Human Resources

Human Resource Management, May–June 2016, Vol. 55, No. 3. Pp. 413–435

© 2014 Wiley Periodicals, Inc.

Published online in Wiley Online Library (wileyonlinelibrary.com).

DOI:10.1002/hrm.21672

Correspondence to: Jeffrey B. Arthur, Associate Professor, Virginia Tech University, Department of Management (0233), 880 West Campus Drive, Blacksburg, Virginia 24061, Phone: 540-231-5695, E-mail: [email protected]

HOW TOP MANAGEMENT HR

BELIEFS AND VALUES AFFECT

HIGH-PERFORMANCE WORK

SYSTEM ADOPTION AND

IMPLEMENTATION EFFECTIVENESS

J E F F R E Y B . A R T H U R , A N D R E W O . H E R D M A N , A N D   J A E W A N YA N G

What explains why high-performance work systems (HPWSs) are not adopted more widely by fi rms that would appear to benefi t economically by adopting them? We address this question by drawing on the upper-echelons perspective to consider the role of the top managers’ beliefs concerning the fi nancial payoffs from investments in HR (“HR cause-effect belief”) as well as their employee- centered value-based beliefs (labeled “HPWS values”). We propose a concep- tual model in which top management HPWS values moderate the relationship between HR cause-effect beliefs and the intensity of HPWS programs reported by managers as well as the relationship between HPWS programs and employees’ perceptions of implemented HPWS practices. We test our model using a unique multisource data set collected from 120 hotel franchisees that includes survey responses from 648 managers and 1,293 employees. We fi nd that fi rms’ ability to translate top managers’ cause-effect beliefs about the economic value of HR investments into adoption of HPWS programs, as well as their ability to effec- tively implement these HPWS programs, is signifi cantly affected by the value- based HR beliefs held by top managers. © 2014 Wiley Periodicals, Inc.

Keywords: strategic HR, top management teams, strategic decision making

Introduction

T o date, an impressive body of empiri- cal research has demonstrated a positive relationship between the extent to which firms report using high-performance work system (HPWS) programs and vari-

ous firm-level performance outcomes (Combs, Yongmei, Hall, & Ketchen, 2006; Huselid, 1995;

Subramony, 2009). For example, in his highly influential study, Huselid (1995) found that a one standard deviation change in the HPWS index score was associated with substantially higher sales, market value, and profits. These empirical findings have raised a number of issues among strategic HRM researchers including the mea- surement of HPWS programs (Gerhart, Wright,

414 HUMAN RESOURCE MANAGEMENT, MAY–JUNE 2016

Human Resource Management DOI: 10.1002/hrm

If HPWSs are as

universally effective

in contributing

to improved

performance as they

appear, then why

haven’t we seen a

broader diffusion of

these HR systems

across firms? In

other words, why

do some firms adopt

HPWS while other

(seemingly similar)

firms don’t?

future research” (Becker & Huselid, 2006, p. 905). In a recent critique of strategic HRM research, Kaufman (2010, p. 297) concludes that the inabil- ity of strategic HR researchers to fully explain the lack of HPWS diffusion “casts considerable doubt on the intellectual robustness and explanatory power of the mainstream SHRM model.” We pro- pose to contribute to the strategic HR literature by addressing this important gap in the research.

In this study, we develop a conceptual model that draws from “upper-echelons theory” found in the strategic management literature (Hambrick & Mason, 1984). Upper-echelons theory empha- sizes that the strategic context faced by top man- agers can most often be characterized as involving significant ambiguity and uncertainty (March & Simon, 1958). Under these conditions, the cogni- tive structures (i.e., “cause-effect beliefs”) and per- sonal values of top management decision makers provide critical insights for understanding how and why firms look and behave the way they do (Finkelstein, Hambrick, & Cannella, 2009). Applying upper-echelons theory to HPWS adop- tion, we propose that top managers will vary in their implicit and explicit cause-effect beliefs about whether investments in HR programs will lead to improved organizational performance. These beliefs are expected to have a direct effect on the level of HR program investments found in firms. In addition, the extent to which orga- nizational leaders hold employee-centered values (in which employees are viewed as assets to be developed as opposed to costs to be minimized) is expected to significantly influence the intensity of HPWS programs adopted by the firm.

Going further, we expect that top manage- ment values will also significantly affect the extent to which “intended” HPWS programs are actually implemented and experienced by employees. Bowen and Ostroff (2004) labeled the extent to which employees experience a firm’s HR system in a uniform manner, consistent with the intent of organizational leaders, as the “HR system strength.” They argued that HR system strength is affected by the distinctiveness, consistency, and consensus with which the HR system is imple- mented. We add to this explanation by propos- ing that top management’s employee-centered values can affect HPWS strength and thus have a significant moderating effect on the relationship between the use of HPWS programs reported by managers and employees’ experience and percep- tions of these programs.

We test our proposed model using a unique data set that consists of detailed, intra-indus- try, firm-level, and multisource survey data col- lected from a sample of 648 managers and 1,293

McMahan, & Snell, 2000), the order of causal- ity between HPWS programs and performance outcomes (Wright, Gardner, Moynihan, & Allen, 2005), and mechanisms that explain the rela- tionship between HPWS programs and firm-per- formance outcomes (Takeuchi, Lepak, Wang, & Takeuchi, 2007). Although theories and evidence addressing these issues continue to be developed, another basic question in this literature has remained largely unaddressed: If HPWSs are as universally effective in contributing to improved performance as they appear, then why haven’t we seen a broader diffusion of these HR systems across firms? In other words, why do some firms adopt HPWS while other (seemingly similar) firms don’t?

We are not the first to consider the theoretical and practical impor- tance of addressing the gap between evidence of economic success of HPWSs and yet a lack of diffusion across a broader number of firms (Kaufman & Miller, 2011; Pfeffer, 1998; Wright, Dunford, & Snell, 2001). Several authors have proposed various forms of “market failure” to explain the gap. In one early exam- ple, Pfeffer (1998, p. 29) proposed a “one-eighth rule.” He speculated that about one half of organizations wouldn’t invest in HPWSs because they don’t believe that there is a con- nection between investments in HR and financial performance. Of the remaining one half, only 50 percent would engage in the kind of sys- tematic and comprehensive invest- ments in HPWS required for success. Finally, only one half of these firms are expected to “persist with prac- tices long enough to actually derive economic benefits” (Pfeffer & Viega, 1999, p. 47).

