment, rigorous selection, and ment and selection practices are intended to insure
extensive training. Motivation-enhancing HR prac that employees have the skills needed for task per
tices are implemented to enhance employee moti formance, and training and development may fur
vation. Typical ones include developmental perfor ther provide employees with organization-specific
mance management, competitive compensation,skills with which to perform their work. Indeed,
incentives and rewards, extensive benefits, promo Delaney and Huselid (1996) indicated that organi
tion and career development, and job security. Op zations can enhance the skills of their workforces
portunity-enhancing HR practices are designed to both by hiring high-quality individuals and by im
empower employees to use their skills and motiva proving the level of skills in their current work
tion to achieve organizational objectives. Practices
forces. Relatedly, prior research shows that the use
such as flexible job design, work teams, employee of comprehensive selection and training practices
involvement, and information sharing are generallyfostered employees' collective human capital (e.g.,
used to offer these opportunities. The use of the Cabello-Medina et al., 2011; Takeuchi et al., 2007;
three dimensions of HR systems instead of a unidi Yang & Lin, 2009; Youndt & Snell, 2004). Further
mensional or two-dimensional framework is based more, research suggests that practices such as com
on an examination of differential effects of the three petitive compensation, extensive benefits, and job
dimensions of HR systems on different types of HR security may help attract capable employees and
outcomes. retain them in organizations, and practices such as
work teams, employee involvement, and flexible
job design may provide employees with opportuni
Linking HR Dimensions to Multiple Outcomes
ties to share knowledge and to learn new skills.
However, the relationships between the other two
According to the ability-motivation-opportunity
model of HRM, HR outcomes can conceptually HR dimensions
be and human capital are seen as less
divided into human capital, motivation, anddirect.
opporResearch has shown that practices from
tunity to contribute (Becker & Huselid, 1998;
theseDel
two dimensions were less positively related
ery & Shaw, 2001; Guest, 1997), and human to capital
human capital than skill-enhancing HR practices
and employee motivation are two of the most (Cabello-Medina
crit et al., 2011; Yang & Lin, 2009).
ical mediating factors that have been examined in
Therefore, we propose the following:
the literature (e.g., Gardner et al., 2011; Gong et al.,
2009; Liao et al., 2009; Sun et al., 2007; Takeuchi et
Hypothesis la. Skill-enhancing HR practices
are positively related to human capital.
al., 2007; Youndt & Snell, 2004). In line with the
literature, we focus on the mediating roles of hu Hypothesis lb. Motivation-enhancing HR prac
man capital and employee motivation. As previous tices are positively related to human capital.
research suggests, human capital can be viewed as
Hypothesis 1c. Opportunity-enhancing HR
a composition of employees' knowledge, skills, and
practices are positively related to human
abilities (Coff, 2002), and employee motivation re
capital.
fers to the direction, intensity, and duration of em
ployees' effort (Campbell, McCloy, Oppler, & Sager, Hypothesis 2a. Skill-enhancing HR practices
1993), as manifested by positive work attitudes are more positively related to human capital
(e.g., collective job satisfaction, commitment, per than motivation-enhancing HR practices.
ceived organizational support) and work behaviors
Hypothesis 2b. Skill-enhancing HR practices
(e.g., organizational citizenship behavior).
are more positively related to human capital
Although we anticipate that all three HR dimen
than opportunity-enhancing HR practices.
sions are positively related to both human capital
and employee motivation, we also anticipate that We also posit that the three dimensions of HR
the three HR dimensions may play different roles in systems are positively related to employee motiva
building human capital and enhancing employee tion to different degrees. First, investment in all
motivation. We expect that, compared with moti three HR dimensions generally indicates that organ
vation-enhancing and opportunity-enhancing HR izations value and support employees' contribu
practices, skill-enhancing HR practices will likely tions. According to social exchange theory (Blau,
have a stronger impact on human capital and a 1964) and the norm of reciprocity (Gouldner, 1960),
weaker impact on employee motivation. employees who perceive an organization's actions
According to the ability-motivation-opportunity toward them as beneficial may feel obligated to
framework, skill-enhancing HR practices can di reciprocate and be motivated to exert more effort at
rectly help to optimize the levels or types of em work. More specifically, motivation-enhancing HR
This content downloaded from 152.118.24.10 on Wed, 07 Nov 2018 05:11:02 UTC
All use subject to https://about.jstor.org/terms
1268 Academy of Management Journal December
practices (e.g., performance-based compensation, tinct from human capital and employee motivation
incentives and benefit, promotion opportunities, (e.g., Batt, 2002; Batt & Colvin, 2011; Gardner et al.,
and job security) are more likely to provide em 2011; Guthrie, 2001; Shaw et al., 1998, 2005, 2009;
ployees with extrinsic motivation that links their Sun et al., 2007). Research has consistently demon
work efforts to external rewards. Practices such as strated that HR practices designed to enhance em
work teams, employee involvement, and flexible ployee skills and motivation are significantly and
job design help to generate employees' intrinsic negatively associated with voluntary turnover (e.g.,
motivation, which encourages them to seek out Arthur, 1994; Batt, 2002; Guthrie, 2001; Huselid,
challenges at work (Ryan & Deci, 2000). In addition, 1995). Some researchers attribute the negative rela
skill-enhancing HR practices can enhance employ tionships to the emotional bond between employ
ees' skills and abilities, which may help career ees and organizations formed by HR practices. In
development and induce promotion opportunities other words, because HR practices enhance em
in their organizations (Tharenou, Saks, & Moore, ployees' motivation at work, these employees are
2007). However, the effect of skill-enhancing HR reluctant to leave their organizations (e.g., Gardner
practices on employee motivation is relatively in et al., 2011; Sun et al., 2007). Investment in the
direct and likely to be contingent on the practices three aspects of HR systems implies that organiza
in the other two HR dimensions. For example, even tions value employees' contribution and expect to
though training can improve employees' skills at establish long-term employment relationships with
work, the increased skills may not necessarily lead their employees. As a result, employees are encour
to promotion in their organization. Therefore, we aged to work harder to reciprocate and thus are less
expect all three HR dimensions to be positively prone to quit their jobs.
associated with employee motivation and, com Human capital theory and the resource-based
pared with the other two dimensions, skill-enhanc view of the firm indicate that employees with ap
ing HR practices are less positively related to em propriate human capital resulting from HR invest
ployee motivation. Recent empirical research that ments may be less likely to leave their organiza
examined the influence of three HR dimensions on tions. First, researchers have suggested that
employee affective commitment (Gardner et al.,employees with high levels of human capital are
2011) has also supported this reasoning. Therefore, more capable of meeting job demands, receiving
we hypothesize: positive performance appraisals, obtaining promo
tions, and participating in decision making (Batt &
Hypothesis 3a. Skill-enhancing HR practices Colvin, 2011; Shaw et al., 2009). Therefore, com
are positively related to employee motivation.
pared with those with less human capital, employ
Hypothesis 3b. Motivation-enhancing HR prac ees with higher levels of human capital will be less
tices are positively related to employee likely to leave their organizations. In addition, em
motivation. ployees with high levels of human capital are better
able to learn at work, which facilitates the devel
Hypothesis 3c. Opportunity-enhancing HR
opment of specific human capital (Ployhart & Mo
practices are positively related to employee
motivation.
literno, 2011). The accumulated specific human
capital may in turn reduce the likelihood employ
Hypothesis 4a. Skill-enhancing HR practices ees leave, because the specific human capital that is
are less positively related to employee motiva unique and valuable for their current organization
tion than motivation-enhancing HR practices. may not provide value to other organizations (Bar
ney, 1991; Lepak & Snell, 1999). Employees are
Hypothesis 4b. Skill-enhancing HR practices
unable to obtain return on their input in developing
are less positively related to employee motiva
the specific human capital if they quit (Shaw et al.,
tion than opportunity-enhancing HR practices.