Nearly 10 years later, Becker and Huselid (2006) noted that results showing persis- tent large effects of HPWSs on firm performance may be related to heterogeneity in “managerial quality” across firms (Chadwick & Dabu, 2009). They suggested that differences in the ability of managers to successfully implement HPWSs may help explain why the market for HPWSs has failed to reach equilibrium (and thus eliminate competi- tive advantage derived from the use of these sys- tems): “Whether this market failure is due to lack of knowledge, a lack of managerial competence, or an inability to execute (or more likely some com- bination) is open to conjecture and, it is hoped,

Human Resource Management DOI: 10.1002/hrm

TOP MANAGEMENT HR VALUES 415

Cognitive base

refers to beliefs

about factors that

executives consider

to be strategically

important as well

as beliefs about the

causal relationships

between these

factors and firm

performance.

information that enters an executive’s field of vision, and how that information is perceived and interpreted. In considering the role of mana- gerial values, researchers have generally adopted Rokeach’s (1973, p. 5) conception of a value as an “enduring belief that a specific mode of conduct or end-state of existence is personally or socially preferable to an opposite or converse mode of conduct or end-state of existence.” Values are distinguished from other types of beliefs in that they have an “oughtness” characteristic to them that lead managers to behave in ways that they believe to be socially desirable or beneficial (Hitlin & Piliavin, 2004; Rokeach, 1973; Schwartz, 1992). Values are also seen as relatively stable aspects of a manager’s orientation developed in part through experience and not subject to frequent change (Meglino & Ravlin, 1998; Rokeach, 1973). Both the values and cause-effect beliefs of top managers are expected to affect strategic decision making by affecting what information makes it into an executive’s field of vision, as well as how that information is per- ceived, interpreted, and acted upon (Thomas, Clark, & Gioa, 1993).

Top Management HPWS Cause- Effect and Value-Based Beliefs

Next, we turn to the question of the specific cause-effect and value- based beliefs that are expected to be positively associated with strategic decisions by top managers to invest in an HPWS. Regarding cause-effect beliefs, a basic premise of an HPWS is that employees are viewed as assets to be developed as opposed to simply variable costs to be mini- mized (Bamberger & Meshoulam, 2000, p. 42). As part of viewing employees as assets, we propose that an underlying managerial cause-effect belief in a profit-making enterprise is that there will be some positive financial return to the firm for investments made in human resources and HR programs. Although the specific type of HR pro- gram may vary (e.g., selection, recruitment, train- ing, performance management), a belief by top managers that investing in HR will yield positive returns for the company would appear to a basic condition underlying the use of HPWSs. We label this top management belief as “HR cause-effect belief.” As noted previously, Pfeffer (1998) pro- posed that absence of this basic belief regarding the link between HR and profits in “about half” of firms represented an important factor limiting the more widespread adoption of HPWS programs.

employees in 120 hotel franchises. In contrast to data used in much of the upper-echelons empirical research, our survey included direct measures of the cause-effect and value-based beliefs collected from the group of top managers at each location. In addition, data on the firm’s HPWS programs were collected using a separate survey completed by the HR or general manager at each location. Finally, data on perceived HPWS practices were collected from the hotel’s front desk and house- keeping employees. These data and sample char- acteristics allow us to contribute to the strategic HR literature by assessing not only the effects of top management beliefs on firm-level variation in HPWS programs as reported by managers, but also the relationship between these stated HPWS programs and employees’ perceptions of imple- mented HWPS practices within each firm.

Literature Review

In applying upper-echelons theory to understand- ing interfirm HPWS variation, we begin by con- sidering the types of managerial beliefs that are predicted to be associated with firm-level strate- gic choices and outcomes generally. Hambrick and Mason (1984) identified the “cognitive base” and “values” of executives as major psychologi- cal characteristics affecting their strategic choices. Cognitive base refers to beliefs about factors that executives consider to be strategically impor- tant as well as beliefs about the causal relation- ships between these factors and firm performance (Chattopadhyay, Glick, Miller, & Huber, 1999; Finkelstein et al., 2009; Markoczy, 1997). The con- tent of what a leader knows (through experience or education) is expected to affect what informa- tion he/she notices and how he/she interprets that information. For example, a leader with training and experience in engineering is expected to be more likely to seek out and notice information regarding technical specifications of products and to evaluate the importance of that informa- tion differently than someone with a background in sales or marketing. Cognitive structures refer to beliefs about how strategic factors are caus- ally related to each other and to performance (Chattopadhyay et al., 1999). These beliefs repre- sent implicit or explicit mental models that take the form of causal inferences such as “if X, then Y.” Following previous work, we label beliefs asso- ciated with cognitive structure as top managers’ “cause-effect beliefs” (Chattopadhyay et al., 1999; Markoczy, 1997).

Beyond cognitive structures, Hambrick and Mason (1984) also proposed that the personal val- ues of executives would affect strategic decision making by affecting (or filtering) the external

416 HUMAN RESOURCE MANAGEMENT, MAY–JUNE 2016

Human Resource Management DOI: 10.1002/hrm

Although the

existence of HR

cause-effect beliefs

by top managers

is considered a

necessary condition

for the adoption of

HPWSs, it is not

sufficient. In addition,

we propose that

top management

employee-centered

values supporting

the use of HPWSs

represent an

additional necessary

condition that must

be present for firms to

fully adopt HPWSs.

nature of mutual obligations is uncertain and ambiguous, social exchange relationships require a degree of trust between exchange partners.

To date, empirical research linking the values or philosophy of top managers and the use of HR programs has been quite limited. We were able to locate two relevant studies. First, Osterman (1994) used data from a 1992 survey of 694 US manufacturing establishments to examine the extent to which various factors, including man- agement values, might explain variation in the adoption of “innovative work practices” includ- ing quality circles, Total Quality Management (TQM), job rotation, and autonomous teams. The values measure used in the study was a sin- gle dichotomously scored item that asked man- agement respondents whether they agreed that helping to “increase the well-being of employees with respect to their personal or family situation” was considered appropriate in their establish- ment (Osterman, 1994, p. 179). Osterman (1994) found that this “value” variable was positive and significantly related to the use of innovative work practices in all of his regression models. He concludes, “Evidently, independent of any productivity gains to be had from flexible work organization, establishments that believe that they have responsibility for employee welfare are more likely to adopt innovative work practices” (p. 182).

More recently, Lepak et al. (2007) investi- gated the extent to which “high-investment work practices” applied to core and support employees were contingent on industry, strat- egy, and an “employee-centered HR philosophy.” The employee-centered HR philosophy variable reflected managers’ “general beliefs of high regard for employee welfare and treatment” (Lepak et al., 2007, p. 229). They found strong support for the hypothesis that HR philosophy predicted variation in the use of HR practices beyond that explained by industry and firm strategy. Although these pre- vious studies are useful, they are limited in that the data included only one or two respondents per organization for measures of HR programs and HR values. In addition, previous empirical studies did not include measures of other top management characteristics such as education or tenure in the firm that might provide alternate explanations for the relationship found between top managers’ values and HR programs (Finkelstein et al., 2009). Finally, previous empirical strategic HR research on values has not used a comprehensive theoreti- cal model such as the upper-echelons perspective that explains how managerial value-based beliefs are expected to be related to the choice of HPWS programs.

We are not aware, however, of any existing empir- ical test of this proposal.

Although the existence of HR cause-effect beliefs by top managers is considered a necessary condition for the adoption of HPWSs, it is not suf- ficient. In addition, we propose that top manage- ment employee-centered values supporting the use of HPWSs represent an additional necessary condition that must be present for firms to fully adopt HPWSs. The importance of top manage- ment value-based beliefs, or what has sometimes

been labeled “management phi- losophy” (Lepak, Taylor, Tekleab, Marrone, & Cohen, 2007), was iden- tified in early conceptual strategic HR research (Dyer & Holder, 1988; Kochan, Katz, & McKersie, 1986; Schuler, 1992; Walton, 1985). The discussion of management values reemerged in the discussion of the “architectural” perspective (Becker & Gerhart, 1996; Lepak & Snell, 1999). According to this perspec- tive, the defining characteristic of an HR system (including HPWS) is not the specific HR practices that are used, but rather the extent to which HR practices are consistent with an underlying philosophy or logic for managing people in organizations (Becker & Gerhart, 1996; Lepak & Snell, 1999).

Like all employment systems, HPWSs represent an exchange rela- tionship between employees and employers (Tsui, Pearce, Porter, & Tripoli, 1997). Unlike HR systems based on short-term and specific contractual terms (i.e., economic exchange), HPWSs are based on the underlying logic of social exchange and employee-management trust (Bamberger & Meshoulam, 2000; Delery & Doty, 1996; Tsui et al., 1997). According to social exchange

theory (Blau, 1964), repeated interactions between interdependent exchange partners generate recip- rocal social obligations in which parties experience strong obligations to return in kind benefits they receive from their exchange partner (Gouldner, 1960). In addition to extrinsic rewards, resources included in social exchanges involve intrinsic aspects of social relationships such as personal recognition and social support. Also, in contrast to purely economic exchange relationships, social exchange relationships involve more long-term and open-ended obligations. Because the precise

Human Resource Management DOI: 10.1002/hrm

TOP MANAGEMENT HR VALUES 417

programs and the extent to which programs apply to all employees) of HPWS programs, and (2) this variation in HPWS intensity is systematically related to differences in firm performance across a large number of studies. One way to understand the importance of stated HPWS programs is to view them as providing important information about top managers’ cause-effect and value-based beliefs (cf. Bowen & Ostroff, 2004). Thus, the intensity of the HPWS programs reported by man- agers can be seen as representing an important mechanism for communicating values of organi- zational leaders to employees as well as to lower- level managers charged with implementing the HPWS programs (Bowen & Ostoff, 2004; O’Reilly & Pfeffer, 2000; White & Bryson, 2013). Put dif- ferently, a consistent and coherent set of HPWS programs may improve the possibility that these programs will be implemented in a more consis- tent and coherent way by lower-level managers. As shown in Figure 1, we consider the intensity of HPWS programs reported by managers to repre- sent an important mediating mechanism linking top management cause-effect beliefs and employ- ees’ HPWS perceptions.

HR Cause-Effect Belief and HPWS Programs

Regarding the first link in the mediated relation- ship shown in Figure 1, we propose that there will be a positive relationship between the level of HR cause-effect beliefs held by top managers and the intensity of HPWS adopted by the firm. When considering the relationship between HR cause- effect beliefs and HPWS, it is especially important to take into account the level or strength of HR cause-effect beliefs held by organizational lead- ers. In general, we would expect most managers to indicate that they believe there is some posi- tive financial return to investing in HR programs. We also expect, however, that managers will dif- fer in the extent to which they believe the return on investment for HR programs is higher or lower than potential returns on investment for other strategic assets. For example, a manager holding low-level HR cause-effect beliefs may consider investments in HR to be somewhat important, but not nearly so cost effective for improving firm performance as investments in new technology,

Hypotheses

We present our conceptual model for the effects of top managers’ cause-effect and value-based beliefs on firms’ adoption of HPWS in Figure 1.

As shown in Figure 1, the dependent variable in our model is employee perceptions of imple- mented HPWS practices (or simply “HPWS percep- tions”). Although the majority of previous strategic HR research use measures of HPWS policies and programs reported by HR managers as indicators of the extent of HPWS in a firm, we follow more recent strategic HR research that makes important distinctions between stated (or intended) HPWS policies and programs and actual HR practices experienced by employees (Liao, Toya, Lepak, & Hong, 2009; Nishii, Lepak, & Schneider, 2008; Peccei, Van de Voorde, & Van Veldhoven, 2012). According to Wright and Boswell (2002, p. 264):

HR policies represent the fi rm or busi- ness unit’s stated intentions about the kinds of HR programs, processes, and techniques that should be carried out in the organization. HR practices consist of actual programs, processes, and tech- niques that actually get operationalized in the unit. … Because employees can only respond to actual practices, any research attempting to demonstrate a relationship between HRM and fi rm per- formance stands on fi rmer ground when assessing the actual practices rather than intended policies.

Because strategic HR research is ultimately concerned with predicting how the use of HR sys- tems impacts firm-level performance outcomes, firm-level variation in employees’ perceptions of HPWS practices provides the most relevant depen- dent variable for this study.