2005). Therefore, we hypothesize:
In addition to the direct effects of the three HR
dimensions on human capital and employee moti Hypothesis 5a. Human capital mediates the
vation, we propose that human capital and em negative relationships between the three di
ployee motivation mediate the relationships be mensions of HR systems and voluntary
tween the three HR dimensions and more distal turnover.
This content downloaded from 152.118.24.10 on Wed, 07 Nov 2018 05:11:02 UTC
All use subject to https://about.jstor.org/terms
2012 Jiang, Lepak, Hu, and Baer 1269
This content downloaded from 152.118.24.10 on Wed, 07 Nov 2018 05:11:02 UTC
All use subject to https://about.jstor.org/terms
1270 Academy of Management Journal December
comes and financial outcomes. In view of these to search conference programs from the Academy
findings, we propose a positive relationship
ofbe
Management (AOM) and the Society of Indus
tween operational and financial outcomes. trial and Organizational Psychology from 2000 to
2010. Second, we referred to the reference lists of
In sum, drawing upon the behavioral perspective
of HRM, human capital theory, and the resource the prior reviews on this topic, including theoreti
based view of the firm, we propose a mediating cal reviews (e.g., Becker & Gerhart, 1996; Becker &
model in which the three dimensions of HR Huselid,
sys 1998; Lengnick-Hall, Lengnick-Hall, An
tems are indirectly related to financial outcomes
drade, & Drake, 2009; Lepak et al., 2006; Wright &
through human capital, employee motivation,Boswell,
vol 2002) and meta-analytic reviews (Combs
untary turnover, and operational outcomes in et se
al., 2006; Subramony, 2009). Third, we made an
quence. In building this framework, we focuseffort
on to identify unpublished studies through the
the mediating role of employees in the link of HRM
listservs of the AOM's Human Resources and Organ
with financial performance. However, our model izational Behavior Divisions.
does not exclude other paths through which HRM Four inclusion criteria were used to select stud
can help increase financial outcomes. In fact, both
ies. First, we focused only on studies that examined
theoretical and empirical research has suggested
the relationships between HR practices and organ
that HRM can provide firms with organizational izational outcomes at the organizational level (e.g.,
establishment, business unit, or firm). We did not
capital reflected by internal fit and flexibility
(Evans & Davis, 2005; Wright & Snell, 1998) andinclude studies that investigated individual-level
social capital (Collins & Clark, 2003; Delery &
relationships between employee-perceived HR
Shaw, 2001; Gittell, Seidner, & Wimbush, 2010), practices/systems and individual outcomes (e.g.,
both of which can be sources of competitive advan Agarwala, 2003; Barling, Kelloway, & Iverson,
tage for organizations. Given these alternative pos 2003) or cross-level relationships between organi
sibilities, we hypothesize that the intermediate out zation-level HR practices and individual-level out
comes proposed in our model partially mediate the comes (e.g., Liao et al., 2009; Takeuchi, Chen, &
positive relationships between the three HR dimen Lepak, 2009). Second, we only included studies
sions and financial outcomes. that emphasized the use of HR practices/systems in
organizations but not the effectiveness or the value
Hypothesis 7. Human capital, employee moti
of these practices or systems (e.g., Huselid, Jackson,
vation, voluntary turnover, and operational
& Schüler, 1997; Richard & Johnson, 2004). Third,
outcomes partially mediate the positive rela
we included studies in the meta-analysis if they
tionships between the three dimensions of HR
reported at least one correlation among individual
systems and financial outcomes.
HR practices and various organizational outcomes.
We excluded the studies that only presented the
METHODS correlations of HR systems rather than those of
Data Collection
individual HR practices with organizational out
comes (e.g., Bae & Lawler, 2000). Studies without
We tested the mediating hypotheses with thethe statistical information (e.g., sample sizes, cor
help of meta-analytic structural equation modeling relation coefficients) necessary to calculate effect
(SEM) techniques (Cheung & Chan, 2005, 2009;sizes were also excluded (e.g., Cappelli & Neumark,
Viswesvaran & Ones, 1995). To identify studies that 2001; Ichniowski, Shaw, & Prennushi, 1997). Fi
could be used in the meta-analysis, we firstnally, when the same sample was used in two or
searched the PsycINFO, Web of Science, and Pro more articles, we considered only the one that pro
Quest Digital Dissertations databases for studiesvided more information. In contrast, when a study
published before May 2011. We used multiple key used two or more independent samples, we coded
words. For HRM, we used the keywords "human these independent samples separately. The inclu
resource work practice/system," "high-perfor sion criteria yielded a final set of 116 articles rep
mance work practice/system," "high-involvement resenting 120 independent samples that included a
work practice/system," or "high-commitment work total of 31,463 organizations.
practice/system," whereas for organizational out We first developed the coding sheet and instruc
comes, we searched for studies that also included tions as recommended by Lipsey and Wilson
the keywords "performance," "outcome," "atti (2001). The first author and the third author then
tudes," "satisfaction," "commitment," "motiva independently coded a random selection of 15 ar
tion," "human capital," "turnover," "productivity," ticles to assess the level of agreement regarding
"quality," "service," "safety," "growth," or "profit sample sizes, effect sizes, and reliability. After both
ability." Moreover, we used the same search terms coders checked data entry and resolved errors, they
This content downloaded from 152.118.24.10 on Wed, 07 Nov 2018 05:11:02 UTC
All use subject to https://about.jstor.org/terms
2012 Jiang, Lepak, Hu, and Baer 1271
independently coded the rest of studies. The con comes by correcting for measurement error and
sensus rate was 96 percent, and disagreements sampling error (Hunter & Schmidt, 2004). We first
were solved through discussion between the two performed reliability corrections for informant-re
coders. ported measures of HR practices and organizational
outcomes to correct for measurement error. For
those studies that did not report the reliabilities of
Operationalization of Variables
the informant-reported measures, we imputed the
Three dimensions of HR systems. We identified reliabilities using the weighted mean of the avail
14 HR practices frequently examined in the litera able reliabilities estimated from the other studies
ture. By following previous research using the abil (Lipsey & Wilson, 2001). Regarding the variables
ity-motivation-opportunity framework (e.g., Appel that were measured with archival data (e.g., return
baum et al., 2000; Batt, 2002; Gardner et al., 2011; on assets), we adopted a more conservative .80
Guest, 1997; Lepak et al., 2006; Subramony, 2009), reliability estimate, which has been used in previ
we categorized these practices into three dimen ous meta-analyses in management (e.g., Dalton,
sions. Skill-enhancing HR practices included re Daily, Certo, & Roengpitya, 2003; Dalton, Daily,
cruitment, selection, and training. Motivation-en Ellstrand, & Johnson, 1998; Dalton, Daily, Johnson,
hancing HR practices consisted of performance & Ellstrand, 1999). For example, if training prac
appraisal, compensation, incentive, benefit, pro tices were measured by reflective items (e.g., "This
motion and career development, and job security. firm invests considerable time and money in train
In addition, opportunity-enhancing HR practices ing") in a study that reported the reliability of these
covered job design, work teams, employee involve items, we would correct for the reliability for train
ment, formal grievance and complaint processes, ing. In contrast, if training practices were measured
and information sharing. by archival data (e.g., "On average how many hours
Organizational outcomes. We summarized var of formal training do employees in this firm receive
ious organizational outcomes into five categories. each year?"), we would correct for a reliability of
Human capital included overall organizational hu .80 for this measure. For comparison purposes, we
man capital measured via established scales (e.g., also calculated the reliability-corrected correlation
Subramaniam & Youndt, 2005; Youndt, Subrama
by using a reliability of 1.00 for archival measures
niam, & Snell, 2004) and the education level of a
and did not find changes in the main findings of
workforce. Employee motivation was reflected by this study.