This is not to say, however, that a firm’s stated HPWS programs are unimportant. As previously discussed, the majority of previous empirical strategic HR research has used HPWS programs reported by a firm’s top HR manager as the inde- pendent variable in predicting variation in firm performance (Combs et al., 2006; Subramony, 2009). This research has consistently shown that (1) firms vary in the intensity (i.e., number of HR

FIGURE 1. Conceptual Model

Level of Top-Mgt HR Cause-Effect Beliefs

Intensity HPWS Programs

Perceived HPWS Practices

Top-management HPWS Values

H1

H2 H3

418 HUMAN RESOURCE MANAGEMENT, MAY–JUNE 2016

Human Resource Management DOI: 10.1002/hrm

across a range of employee-focused HR programs, we propose that high levels of systematic HPWS program investment are more likely to be found in firms with both high HR cause-effect belief and high HPWS value-based beliefs among top managers.

Hypothesis 2: The level of top management HPWS values will moderate the relationship between the strength of HR cause-effect beliefs and HPWS intensity. Specifi cally, the relationship between HR cause-effect beliefs and HPWSs will be more strongly positive under conditions of high top management HPWS values.

HPWS Programs and Employee Perceptions: The Moderating Role of HPWS Values

As shown in Figure 1, we also expect that top management HWPS values will moderate the relationship between HPWS programs and HPWS perceptions. To understand this proposed moder- ating effect, we need to consider how top manage- ment HPWS value-based beliefs affect the quality of HPWS program implementation by lower-level managers. As Nishii and Wright (2008) point out, the relationship between an organization’s stated HR policies and actual HR practices depends on the quality of HR program implementation. Because employees are the recipients of imple- mented HR programs, they are in the best posi- tion to report on the extent and quality of HPWS program implementation (Arthur & Boyles, 2007; Wright & Boswell, 2002).

In line with Bowen and Ostroff’s (2004) dis- cussion of HR programs as communication mech- anisms, formal HPWS programs can be seen as sending important messages to employees and lower-level managers about the stated values held by the organization. In implementing these pro- grams, however, we expect that lower-level man- agers’ perceptions and interpretations of these programs will be affected by both the stated val- ues and the actions of top managers. We posit that top managers whose personal value-based beliefs are consistent with the values underlying HPWS programs are more likely to be perceived by lower- level managers as “walking the talk” (Simons, 2002)— that is, acting in a way that is consistent with stated HPWS values. Further, we expect that HPWS implementation effectiveness will improve when there is a correspondence between the level of HPWS programs and HPWS values in part because of improvements in HPWS distinctiveness that affect the perceived strength of the HR system by lower-level managers and employees (Bowen & Ostroff, 2004).

financial instruments, or marketing programs. The more managers believe that investments in HR programs will lead to positive financial returns relative to other potential strategic investments, the stronger their HR cause-effect beliefs.

Hypothesis 1: The level of top management HR cause- effect beliefs will be positively related to the intensity of HPWS programs reported by the fi rm.

As previously discussed, we posit that top management HR cause-effect beliefs alone are unlikely to lead firms to fully adopt HPWS. Instead, we propose that the degree to which top managers’ value-based beliefs are consistent with the use of HPWS programs will interact with top management HR cause-effect belief to affect the intensity of HPWS adopted by the firm. We define top management HPWS value-based beliefs (or simply “HPWS values”) as the extent to which managers believe that they should take employee welfare into account in making business decisions and to seek out employee input. This set of beliefs is similar to what values researchers have labeled “other-oriented” (Korsgaard, Meglino, & Lester, 1997) values that include characteristics such as “concern for others” (Ravlin & Meglino, 1987), “self-transcendence” (Schwartz, 1992), and “collectivism” (Hambrick & Brandon, 1988). These other-oriented values encompass a basic belief in the importance of promoting the wel- fare of others (even when acting on this value may harm one’s own individual interests). Those higher in other-oriented values have been found to increase their willingness to seek input from those groups affected by their decisions (Ling, Zhao, & Baron, 2007) as well as to develop a “sup- portive” organizational culture (Berson, Oreg, & Dvir, 2008).

Top managers with weak HPWS values may choose not to invest in the full range of HPWS pro- grams. For example, organizational leaders with low HPWS values may choose to invest (or main- tain investments) in some HR programs such as rigorous selection, or pay based on performance, but not in other HPWS programs that may be seen as directly affecting employee welfare such as those that provide employees with employment security, internal promotion opportunities, formal training, information sharing, or participation in decision making. These additional employee-ori- ented HR programs, however, are integral parts of the HPWS and the social-exchange relation- ship that underlies this HR system (Bamberger & Meshoulam, 2000; Hom et al., 2009; White & Bryson, 2013). Because an HPWS is defined as consisting of a broad-based set of HR investments

Human Resource Management DOI: 10.1002/hrm

TOP MANAGEMENT HR VALUES 419

If the HPWS program

calls for managers to

engage in “frequent

developmental

reviews” with

employees, and yet

top managers engage

in infrequent and

nondevelopmental

reviews with

managers reporting

to them, then the

true meaning and

understanding of

this HR program

may be unclear

to those called on

to implement it to

employees.

and top management HPWS values can be seen as enhancing the consistency of HR messages and thus helping to avoid sending ambiguous “dou- ble-bind communications” to employees (Bowen & Ostroff, 2004, p. 211). The consistency of com- munication may be enhanced in this case by both the consistency in the content of the HPWS pro- grams in place as well as by stability over time in the HR-related communication sent to lower-level managers and employees by top managers.

It follows that when top management HPWS values are high, lower-level managers are more likely to implement the full range of HPWS pro- grams in ways that are consistent with the values underlying these programs—leading to a signifi- cant positive relationship between HPWS programs and employees’ HPWS perceptions of implemented HPWS practices. In contrast, when top management HPWS values are low, we expect to find a disconnec- tion between the level of HPWS pro- grams and HPWS perceptions. This disconnection (or lack of a signifi- cant positive relationship) between HPWS programs and perceptions is expected because lower-level man- agers may choose either not to implement the full range of HPWS programs or to implement them in an inconsistent or idiosyncratic way. In either case, we expect to find a weak (nonsignificant) relationship between HPWS programs and per- ceptions under this condition.

Hypothesis 3: Top management HPWS values will moderate the relationship between HPWS programs and employee perceptions of implemented HPWS practices. Specifi cally, a signifi cant pos- itive relationship between HPWS pro- grams and HPWS perceptions will be found only under conditions of high top management HPWS values.

Finally, we propose to test the overall model shown in Figure 1. The logic for this model is based on the hypotheses discussed previously. Our conceptual model represents a special case of moderated mediation labeled as “first- and sec- ond-stage moderation” by Edwards and Lambert (2007) and described by Muller, Judd, and Yzerbyt (2005). As Hayes (2013) explained, in hypoth- esizing moderated mediation models, research- ers attempt to address questions of “how” and

HR system distinctiveness refers to “features [of the HR system] that allow it to stand out in the environment, thereby capturing attention and arousing interest” (Bowen & Ostroff, 2004, p. 208). Congruence between high HPWS pro- grams and high top management HPWS values are expected to enhance HPWS distinctiveness in three ways. First, understandability of HPWS pro- grams by lower-level managers is expected to be enhanced because there is less ambiguity in the messages being sent. For example, if the HPWS program calls for managers to engage in “frequent developmental reviews” with employees, and yet top managers engage in infrequent and nonde- velopmental reviews with managers reporting to them, then the true meaning and understanding of this HR program may be unclear to those called on to implement it to employees.

Second, the perceived legitimacy of stated HPWS programs in the eyes of lower-level manag- ers and employees is expected to be undermined by perceived differences between HPWS programs and top-management HPWS values. Researchers refer to the difference between leaders’ words and deeds as leader “hypocrisy” (Brunsson, 1989; Greenbaum, Mawritz, & Piccolo, 2012; Simons, 2002). Perceived leader hypocrisy regarding HPWS programs may be associated with cynicism and disenchantment (Cha & Edmonson, 2006). HPWS programs may be seen as a management fad or fashion that will quickly pass and don’t need to be taken seriously (Abrahamson, 1991).

Third, high levels of top management HPWS values are expected to be associated with enhanced relevance of HPWS programs for lower- level managers and employees.1 Drawing on Kelman and Hamilton (1989), Bowen and Ostroff (2004) described the relevance of an HR system as the extent to which individuals perceive that the content of the HR system is instrumental in helping them to achieve both organizational and individual goals. Words and actions of top man- agers with high levels of HPWS cause-effect and value-based beliefs are expected to enhance the “motivational significance” (Kelman & Hamilton, 1989) of HPWS program implementation through (1) their ability to articulate the overall goals of an HPWS and (2) their ability to convince lower- level managers that implementation of HPWS programs will lead to both positive organizational outcomes and successful achievements for indi- vidual managers.

Top management HPWS values and beliefs can be seen as affecting other dimensions of HR system strength identified by Bowen and Ostroff (2004) as well. For example, congruence between the extent of stated HPWS programs

420 HUMAN RESOURCE MANAGEMENT, MAY–JUNE 2016

Human Resource Management DOI: 10.1002/hrm

employees. Using job titles, we identified the top management group at each location as consist- ing of the general hotel manager and the group of managers reporting directly to him or her (i.e., assistant GM, rooms division manager, front desk manager, engineering manager, executive house- keeper, and director of sales). Previous strategic HR research indicates that asking managers to describe HR programs that pertain to a specific group of “core” employees (as opposed to all employees in the firm) may improve the accuracy and reli- ability of HPWS program measures (Gerhart et al., 2000). We therefore asked managers specifically to describe HR programs that applied to front desk and housekeeping employees.

Surveys were mailed to each franchise loca- tion, and managers were asked to distribute these to other top managers and all hotel front desk and housekeeping employees. Employees and manag- ers were assured confidentiality of their responses and provided an envelope for their completed survey to be deposited into the slot of a sealable mailing box that was mailed directly back to the researchers. In order to generate meaningful unit- level averages for our variables, we omitted from our final sample any hotel from which we did not receive at least 3 management responses and 3 employee survey responses. The final sample con- sisted of 120 hotels with 648 responses from top management team members (5.4 managers per firm; 68 percent response rate) and 1,293 responses from employees (10.78 per hotel; approximately 24 percent of employees in the departments sur- veyed). The final sample of hotel franchises did not differ significantly from the original sample in terms of their average age, size, or brands.

Independent and Dependent Variables

Top Management HR Cause-Effect Belief

Based on previous research (Chattopadhyay et al., 1999; Markoczy, 1997), we constructed a cause- effect beliefs measure using success factors rel- evant to firms in our sample. With the help of firm representatives, we generated a list of five types of strategic investments for improving hotel performance over which managers in our sample had discretion: (1) technology, (2) improvements in the quality of materials, (3) HR programs, (4) physical plant improvements, and (5) customer service programs. In our survey, we asked each manager to rank each factor from 1 to 5 based on “the degree to which investments in this area impact hotel performance.” We interpreted a manager’s numerical ranking given to the “HR programs” factor to indicate the strength of his or her “cause-effect belief” linking HR programs and

“when” simultaneously. The “how” in this case is represented by the mediated relationship between the level of HR cause-effect belief and employ- ees’ HPWS perceptions of implemented HPWS practices. We propose that the effect of HR cause- effect belief on HPWS perceptions occurs through the effect of HR cause-effect belief on HPWS pro- grams. HR cause-effect beliefs are expected to lead to investments in HPWS programs, which then lead to implemented HPWS practices that are per- ceived by employees. The existence of this medi- ated relationship, however, is not expected to be present under all conditions. The “when” in this case is represented by the presence of top man- agement HPWS values. As argued in Hypotheses 2 and 3, we expect to find this mediated relation- ship only under conditions of high HPWS values.

Hypothesis 4: The mediated relationship between top management HR cause-effect belief and HPWS percep- tions through HPWS programs will be moderated by top management HPWS value-based beliefs.

Methods

Research Setting and Sample Characteristics

The setting for this study is a single large com- pany that operates, through franchise agreements, approximately 3,600 hotels worldwide composed of five brands and representing a variety of mar- ket orientations, including extended-stay, low- cost, business traveler, and luxury. The franchise agreements for independently operated hotels in our sample required that the operators follow corporate guidelines for the location’s physical layout, operations, technology, and marketing. However, the management team at each location was able to choose the approach to managing employees that they thought best. Although the holding company suggested guidelines regarding appropriate HR practices, franchisees were not required to follow these guidelines. These sample characteristics allowed us to isolate the effects of firm-level differences in top-management charac- teristics (including HPWS values) on the presence of HWPS programs and implementation effective- ness by holding constant potentially confound- ing effects of differences in industry, technology, business unit strategy, and other organization characteristics.