collective job satisfaction, organizational commit Second, to calculate the composites of HR prac
ment, organizational climate, perceived organiza tices (i.e., HR dimensions) and the composites of
tional support, and organizational citizenship be outcome variables (i.e., organizational outcomes
havior. Voluntary turnover only represented the categories), we combined the correlations among
percentage of employees who quit or voluntarily individual HR practices and outcomes using the
left the organizations. Dismissal rate and overall
formula provided by Hunter and Schmidt (2004:
turnover rate were not included. In addition, we
435-439):
viewed productivity, quality, service, innovation,
2 rxiyj
and overall operational performance as operational
outcomes, and we viewed return on assets, return Txy \ln + ri(n -1) rxixjylm + m(m -1) rylyj'
on equity, market return, sale growth, and overall If it is assumed that x represents a dimension of HR
financial performance as financial outcomes. systems (e.g., skill-enhancing HR practices) and y
As suggested by Aguinis, Pierce, Bosco, Dalton, represents a category of organizational outcomes
and Dalton (2011), we provide a table, in Appendix (e.g., employee motivation), Xrxiyj is the sum of the
A, that lists all the included studies and our cate correlations between HR practices (e.g., recruit
gorizations of the three HR dimensions and differ ment, selection, and training) and outcome vari
ent types of outcomes. This information is impor ables (e.g., collective satisfaction and commit
tant to allow future research to replicate and extend ment); n and m are the numbers of HR practices
this study. and outcome variables respectively; rxixj is the av
erage correlation among HR practices; and ryiyj is
the average correlation among outcome variables.
Meta-analytic and Model-Testing Procedures
By using this formula, we created a single effect
To test the mediating model through meta-ana size for each relationship within each study.
lytic SEM, we needed to calculate meta-analytic Third, we used a random-effects model to correct
correlations among three dimensions of HR sys for the sampling error by weighting each study's
tems and different types of organizational out effect size by its sample size (Hunter & Schmidt,
This content downloaded from 152.118.24.10 on Wed, 07 Nov 2018 05:11:02 UTC
All use subject to https://about.jstor.org/terms
1272 Academy of Management Journal December
2004). We also computed the 95% confidence in 3, and 4, we included all three dimensions of HR
terval (CI) around the sample-weighted mean cor systems in regressions examining their effects on
relation and Q homogeneity statistic. Confidence human capital and employee motivation. As shown
intervals provide an estimate of the variability in Table 2, all three HR dimensions had significant
around the estimated average correlation; a 95% CI and positive effects on human capital. The results
excluding zero indicates that one can be 95 percent of Z-tests show that the regression coefficient of
confident that the confidence interval includes the skill-enhancing HR practices (ß = .29, p < .01) was
average mean true score. The Q statistic indicates significantly larger than the coefficients of motiva
the variance in the sample-weighted mean correla tion-enhancing HR practices (ß = .22, p < .01, Z =
tion; a significant Q suggests the heterogeneity of a 2.74, p < .01) and opportunity-enhancing HR prac
given relationship. Research has suggested that a tices (ß = .07, p < .01, Z = 8.68, p < .01). Moreover,
random-effects model provides a more accurate es the analyses of relative weights indicate that skill
timate than a fixed-effects model when relation enhancing HR practices explained the largest per
ships are heterogeneous (Cheung & Chan, 2005; centage of variance in human capital (48%), fol
Erez, Bloom, & Wells, 1996; Overton, 1998). lowed by motivation-enhancing HR practices
Finally, we used the created correlation matrices (36%) and opportunity-enhancing HR prac
in SEM computed in LISREL 8.72 (Jöreskog & Sör tices (16%).
bom, 2005). Because the sample sizes for differentSimilarly, we found significantly positive effects
correlations were not identical, we imputed the of three HR dimensions on employee motivation.
sample size for the SEM analyses by calculating theConsistently with our prediction, the influences of
harmonic mean of the correlation sample sizes motivation-enhancing HR practices (ß = .29, p <
(Viswesaran & Ones, 1995). Compared with the
.01, Z = -8.64, p < .01) and opportunity-enhanc
arithmetic mean, the harmonic mean gives much ing HR practices (ß = .25, p < .01, Z = -7.07, p <
less weight to large sample sizes and thus results in.01) were significantly stronger than that of skill
a more conservative parameter estimate. Four es enhancing HR practices (ß = .07, p < .01). Motiva
tablished model fit statistics—chi-square (x2), the tion-enhancing HR practices and opportunity-en
root-mean-square error of approximation (RMSEA),hancing HR practices respectively explained 45
the comparative fit index (CFI), and the standard
and 38 percent of the variance of employee moti
ized root-mean-square residual (SRMR)—werevation, whereas skill-enhancing HR practices ex
used to examine the viability of the structural mod
plained 17 percent. In sum, Hypotheses 1 through 4
were supported.
els (Kline, 2005). Acceptable model fit is associated
with nonsignificant chi-square values and with a
CFI greater than .90, an RMSEA less than or equal Mediation Results
to .08, and an SRMR less than .10 (Kline, 2005). We
used two statistics to test the hypotheses predicting Hypotheses 5 through 7 predict that the three HR
relative effects of three HR dimensions on human dimensions have both direct effects and indirect
capital and employee motivation. One was the Z effects through human capital, employee motiva
test, which shows the significance of the differencetion, voluntary turnover, and operational outcomes
between regression coefficients (Clogg, Petkova, &on financial outcomes. We tested the proposed
Haritou, 1995), and the other was the epsilon stamodel (Figure 1) by inputting correlation matrices
tistic, which has been commonly used to determine (Table 1) into LISREL 8.72 (Jöreskog & Sörbom,
the relative weight of each predictor in explaining 2005). As shown in Table 3, the model fit of the
the variance of dependent variables (Johnson, 2000; proposed model was acceptable (;^[9] = 264.82,
Johnson & LeBreton, 2004). The results of relativeRMSEA = .09, CFI = .98, SRMR = .04). All the
weights represent the proportion of total variance proposed relationships among HR dimensions and
(R2) explained by each HR dimension. To analyzeorganizational outcomes categories were signifi
mediation, we used Sobel's (1982) test to examinecant and consistent with our prediction except
the statistical significance of indirect effects. for the direct relationship between opportunity
enhancing HR practices and financial outcomes
(ß = -.03, n.s.). Thus, we dropped this direct path
RESULTS from the model, which only marginally impacted
Differential Effects of HR Dimensions
fit (model 1: A^2[l] = 4.65, p < .05). We also tested
the direct relationships between three HR dimen
Table 1 summarizes the correlation results of the sions and voluntary turnover and operational out
relationships among HR dimensions and organiza comes. As presented in Table 3, adding paths from
tional outcomes categories. To test Hypotheses 1,2, skill-enhancing HR practices to both outcomes sig
This content downloaded from 152.118.24.10 on Wed, 07 Nov 2018 05:11:02 UTC
All use subject to https://about.jstor.org/terms
2012 Jiang, Lepak, Hu, and Baer 1273
TABLE 1
1. Skill-enhancing practices
2. Motivation-enhancing practices (r, rj .38, 46
k IN) 55 (14,670)
95% CI .40: .53
Q 822.75**
Q 557.95** 855.49**
7. Operational outcomes (r, rc) .25, 32 .19, 25 .25, 32 .25, 29 .32, 38 -.15, -.19
k (N) 36 (10,224) 37 (11,041) 35 (9,576) 8 (1,198) 23 (4,618) 22 (6,002)
95% CI .25: .39 .17: .33 .25: .39 .06: .53 .30: .47 -.10: -.27
Q 436.07** 626.61** 354.35** 108.92** 142.24** 189.14**
8. Financial outcomes (r, rc) .22, 26 .22, 27 .15, 20 .19, 24 .32, 38 -.15, -.19 .38, 48
k[N) 41 (9,966) 41 (12,219) 27 (5,610) 12 (2,028) 17 (3,354) 17 (4,055) 33 (8,863)
95% CI .21: .32 .21: .33 .13: .26 .16: .32 .25: .51 -.08: -.30 .39: .57
a The mean sample-size-weighted correlation (r) and mean sample-sized-weighted correlation corrected for attenuation
ability (rc) are presented. A "k" indicates the number of independent samples, and "N" is the total sample size. The
95% confidence interval around the mean sample-size-weighted corrected correlation (rc). Q is the chi-square-test for the hom
corrected correlations (rj across studies.