In total, 204 hotels throughout the United States were identified by the franchise company to represent a diversity of locations, brands, and performance levels. For each location, we sought to collect data from the full set of top-level hotel managers as well as a representative set of

Human Resource Management DOI: 10.1002/hrm

TOP MANAGEMENT HR VALUES 421

& Smith, 2006; Delery & Doty, 1996; Zacharatos, Barling, & Iverson, 2005). Following previous stra- tegic HR research, we summed the scores from the 24 HPWS program items to create an additive HPWS program index score for each hotel (Batt, 2002; Guthrie, 2001; Huselid, 1995). A complete listing of HPWS program items is provided in Appendix A.

Data on the use of HPWS programs were pro- vided by the hotel’s HR manager or, in cases in which there was no dedicated HR manager posi- tion, the general manager of the location. We assumed that the person occupying one of these positions would be the most knowledgeable about the HR programs being offered at the hotel and could act as a key informant in providing data on these programs (Arthur & Boyles, 2007). In addi- tion, data for HPWS programs were collected using a separate survey from that used to collect data on HR cause-effect beliefs and HPWS values in order to reduce common source variance.

HPWS Perceptions

When we began our survey design in 2006, there were very few published articles from which to draw items on employee perceptions regard- ing HPWS programs in their firm. We adapted as many items as possible from these published sources (e.g., Riordan, Vandenberg, & Richardson, 2005; Zacharatos et al., 2005) to generate items that referred to each HPWS program. Feedback from a pilot study of employees in a separate group of hotels (not used in the current study) indicated that we needed to shorten and simplify these items significantly in order to fit with the education level and time constraints of employ- ees in our sample. This process resulted in a list of 24 simplified items that asked employees to indi- cate their experience of HPWS practices at their location. Because our interest was in assessing the extent to which these perceptions were shared by employees at the hotel, we asked employees to report on their perceptions of experiences of employees as a group (Chan, 1998; Klein, Conn, Smith, & Sorra, 2001).

We conducted an exploratory factor analysis on the responses to these 24 items to identify any underlying dimensions. After eliminating items that loaded highly on multiple factors, two fac- tors were retained using a minimum eigenvalue criterion. These two factors explained 54 percent of the variance in the items. The 15 retained items and results of this factor analysis are pre- sented in Appendix B. The first factor contained nine items that indicated the extent to which employees agreed that employees in their hotel were provided with job security, were developed

firm performance outcomes. We averaged the “HR programs” ranking score provided by each man- ager at the hotel to obtain the score for top man- agement “HR cause-effect belief” at each location (sample mean 3.15; standard deviation [SD] = .69).

Top Management HPWS Value-Based Beliefs

A six-item scale completed by the hotel’s top- management group2 was used to assess the extent of top management HPWS values at each loca- tion. These items indicate the strength of manag- ers’ employee-centered beliefs regarding concern for employee welfare as well as the value of employee contributions. For example, agreement with the item “My primary responsibility is to create an environment that maximizes employee productivity and contribution” suggests how strongly managers’ values are directed toward employee stakeholders. In addition, managers’ agreement with items such as “As long as employ- ees continue to do a good job, they can expect to stay here as long as they want” and “I believe that it is management’s obligation to provide all employees with long-term career opportunities” are consistent with a concern for employees’ wel- fare and cultivating a social exchange relation- ship based on long-term, open-ended mutual obligations with employees. Finally, items such as “When making key decisions, I always seek the opinion and suggestions of the employees who will be affected by the decisions” indicate the willingness to take employees’ input and welfare into account in their decision making. The com- plete set of items used in the assessment of HPWS values along with means and standard deviations is provided in Appendix A. The internal reliability of the scale items (coefficient alpha) is .75.

HPWS Programs

For our measure of HPWS programs, we chose to assess the use of eight HR programs identified in the strategic HR literature as representing the con- tent of HPWS (e.g., Lepak et al., 2007):

1. Autonomous job design 2. Internal promotions 3. Performance appraisals 4. Information sharing 5. Performance-based pay 6. Employee participation 7. Employee training and development 8. Employment security

The extent to which HPWS programs existed at each hotel was assessed using management responses to 24 items (3 items per practice) adapted from previous HPWS studies (e.g., Collins

422 HUMAN RESOURCE MANAGEMENT, MAY–JUNE 2016

Human Resource Management DOI: 10.1002/hrm

We assume

that employees’

perceptions of

implemented HPWS

practices within a

given hotel location

will be significantly

affected by their

common experiences

with HPWS programs

as well as with top

managers at their

hotel.

is similar to that found in other organizational- level strategic human resource studies (Liao et al., 2009; Nishii et al., 2008; Takeuchi et al., 2007). Based on these results, we judged that the level of agreement for HPWS Perceptions was sufficient to justify aggregation.

We also considered a variety of issues and cri- teria for averaging the scores from multiple top management respondents for the HPWS values and HR cause-effect beliefs variables. Within the upper-echelon literature, there has been some discussion regarding the efficacy and need for researchers to gather data on the personal char- acteristics and beliefs of the group of top manag- ers (i.e., “dominant coalition”) versus only using data on the attributes of the CEO (Cannella & Holcomb, 2005). This discussion differs signifi- cantly from issues generally addressed in the psy- chological literature regarding the appropriate level of a construct and the need for statistical justification for data aggregation. A basic premise of the upper-echelons perspective is that execu- tives, because of their position at the top of the firm, are able to affect firm-level strategic choices and performance outcomes (Hambrick & Mason, 1984). Conceptually, researchers have argued that given the complexity of organizations and strate- gic decisions, it is logical to assume that multiple members of the firm’s top management team (in addition to the CEO) may be involved in firm-level decisions and affect performance outcomes. Thus, including attributes of multiple top management team members should improve the validity and predictive power of the upper-echelons’ models by taking into account characteristics of more of the relevant decision makers in the analysis. Empirical research has generally supported using characteristics of multiple top managers by dem- onstrating that additional variance in organiza- tional outcomes can be explained by examining the attributes of non-CEO executives (Bertrand & Schoar, 2003).