* p < .05
** p < .01
nificantly improved fit over that of model 1 (model n.s.). Furthermore, we added the direct paths from
2: A*2[2] = 80.71, p < .01). However, the path from motivation-enhancing HR practices to voluntary
skill-enhancing HR practices to voluntary turnover turnover and operational outcomes and found a
was not significant (ß = —.02, p > .05). Dropping significant improvement in the fit over that of
this path did not impact fit (model 3: A*2[l] = 1.56, model 3 (model 4: A^[2] = 10.54, p < .01). The
TABLE 2
a Standardized coefficients are presented. Z is the test for the significance of the difference between the regression
* p < .05
** p < .01
This content downloaded from 152.118.24.10 on Wed, 07 Nov 2018 05:11:02 UTC
All use subject to https://about.jstor.org/terms
Academy of Management Journal December
FIGURE 1
Theoretical Model of Effects of HR Dimensions on Organizational Outcomes
TABLE 3
Fit Statistics for Alternative Models®
Three HR dimensions
264.82 9 .98 .09 .04
Theoretical model (Figure 1)
Alternative model lb 269.47 10 4.65*c .98 .08 .05
a n = 3,724.
b Deletes the direct paths from opportunity-enhancing HR practices to financial outcomes.
c Model fit compared with the theoretical model of the effects of three HR dimensions on organizational outcomes (Figure 1).
d Adds the direct paths from skill-enhancing HR practices to voluntary turnover and operational outcomes.
G Model fit compared with the previous model.
f Deletes the direct paths from skill-enhancing HR practices to voluntary turnover.
8 Adds the direct paths from motivation-enhancing HR practices to both voluntary turnover and operational outcomes.
h Deletes the direct path from motivation-enhancing HR practices to operational outcomes.
' Adds the direct path from opportunity-enhancing HR practices to voluntary turnover and operational outcomes.
' Adds the direct path from HPWS to both voluntary turnover and operational outcomes.
k Model fit compared with the theoretical model of the effects of HPWS on organizational outcomes (Figure 3).
* p < .05
** p < .01
This content downloaded from 152.118.24.10 on Wed, 07 Nov 2018 05:11:02 UTC
All use subject to https://about.jstor.org/terms
2012 Jiang, Lepak, Hu, and Baer 1275
path from motivation-enhancing HR practices and that human capital, employee motivation, volun
operational outcomes was not significant (ßtary = turnover, and operational outcomes partially
mediated the relationships between skill-enhanc
-.02, n.s.), and we dropped it without impacting fit
(model 5: Ax2[l] = 0.76, n.s.). Finally, we added the
ing and motivation-enhancing HR dimensions and
direct paths from opportunity-enhancing HR prac financial outcomes and fully mediated the relation
tices to voluntary turnover and operational out ship between opportunity-enhancing HR practices
comes, and both paths were significant (modeland 6: financial outcomes. Hypotheses 5 through 7
A^2[2] = 50.22, p < .01). Therefore, we kept modelwere generally supported.
6 as the final model for the mediation analyses. We obtained the indirect effects and total effects
Figure 2 presents the standardized path estimates
of the three HR dimensions on financial outcomes
for the final mediating model. Both human capital from the estimates in SEM. The total effects of
and employee motivation were negatively related
skill-enhancing, motivation-enhancing, and oppor
to voluntary turnover (ß = -.20, p < .01 for human
tunity-enhancing HR dimensions on financial ou
capital; ß = -.34, p < .01 for employee motivation)
comes were .13, 18, and .09 respectively (all p's <
and were positively related to operational out
.01). The indirect effects mediated by human cap
comes (ß = .15, p < .01 for human capital; ß = .26,
tal, employee motivation, voluntary turnover, and
p < .01 for employee motivation). In turn, volun
tary turnover was negatively related to financial operational outcomes were .08, 05, and .09 for the
outcomes (ß = -.08, p < .01), whereas operational three HR dimensions respectively. We also calcu
outcomes were positively associated with financiallated the squared multiple correlations (i.e., R2s) for
outcomes (ß = .42, p < .01). Sobel (1982) tests structural equations predicting human capital (.22),
showed that the indirect relationships betweenemployee
all motivation (.25), voluntary turnover
three HR dimensions and voluntary turnover, (.18),
op operational outcomes (.22), and financial out
erational outcomes, and financial outcomes were comes (.26). The results indicate that the final
significant (Z varied from 8.05 to 13.89, all p-values model explained a moderate amount of variance in
were less than .01). In sum, these results suggest these variables.
FIGURE 2
Final Model of Effects of HR Dimensions on Organizational Outcomes®
This content downloaded from 152.118.24.10 on Wed, 07 Nov 2018 05:11:02 UTC
All use subject to https://about.jstor.org/terms
1276 Academy of Management Journal December
FIGURE 3
Theoretical Model of Effects of HPWS on Organizational Outcomes
Skill-Enhancing
HR Practices Human
Capital
Motivation
Enhancing HR
Practices
Opportunity Employee
Enhancing HR Motivation
Practices
This content downloaded from 152.118.24.10 on Wed, 07 Nov 2018 05:11:02 UTC
All use subject to https://about.jstor.org/terms
2012 Jiang, Lepak, Hu, and Baei 1277
FIGURE 4
Effects of HPYVS on Organizational Outcomes8
1
Voluntary
Turnover
R2= .15
—.05*
Financial
if Outcomes
R2= .27
.37*
Operational
Outcomes
R2= .22
age between HRM and operational and financial Davis, 2005; Gittell et al., 2010) and enhancing the
outcomes. social capital of organizations (Collins & Clark,
Moreover, this study embraced the multidimen
2003). The findings of the current study and others
sionality of performance as well as the potential
suggest
forthat it is meaningful for future research to
different relationships with proximal andfurther
distal explore other mediators of the relationship
outcomes. Researchers have recently called for HRM and organizational outcomes.