Consistent with the previously cited upper- echelons research, support for considering the effects of top management HPWS values and cause-effect beliefs at the hotel level stems from the fact that the organizational leaders who hold these values and beliefs are in a position to affect firm-level decisions and outcomes regarding hotel- level HPWS investments and implementation. To empirically examine the effect of including scores from non-GMs on our results, we reran our regres- sions with only the GM scores for HPWS values and cause-effect beliefs variables. In all cases, the predictive power of our results in explaining variation in HPWS programs as well as HPWS per- ceptions across hotels declined when only GM

and promoted internally within the hotel, and received feedback on their performance that was tied to rewards. These perceived HPWS practices are directly related to the HPWS programs in which the employer invests in employees and demonstrates a concern for their well-being (e.g., job security and career development).

The second factor consists of six items that indicate the extent to which employees agree that employee participation in decision making is encouraged and facilitated at their hotel. Sample items include “It’s easy for employees to share their ideas with management,” and “Employees

are encouraged to participate in decisions that affect their work.” Taken together, these two factors appear to capture the intended message of HPWS being sent to employees—namely, that employ- ees and their inputs are valued by the hotel and that the employer is committed to a long-term employ- ment relationship based on princi- ples of social exchange. In addition, the items in each factor showed good internal reliability (.89 and .83, respectively). We therefore decided to combine these items to form two variables. The correlation between the two variables was quite high (r = .65; p < .001). Because we view each variable as a dimension of the extent to which employees per- ceive the intended messages sent by implemented HPWS programs, we averaged the two variables to create the HPWS perceptions variable. An

alternate approach of using all 15 items to create a single HPWS perceptions variable yielded identi- cal results.

Data Aggregation

We assume that employees’ perceptions of imple- mented HPWS practices within a given hotel loca- tion will be significantly affected by their common experiences with HPWS programs as well as with top managers at their hotel. To determine the level of within-hotel agreement (and between- hotel differences) in perceived HPWS practices, we calculated intraclass correlations (ICCs) based on one-way ANOVA results (Bleise, 2000; Gerhart et al., 2000). ICC(1) for the HPWS perceptions variable is .08 (p < .001), indicating significant between-hotel differences in employees’ HPWS perceptions. ICC(2) is .51, suggesting a modest level of within-hotel agreement. ICC(2) is sensi- tive to group size and the level found in this study

Human Resource Management DOI: 10.1002/hrm

TOP MANAGEMENT HR VALUES 423

Empirical upper-

echelon research

supports the claim

that executive

characteristics

such as firm tenure,

experience in

the industry, and

education level

are predictive of

firm-level strategic

choices and

organizational

outcomes.

hotel brand, we constructed a set of three dummy- coded hotel brand variables (with business class as the omitted brand category) and included these in our regression models. In addition, because they may be associated with the use of HPWS pro- grams as well as employee perceptions, hotel size (number of rooms at each location) and hotel age (number of years since founding) were included as control variables in our analyses.

We included a dummy variable indicating whether the hotel’s workforce was represented by a labor union (union = 1) as well as a variable indicating the percentage of the hotel’s workforce that was classified as “full time.” We expected that the presence of a labor union may affect the will- ingness and ability of the hotel to adopt a HPWS as well as employee’s perception of these practices. The existence of a larger percentage of full-time employees may be positively related to a hotel’s adoption of HPWS pro- grams and may also affect employ- ees’ experience and perceptions of these programs.

Other Top Management Demographic Characteristics

Empirical upper-echelon research supports the claim that executive characteristics such as firm tenure, experience in the industry, and education level are predictive of firm-level strategic choices and orga- nizational outcomes (see Finkelstein et al., 2009, for a review). To control for the possibility that top man- agement characteristics other than HPWS values or cause-effect beliefs provide alternate explanations for our findings, we included in our regression analyses self-reported measures of managers’ years of firm- and industry-level experience as well as their highest level of education (1 = some high school; 2 = graduated high school; 3 = some college; 4 = graduated college; 5 = postgraduate education). We calculated the average score of top manage- ment respondents in the hotel to determine the hotel score for each of the demographic variables.

Results

Means, standard deviations, and intercorrela- tions between variables are presented in Table I. Hypothesis 1 predicted that the level of top-man- agement HR cause-effect belief would be posi- tively associated with HPWS program intensity. Ordinary least squares (OLS) regression analysis

measures where included. These results are con- sistent with previous upper-echelons empirical research and point to the practical importance of including the values and beliefs of multiple top management respondents (including the GM/ CEO).

ICC(1) statistics for our aggregate measures of top managers’ HPWS values (.07, p < .05) and cause-effect beliefs (.04, p = .05) indicate a mod- erate level of hotel-level effects (between-hotel differences) on the average scores for these vari- ables. These hotel-level effects may be the result of attraction, selection, and attrition processes within hotels based on similarities in top manage- ment values and beliefs. The ICC(2) statistics for HPWS values (.30) and cause-effect beliefs (.18), however, indicate that hotel-level effects are quite modest and that a significant amount of within- hotel heterogeneity exists. Given that personal values and beliefs are considered to be relatively stable individual-level characteristics, within-hotel heterogeneity on these characteristics among top managers is not surprising (Cannella & Holcomb, 2005). Upper-echelon researchers have generally followed other group-level demographic diver- sity researchers in suggesting that heterogeneity among top managers’ characteristics and beliefs may have both positive and negative effects on performance outcomes (see Finkelstein et al., 2009, for a review of this research on top manage- ment teams). By using average hotel-level scores, we propose that interactions between managers will tend to balance out the effects of individual managers with “extreme” high and low scores on these variables in terms of the overall effect on hotel-level HPWS program investment and imple- mentation effectiveness.

Control Variables

Organizational characteristics such as hotel size, age, and brand, as well as measures of the top management demographic characteristics (i.e., average education level, years in industry, years with the hotel) were included in the analysis to statistically control for alternate explanations for our findings.