between
studies to simultaneously examine multiple One
outmajor contribution of this study to the stra
come variables that have only been studiedtegic
indeHRM literature is that the results suggest dif
pendently before (Lengnick-Hall et al., 2009). With
ferential effects of the three dimensions of HR sys
the help of meta-analytic techniques, we tested a finding is important both in theory and
tems. This
comprehensive mediating model and provided em
in the methodology of measuring HR systems. The
pirical support for the theoretical propositionoretically,
that this finding challenges previous re
HRM first relates to proximal outcomes, which search,
fur in which the assumption has been that all
ther relate to distal outcomes (Becker & Huselid, HR practices in an HR system function in the same
1998; Delery & Shaw, 2001; Dyer & Reeves, 1995; pattern. Our findings remind researchers that dif
Guest, 1997) and revealed that the relationships ferent dimensions of HR systems may have unique
between HRM and distal outcomes (e.g., opera relationships with specific organizational out
tional and financial outcomes) could be mediated comes. For example, skill-enhancing HR practices
through multiple pathways (e.g., through human were more effective in enhancing human capital,
capital and employee motivation). Moreover, as we whereas motivation-enhancing HR practices and
expected, there were direct relationships between opportunity-enhancing HR practices were more
skill-enhancing HR practices and motivation-en likely to improve employee motivation. This result
hancing HR practices and financial outcomes that is also consistent with recent research suggesting
could not be explained by the mediating process. the heterogeneous effects of the components of HR
This is consistent with prior research suggesting systems on organizational outcomes (e.g., Batt &
that HRM can improve organizational effectives Colvin, 2011; Gardner et al., 2011; Gong et al., 2009;
through alternative approaches such as affecting Liao et al., 2009; Shaw et al., 2009; Subramony,
internal interaction within organizations (Evans & 2009). HR practices are not only distinct, but also
This content downloaded from 152.118.24.10 on Wed, 07 Nov 2018 05:11:02 UTC
All use subject to https://about.jstor.org/terms
1278 Academy of Management Journal December
with the increase in financial outcomes. Specifi HR practices. For example, researchers have re
cally, we found that given no change in other con ported that leadership and organizational culture
ditions, a one standard deviation increase in skill have an important impact on employee motivation
enhancing, motivation-enhancing, or opportunity (Hartnell, Ou, & Kinicki, 2011; Hies, Nahrgang, &
enhancing HR practices was related to a .13, .18, or Morgeson, 2007). Therefore, managers may con
.09 standard deviation increase in financial out
sider how these factors can complement the effects
comes. For example, Huselid (1995) examined of
theHR practices in enhancing employee motivation.
This content downloaded from 152.118.24.10 on Wed, 07 Nov 2018 05:11:02 UTC
All use subject to https://about.jstor.org/terms
2012 Jiang, Lepak, Hu, and Baer 1279
Limitations and Future Research Fourth, in the current study we examined volun
Several limitations should be noted in the cur
tary turnover as an intermediate outcome mediat
ing the relationships between the three HR dimen
rent study. First, some studies included in this
sions as well as employee human capital and
meta-analysis used informant-reported measures to
motivation and financial outcomes. However, a
evaluate HR practices and organizational outcomes
growing literature indicates that voluntary turnover
from the same source. This may lead to common may moderate the relationship between HRM and
method bias, which might inflate the correlations
financial outcomes (e.g., Guthrie, 2001; Haus
between HR practices and organizational out
knecht & Trevor, 2011; Shaw, 2011; Shaw et al.,
comes. Relatedly, most of the studies included 2005).
in However, we were not able to test the inter
the analysis had cross-sectional designs, which
actions between the three HR dimensions and vol
may limit conclusions regarding the direction ofuntary turnover because very few studies reported
the mediating mechanism. The results from the the correlations between the interaction terms and
current investigation should be interpreted withthe variables examined. We encourage scholars to
these limitations in mind. We encourage more lonexplore this issue in future research. In addition,
gitudinal studies that collect information on HRrecent turnover research suggests that involuntary
practices and organizational outcomes from differ
turnover or dismissal is also influenced by HR
ent sources. Future meta-analysis can explore ifpractices
a and negatively related to operational and
financial
longitudinal research design may influence the es outcomes (Batt & Colvin, 2011; Haus
timates of effect sizes and the mediating mecha knecht & Trevor, 2011). It is worth considering the
nisms examined in this study. roles of both types of turnover in the mediating
Second, potential moderators may exist in the process rather than just focusing on voluntary
relationships among HR dimensions and organiza turnover.
tional outcome categories. For example, recentFifth, like other meta-analyses testing mediating
meta-analytic reviews have reported that industryprocess (e.g., Chang, Rosen, & Levy, 2009; Colquitt,
Scott, & LePine, 2007; Robbins, Oh, Le, & Button,
type (manufacturing industry vs. service industry)
2009), the current meta-analysis did not include
and country moderated the relationship between
control
HPWS and organizational outcomes (Combs et al., variables in the regression models (e.g., in
dustry,
2006; Rabl, Jayasinghe, Gerhart, & Kuehlmann, size, unionization, strategy) because many
2011; Subramony, 2009). Researchers have alsostudies did not provide correlations with these
variables.
suggested that HR practices applied to a specific
group of employees, or used for employees in genFinally, our study only focused on the relation
ships between HRM and organizational outcomes
eral, may influence their effects on organizational
at the organizational level, even though there is a
outcomes (Gerhart, Wright, McMahan, & Snell,
growing research focus on cross-level influences of
2000). However, owing to the relatively few studies
organization-level HRM on individual-level out
in the subgroups divided by the potential modera
comes (e.g., Liao et al., 2009; Snape & Redman,
tors, we were not able to test the mediating model
2010; Takeuchi et al., 2009) and on the influence of
separately in each subgroup. Future research can
employee-perceived HR systems on individual out
examine this mediating model by using samples
comes (e.g., Butts, Vandenberg, Dejoy, Schaffer, &
from different industries, different countries, and
Wilson, 2009; Kehoe & Wright, in press). We en
different job groups.
courage more empirical studies on the effects of
A third limitation of this study is that we were
organization-level HR systems and employee-per
unable to explore synergy among the three HR ceived di HR systems on individual outcomes. Over
mensions by examining their interactions, even time, there may be enough studies for a future
though the synergies within HR systems have been meta-analysis summarizing these effects on indi
suggested in the literature (e.g., Delery, 1998; Ger
vidual outcomes.
hart, 2007; Jiang et al., 2012). Operationally, this
was impossible owing to how existing studies were
Conclusions
measured. However, moving forward, if a good
amount of research includes all three HR dimen
This meta-analysis examined and extended the
sions while reporting the correlations of HR dimentheoretical model linking human resource manage
sions and organizational outcomes with interaction ment with organizational outcomes (e.g., Becker &
terms comprised of the three HR dimensions, fuHuselid, 1998; Delery & Shaw, 2001; Guest, 1997).
ture meta-analytic review will be able to exam We found that three dimensions of HR systems (i.e.,
ine this. skill-enhancing, motivation-enhancing, and oppor
This content downloaded from 152.118.24.10 on Wed, 07 Nov 2018 05:11:02 UTC
All use subject to https://about.jstor.org/terms
Academy of Management Journal December
This content downloaded from 152.118.24.10 on Wed, 07 Nov 2018 05:11:02 UTC
All use subject to https://about.jstor.org/terms
2012 Jiang, Lepak, Hu, and Baer 1281
This content downloaded from 152.118.24.10 on Wed, 07 Nov 2018 05:11:02 UTC
All use subject to https://about.jstor.org/terms
Academy of Management Journal December
sation, matter?
much do high-performance work practices and voice effects
A on quits and retention. In
dustrial Relations,
meta-analysis of their effects on organizational 39: 625-645.
per
formance. Personnel Psychology, 59: 501-528.
Delery, J. E., & Shaw, }. D. 2001. The strategic manage
Crook, T. R., Todd, S. Y., Combs, J. G., Woehr, D. J.,
ment of people & organizations: Review, syn
in work
Ketchen, D. J. 2011. Does human capital thesis, and extension. In
matter? A G. R. Ferris (Ed.), Research
meta-analysis of the relationship between
in personnel human
and human resource management,
capital and firm performance. Journal vol. 20:of Applied
167-197. Stamford, CT: JAI.
Psychology, 96: 443-456.
*Den Hartog, D. N., & Verbürg, R. M. 2004. High perfor
mance work
Curtis, B., Hefley, W. E., & Miller, S. 1995. Peoplesystems, organizational culture and
capa
firm effectiveness.
bility maturity model. Pittsburgh: Pittsburgh Soft Human Resource Management
Journal, 14:
ware Engineering Institute, Carnegie Mellon 55-78.
Univer
sity.
Dess, G. G., & Shaw, J. D. 2001. Voluntary turnover, social
Dalton, D. R., Daily, C. M., Certo, S. T., & Roengpitya, R. capital, and organizational performance. Academy
2003. Meta-analyses of financial performance and of Management Review, 26: 446-456.
equity: Fusion or confusion? Academy of Manage
ment Journal, 46: 13-26. Dyer, L., & Reeves, T. 1995. Human resource strategies
and firm performance: What do we know and where
Dalton, D. R., Daily, C. M., Ellstrand, A. E., & Johnson, do we need to go? International Journal of Human
J. L. 1998. Meta-analytic reviews of board composi Resource Management, 6: 656-670.
tion, leadership structure, and financial perfor
Erez, A., Bloom, M. C., & Wells, M. T. 1996. Using ran
mance. Strategic Management Journal, 19: 269
290. dom rather than fixed effects models in meta-analy
sis: Implications for situational specificity and valid
Dalton, D. R., Daily, C. M., Johnson, J. L., & Ellstrand, ity generalization. Personnel Psychology, 49: 275
A. E. 1999. Number of directors and financial per 306.
formance: A meta-analysis. Academy of Manage
*Ericksen, J. 2006. High-performance work systems, dy
ment Journal, 42: 674-686.
namic workforce alignment, and firm perfor
Dalton, D. R., & Todor, W. D. 1979. Turnover turned over: mance. Unpublished doctoral dissertation, Cornell
An expanded and positive perspective. Academy of University.
Management Review, 4: 225-236.
Evans, W. R., & Davis, W. D. 2005. High-performance
*Datta, D. K., Guthrie, J. P., & Wright, P. M. 2005. Human work systems and organizational performance: The
resource management and labor productivity: Does mediating role of internal social structure. Journal of
industry matter? Academy of Management Journal, Management, 31: 758-775.
48: 135-145.
*Faems, D., Sels, L„ De Winne, S., & Maes, J. 2005. The
*De Winne, S., & Sels, L. 2010. Interrelationships effectbe
of individual HR domains on financial perfor
tween human capital, HRM and innovation mance: in BelEvidence from Belgian small businesses. In
gian start-ups aiming at an innovation strategy. ternational
In Journal of Human Resource Manage
ternational Journal of Human Resource ment, 16: 676-700.
Management, 21: 1863-1883.
*Fey, C. F., & Björkman, I. 2001. The effect of human
Deci, E. L., Connell, J. P., & Ryan, R. M. 1989. Self resource management practices on MNC subsidiary
determination in a work organization. Journal of performance in Russia. Journal of International
Applied Psychology, 74: 580-590. Business Studies, 32: 59-75.
This content downloaded from 152.118.24.10 on Wed, 07 Nov 2018 05:11:02 UTC
All use subject to https://about.jstor.org/terms
2012 Jiang, Lepak, Hu, and Baer 1283
This content downloaded from 152.118.24.10 on Wed, 07 Nov 2018 05:11:02 UTC
All use subject to https://about.jstor.org/terms
Academy of Management Journal December
Kehoe,
Ilies, R., Wagner, D. T., & Morgeson, F. P. 2007. R. R., & Wright, P. M. In press. The impact of
Explain
high-performance human resource practices on em
ing affective linkages in teams: Individual differ
ployees' attitudes and behaviors. Journal of Man
ences in susceptibility to contagion and individual
ism-collectivism. Journal of Applied Psychology,
agement.
92: 1140-1148.
*Kepes, S., Delery, J., & Gupta, N. 2009. Contingencies in
*Iverson, R. D., & Zatzick, C. D. 2011. The effects of the effects of pay range on organizational effective
downsizing on labor productivity: The value of ness. Personnel Psychology, 62: 497-531.
showing consideration for employees' morale and
*Khatri, N. 2000. Managing human resource for compet
welfare in high-performance work systems. Human
itive advantage: A study of companies in Singapore.
Resource Management, 50: 29-44.
International Journal of Human Resource Manage
Jackson, S. E., & Schüler, R. S. 1995. Understanding ment, 11: 336-365.
human resource management in the context of or
*Kim, H., & Gong, Y. 2009. The roles of tacit knowledge
ganizations and their environments. In J. T. Spence,
and OCB in the relationship between group-based
J. M. Darley, & D. J. Foss (Eds.), Annual review of
pay and firm performance. Human Resource Man
psychology, vol. 46: 237-264. Palo Alto, CA: Annual
Reviews. agement Journal, 19: 120-139.
*Kintana, M. L., Alonso, A. U., & Olaverri, C. G. 2006.
Jackson, S. E., Schüler, R. S., & Rivero, J. 1989. Organi
High-performance work systems and firms' opera
zational characteristics as predictors of personnel
tional performance: The moderating role of technol
practices. Personnel Psychology, 42: 727-786.
ogy. International Journal of Human Resource
Jiang, K., Lepak, D. P., Han, K., Hong, Y., Kim, A., & Management, 17: 70-85.
Winkler, A. 2012. Clarifying the construct of human
*Kirkman, B. L., & Rosen, B. 1999. Beyond self-manage
resource systems: Relating human resource manage
ment: Antecedents and consequences of team em
ment to employee performance. Human Resource
powerment. Academy of Management Journal, 42:
Management Review, 22: 73-85. 58-74.
This content downloaded from 152.118.24.10 on Wed, 07 Nov 2018 05:11:02 UTC
All use subject to https://about.jstor.org/terms
2012 Jiang, Lepak, Hu, and Baer 1285
*Li,}. 2003. Strategic human resource management and *McClean, E., & Collins, C. J. 2011. High-commit
MNEs' performance in China. International Journal practices, employee effort, and firm perform
of Human Resource Management, 14: 157-173. Investigating the effects of HR practices acro
*Li, Y. 2003. The relationships between human re ployee groups within professional services
Human Resource Management, 50: 341-363
source management practices and perceptions of
organizational performance based on 1996-1997 *Miah, M. K., & Bird, A. 2007. The impact of cult
national organizations survey (NOS) data. Unpub HRM styles and firm performance: Evidenc
lished doctoral dissertation, the Pennsylvania State Japanese parents, Japanese subsidiaries/joi
University. tures and south Asian local companies. I
tional Journal of Human Resource Manage
*Liao, H., & Chuang, A. 2004. A multilevel investigation 18: 908-923.
of factors influencing employee service performance
and customer outcomes. Academy of Management *Minbaeva, D., Pedersen, T., Bjorkman, I., Fey, C
Journal, 47: 41-58. Park, H. J. 2003. MNC knowledge transfer, subsi
absorptive capacity, and HRM. Journal of In
*Liao, H., Toya, K., Lepak, D. P., & Hong, Y. 2009. Do tional Business Studies, 34: 586-599.
they see eye to eye? Management and employee per
spectives of high-performance work systems and in Morrow, P., & McElroy, J. 2007. Efficiency as a me
fluence processes on service quality. Journal of Ap in turnover-organizational performance relat
plied Psychology, 94: 371-391. Human Relations, 60: 827-849.
This content downloaded from 152.118.24.10 on Wed, 07 Nov 2018 05:11:02 UTC
All use subject to https://about.jstor.org/terms
Academy of Management Journal December
level: Human resource practices and training and support. Personnel Psychology, 38:
organizational
effectiveness. In K. J. Klein & S. W.849-863.
J. Kozlowski
(Eds.), Multilevel theory, research, and methods in
Ryan, R. M., & Deci, E. L. 2000. Self-determinat
organizations: 211-266. San Francisco: Jossey-Bass.
and the facilitation of intrinsic motivation, social
development, andand
Overton, R. C. 1998. A comparison of fixed-effects well-being. American Psycholo
gist, 55: 68-78. test
mixed (random-effects) models for meta-analysis
of moderator variable effects. Psychological Meth
Schneider, B. 1978. Person-situation selection: A review
ods, 3: 354-379.
of some ability-situation interaction research. Per
*Park, H. J., Mitsuhashi, H., Fey, C. F., & Björkman, I. sonnel Psychology, 31: 281-297.
2003. The effect of human resource management
Shaw, J. D. 2011. Turnover rates and organizational per
practices on Japanese MNC subsidiary performance:
formance: Review, critique, and research agenda.
A partial mediating model. International Journal of
Organizational Psychology Review, 1: 187-213.
Human Resource Management, 14: 1391-1406.
*Shaw, J. D., Delery, }. E., Jenkins, G. D., Jr., & Gupta, N.
*Patterson, M. G., West, M. A., & Wall, T. D. 2004. Inte
1998. An organization-level analysis of voluntary
grated manufacturing, empowerment, and company
and involuntary turnover. Academy of Manage
performance. Journal of Organizational Behavior, ment Journal, 41: 511-525.
25: 641-665.
Rabl, T., Jayasinghe, M. M., Gerhart, B., & Kuehlmann, *Singh, K. 2004. Impact of HR practices on per
T. M. 2011. How much does country matter? A firm performance in India. Asia Pacific Jour
meta-analysis of the HPWP systems-business per Human Resources, 42: 301-317.
formance relationship. Paper presented at the an
nual meeting of the Academy of Management, San *Skaggs, B. C., & Youndt, M. 2004. Strategic positioning,
Antonio. human capital, and performance in service organi
zations: A customer interaction approach. Strategic
Richard, O. C., & Johnson, N. B. 2004. High performance Management Journal, 25: 85-99.
work practices and human resource management ef
fectiveness: Substitutes or complements? Journal ofSnape, E., & Redman, T. 2010. HRM practices, organiza
Business Strategy, 21: 133-148. tional citizenship behaviour and performance: A
multi-level analysis. Journal of Management Stud
Robbins, S. B., Oh, I. S., Le, H., & Button, C. 2009. Inter ies, 47: 1219-1247.
vention effects on college performance and retention
Snell, S. A., & Dean, J. W. 1992. Integrated manufacturing
as mediated by motivational, emotional, and social
and human resource management: A human capital
control factors: Integrated meta-analytic path analy
perspective. Academy of Management Journal, 35:
ses. Journal of Applied Psychology, 94: 1163-1184. 467-504.
This content downloaded from 152.118.24.10 on Wed, 07 Nov 2018 05:11:02 UTC
All use subject to https://about.jstor.org/terms
2012 Jiang, Lepak, Hu, and Baer 1287
Sobel, M.
Asymptotic E. 1982.
lationship: Does investment in employees
confide pay off?
indirect Academy of Management Journal,
structural effects 40: 1089-1121.
in
equation
Leinhard (Ed.), Sociological methodol
*Tzafrir, S. S. 2005a. A universalistic perspective for
San Francisco: Jossey-Bass.
explaining the relationship between HRM practices
*Stavrou, E. T. 2005. Flexible
and firm work
performance at different points in time. bu
zational competitiveness:
Journal of ManagerialA cross-nat
Psychology, 21: 109-130.
the European work context. Journa
*Tzafrir,923-947.
tional Behavior, 26: S. S. 2005b. The relationship between trust,
HRM practices and firm performance. International
*Steingruber, W. G. 1996.
Journal Strategic
of Human Resource Management, 16: i
man resource management:
1600-1622. An analy
lationship between international str
*Veld, M., Paauwe, J., & Boselie, P. 2010. HR
tioning and the degree of integrate
strategic climates in hospitals: Does
human resource management. Unp the m
come across at the ward level? Human Resource
toral dissertation, University of North
Management Journal, 20: 339-356.
Strober, M. H. 1990. Human capital theor
for HR managers. Viswesvaran,
Industrial C., & Ones, D.Relations,
S. 1995. Theory testing:
Combining psychometric meta-analysis and struc
*Stup, R. E. 2006. Human resource
tural equations modeling. ma
Personnel Psychology,
izational commitment,
48: 865-885. and percei
tional support in dairy farm busine
lished doctoral * Vlachos, I. 2008. The effect
dissertation, the of human resource
Pennsy
University. on organizational performance: Evidenc
Greece. International Journal of Human R
Subramaniam, M., & Youndt, M. A. 2005. The influence Management, 19: 74-97.
of intellectual capital on the types of innovative
capabilities. Academy of Management Journal, 48: *Way, S. A. 2002. High performance work sys
450-463. intermediate indicators of firm performanc
the US small business sector. Journal of
Subramony, M. 2009. A meta-analytic investigation of
ment, 28: 765-785.
the relationship between HRM bundles and firm per
formance. Human Resource Management, *Welbourne,
48: 745 T. M., & Andrews, A. O. 1996. Predicting
768. the performance of initial public offerings: Should
human resource management be in the equation?
*Subramony, M., Krause, N., Norton, J., & Burns, G. N. Academy of Management Journal, 39: 891-919.
2008. The relationship between human resource in
vestments and organizational performance: A firm *White, J. E. 1998. Implementation and effect of high
level examination of equilibrium theory. Journal of performance work practices in nuclear power
Applied Psychology, 93: 778-788. plants. Unpublished doctoral dissertation, the Penn
sylvania State University.
*Sun, L. Y., Aryee, S., & Law, K. S. 2007. High perfor
mance human resource practices, citizenship behavWhite, R. E., & Hamermesh, R. C. 1981. Toward a model
ior, and organizational performance: A relational of business unit performance: An integrative ap
perspective. Academy of Management Journal, 50: proach. Academy of Management Review, 6: 213
558-577. 223.
*Whitener,
Takeuchi, R., Chen, G., & Lepak, D. P. 2009. Through the E. M. 2001. Do "high commitment" human
resource
looking glass of a social system: Cross-level effects of practices affect employee commitment? A
high performance work systems on employees' cross-level
atti analysis using hierarchical linear model
tudes. Personnel Psychology, 62: 1-29. ing. Journal of Management, 27: 515-535.
*Wood, S.,K.
*Takeuchi, R., Lepak, D. P., Wang, H., & Takeuchi, Holman, D., & Stride, C. 2006. Human re
source management and performance in UK call cen
2007. An empirical examination of the mechanisms
tres. Rritish Journal of Industrial Relations, 44: 99
mediating between high-performance work systems
124.
and the performance of Japanese organizations. Jour
nal of Applied Psychology, 92: 1069-1083.
Wright, P. M., & Boswell, W. R. 2002. Desegregating
Tharenou, P., Saks, A., & Moore, C. 2007. A review and
HRM: A review and synthesis of micro and macro
human resource management research. Journal of
critique of research on training and organizational
level outcomes. Human Resource Management Management,
Re 28: 247-276.
view, 17: 251-273.
Wright, P. M., Dunford, B. B., & Snell, S. A. 2001. Human
Tsui, A„ Pearce, J., Porter, L., & Tripoli, A. 1997. Alter resources and the resource based view of the firm.
native approaches to the employee-organization re Journal of Management, 27: 701-721.
This content downloaded from 152.118.24.10 on Wed, 07 Nov 2018 05:11:02 UTC
All use subject to https://about.jstor.org/terms
Academy of Management Journal December
This content downloaded from 152.118.24.10 on Wed, 07 Nov 2018 05:11:02 UTC
All use subject to https://about.jstor.org/terms
2012 Jiang, Lepak, Hu, and Baer 1289
APPENDIX
Coding of Studies Included in the Meta-analysis
Skill Motivation
Enhancing Human Open ational Financial
Study HR Practices Enhancing JrfXSBs
HR Practices HR Practices Capital
Employee
Motivation 6ut< Outcomes
opportunities,
profit
sharing
Training Team Labor
participation productivity
Voluntary Productivity,
turnover innovation
Voluntary Productivity
turnover
HR incentive Work design Job skill level Quit rate Percent change
index index
Training Variable pay, Problem Quit rate
pay to cost solving
of living
?e?r
directed
development
Performance Participation
training appraisal,
compensation
Staffing, performance, Involvement Customer Helping behavior Service Market
training compensation, knowledge performance performance
caring
Climate for trust, Sale growth,
cooperation revenue
Continued
This content downloaded from 152.118.24.10 on Wed, 07 Nov 2018 05:11:02 UTC
All use subject to https://about.jstor.org/terms
1290
Academy of Management Journal December
APPENDIX
(Continued)
Skill Motivation Opportunity
Enhancing Enhancing Enhancing Human
Operational Financial
Study HR Practices HR Practices HR Practices Capital Outcomes Outcomes
Datta,
Productivity Sales growth
Guthrie, and
Wright (2005)
De Winne and Selection, Group-based Participation Innovation
Sels (2010) training appraisal
ana
performance
Delanev and Staffing Incentive Grievance Perceived
Huselia selectivity, compensation, procedure, market
(1996) training internal decentralized performance
labor decision
market making
Delery and Training Appraisals, Participation Return on
Doty (1996) job assets, return
security, on equity
career
opportunities,
profit
sharing
Delery, Gupta, 'ay ar Voice Quit rate
Shaw, Dem mechanisms
Jenkins, and
Ganster (2000)
Den Hartog Employee Pay-for Autonomy, Voluntary Overall Economic
and Verbürg skills and performance, information turnover operational outcome
(2004) direction profit sharing performance
sharing, meetings
profit
sharing,
performance
evaluation
Ericksen Workforce Voluntary Sales growth
(2006) alignment turnover
Faems, Sels, Selection, Career Participation Voluntary Productivity Value added
De Winne, training management, turnover
and Maes compensation,
(2005) performance
management
Fey and Training and Pay and Information Overall
Björkman development performance sharing and financial
(2001) appraisal complaint performance
resolution
Fey, Training Performance Decentralization, Overall
Björkman, complaint financial
and compensation, resolution performance
Pavlovskaya job
(2000) security,
career
planning,
salary level
Gardner, Skill HR Motivation Education Affective Voluntary
Wright, and practices HR Empowerment level commitment turnover
Moynihan, practices practices
(2011)
Gelade and Staffing, Job c General climate Staff retention Customer Overall
Ivery (2003) professional satisfaction, financial
development clerical performance
accuracy
Gerhart and Education, Return on
Pay and
Milkovich incentive experience assets, sale
(1990)
Ghebregiorgis Recruitment, Compensation Voluntary Productivity
and Karsten selection, turnover
(2007) training,
development
Gibson, Team, Customer Overall
Porath, information service, financial
Benson, and sharing, quality performance
Lawler (2007) boundary
setting
Selective Pay Participation Collective affective
hiring, contingent in decision commitment,
extensive on making collective organ
training performance, izational
career
citizenship
planning, behavior
performance
appraisal
Selective Employment Participation Affective Overall
hiring, security, in decision commitment financial
extensive pay making performance
training contingent
on
performance,
career
development,
performance
appraisal
Continued
This content downloaded from 152.118.24.10 on Wed, 07 Nov 2018 05:11:02 UTC
All use subject to https://about.jstor.org/terms
2012 Jiang, Lepak, Hu, and Baer
APPENDIX
(Continued)
Skill Motivation- Opportunity
Enhancing Enhancing Enhancing Human Employee Operational Financial
Study HR Practices HR Practices HR Practices Capital Motivation Outcomes Outcomes
Continued
This content downloaded from 152.118.24.10 on Wed, 07 Nov 2018 05:11:02 UTC
All use subject to https://about.jstor.org/terms
1292
Academy of Management Journal December
APPENDIX
(Continued)
Skill Motivation Opportunity
Enhancing Enhancing Enhancing Human
Employee Operational Financial
Study HR Practices HR Practices HR Practices Capital Motivation Outcomes Outcomes
Continued
This content downloaded from 152.118.24.10 on Wed, 07 Nov 2018 05:11:02 UTC
All use subject to https://about.jstor.org/terms
2012 Jiang, Lepak, Hu, and Baer
APPENDIX
(Continued)
Skill Motivation Opportunity
Enhancing Enhancing Enhancing Human Financial
Employee Operational
Study HR Practices HR Practices Capital Motivation Outcomes Outcomes
planning
Skaggs and Human Return on
Youndt (2004) capital equity, return
on
investment
Snell and Performance Return on
Youndt (1995) training appraisal,
and performance growth
development reward
Stavrou Job c Voluntary
(2005) turnover
Steingruber Training
(1996)
Stup (2006) Training, Performance Written job Organizational
selection review, descriptions, commitment
incentives, communication,
benefits participation
Subramony, Compensation Employee morale Productivity,
Krause, customer
Norton, and satisfaction
Burns (2008)
Sun. Aryee, Organizational Voluntary Productivity
and Law
citizenship turnover
(2007) behavior
Takeuchi, Human Social exchange
Lepak, Wang, capital relationship
ana Takeucni
(2007)
Tzafrir Selection, Incentive Overall Market
Employee
(2005a) training compensation, participation operational performance
internal
performance
labor
market
Tzafrir Training Evaluation, Participation Overall Market
(2005b) compensation, operational performance
internal performance
labor
market
Veld, Paauwe, Performance Communication, Education Commitment
and Boselie management level
autonomy,
(2010) information
sharing
Vlachos Selective Compensation, Decentralization, Product Market share,
(2008) hiring, job security information quality
training sharing
and
development
Extensiveness Group-based Job rotation, Labor
Voluntary
of staffing, performance self turnover productivity
formal directed
pay, pay
training level teams,
involvement
Continued
This content downloaded from 152.118.24.10 on Wed, 07 Nov 2018 05:11:02 UTC
All use subject to https://about.jstor.org/terms
1294 Academy of Management Journal December
APPENDIX
(Continued)
Skill- Motivation- Opportunity
Enhancing Enhancing Enhancing Human Employee Operational Financial
Study HR Practices HR Practice« HR Practices Capital Motivation Turnover Outcomes Outcomes
This content downloaded from 152.118.24.10 on Wed, 07 Nov 2018 05:11:02 UTC
All use subject to https://about.jstor.org/terms