Organizational and Workforce Characteristics

Our sample consisted of four brands of franchised hotel chains. Each of these brand categories is essentially a business unit strategy that focuses the hotel on a different segment of the customer market: economy class, business class, extended- stay, and upscale. To statistically control for pos- sible covariation between hotel brands and the use of HPWS as well as systematic differences in employee and managerial perceptions based on

424 HUMAN RESOURCE MANAGEMENT, MAY–JUNE 2016

Human Resource Management DOI: 10.1002/hrm

T A

B L

E

I

M ea

ns , S

ta nd

ar d

D ev

ia tio

ns , a

nd C

or re

la tio

ns a

M ea

n S

.D .

1 2

3 4

5 6

7 8

9 10

11 12

13

1 A

g e

(Y ea

rs s

in ce

f o

u n

d in

g )

15 .0

2 12

.3 9

2 S

iz e

(N u

m b

er o

f ro

o m

s)

16 5.

0 82

.8 1

.3 3*

*

3 B

ra n

d 1

( E

co n

o m

y) 0.

29 0.

46 –.

36 **

–. 51

**

4 B

ra n

d 2

( U

p sc

al e)

0. 08

0. 28

–. 09

.3 9*

* –.

19 *

5 B

ra n

d 3

( E

xt en

d ed

S ta

y) 0.

08 0.

28 –.

25 **

–. 22

* –.

19 *

–. 09

6 To

p -m

g t Y

ea rs

w it

h F

ir m

5. 83

4. 10

.5 5*

* .2

2* –.

22 *

.0 3

–. 22

*

7 To

p -m

g t Y

ea rs

in In

d u

st ry

14 .8

5 5.

77 .3

0* *

.4 1*

* –.

27 **

.2 7*

* –.

17 .4

7* *

8 To

p -m

g t

H ig

h es

t E

d u

c Le

ve l

3. 09

0. 48

.0 4

.2 0*

–. 35

** .2

1* .1

8 –.

02 .0

1

9 U

n io

n iz

ed W

o rk

fo rc

e 0.

06 0.

24 –.

10 .2

5* *

–. 00

.1 8*

–. 08

.0 0

.0 6

–. 00

10 Pe

rc en

t Fu

ll- ti

m e

E m

p lo

ye es

73

.5 3

25 .9

4 –.

19 *

.2 4*

* –.

11 .1

1 .1

3 .1

6 –.

00 .1

8 .1

6

11 H

R C

au se

-E ff

ec t

B el

ie fs

3. 15

0. 69

.0 1

–. 11

.1 4

–. 20

* –.

09 –.

10 –.

07 –.

03 .0

5 –.

07

12 H

P W

S V

al u

es 4.

02 0.

33 –.

07 .0

5 .1

9* –.

11 .1

3 –.

22 *

–. 12

–. 07

–. 01

.1 9*

.3 4*

*

13 H

P W

S P

ro g

ra m

s 20

.6 0

4. 66

–. 10

–. 05

.1 1

–. 03

.0 2

–. 13

–. 13

.0 4

–. 05

.1 6

.2 6*

* .3

1* *

14 H

P W

S P

er ce

p ti

o n

s 3.

49 0.

41 –.

04 –.

11 –.

03 –.

15 .1

3 –.

08 –.

20 *

–. 11

–. 21

* .0

2 .1

4 .0

7 .2

6* *

a N =

1 14

–1 20

h o

te l f

ra n

ch is

es .

*p <

.0 5,

* *p

< .0

1.

Human Resource Management DOI: 10.1002/hrm

TOP MANAGEMENT HR VALUES 425

explanatory power of the model (change in R2 = .07; p < .01).

Table III presents the results of our test of Hypothesis 3, which states that HPWS values moderates the relationship between HPWS pro- grams and HPWS perceptions. We mean-centered the HPWS values and HPWS programs variables prior to creating a multiplicative interaction term. As shown in Table III, the interaction term is a statistically significant predictor of HPWS percep- tions (b = .08; p < .01). Adding this interaction term to Model 3 containing control variables and main effects significantly improves the model as shown in Model 1 (change R2-squared = .07; p < .01). These results support Hypothesis 2.

To gain a better understanding of the form of the interaction effects shown in Tables II and III, we plotted relationships between our inde- pendent and dependent variables at “high” and

results presented in Model 1 of Table II provide support for this hypothesis. Table II shows that the variable HR cause-effect beliefs is significantly related to our HPWS program measure after statis- tically controlling for the effects of hotel age, size, and brand as well as the demographic character- istics of top managers and the union status and percent of full-time employees (b = 1.86; p < .01). Hypothesis 2 predicted that HPWS values would moderate the effect of HR cause-effect beliefs on HPWS programs. We constructed the interaction term HR cause-effect belief * HPWS values after first mean-centering HR cause-effect belief and HPWS values variables (Aiken & West, 1991). As shown in Table II, the coefficient for this interaction term is a statistically significant predictor of HPWS programs (b = 5.18; p < .01). Adding the interac- tion term to the Model 1 containing control vari- ables and main effects significantly improves the

T A B L E I I Regression Results for Moderating Effect of HPWS Values on HPWS Programsa

Model 1 Model 2 Model 3

Constant 20.61** 20.62** 3.58**

Age of hotel –.01 –.02 .01

Size of hotel (rooms)

.00 .00 .00

Brand 1(Economy)b .99 .46 .44

Brand 2 (Upscale)b .97 1.01 –.30

Brand 3 (Ext stay)b –.22 –.80 –.82

Top-mgt years with fi rm

–.01 .03 –.01

Top-mgt years in industry

–.09 –.08 –.07

Top-mgt highest level of education

.32 .45 .29

Unionized work- force

–2.30 –2.08 –1.82

Percent full-time employees

.03* .02 .04*

HR cause-effect beliefs 1.86** 1.43* 1.31

Top-management HPWS values 2.71 2.57

HR Cause-effect beliefs × HPWS Values

5.18**

R-squared .13 .16 .23*

R-squared change 0.03 .07** a N = 112; Unstandarized regression coeffi cients. Variables listed in italics are those that are used to test our hypotheses. bDummy coded, business class hotel is the omitted brand. *p < .05, **p < .01 (two-tailed tests).

T A B L E I I I Regression Results of Moderating Effect of HPWS Values on HPWS Perceptionsa

Model 1 Model 2 Model 3

Constant 3.49 3.49 3.45

Age of hotel –.00 .00 –0.00

Size of hotel (rooms)

.00 .00 0.00

Brand 1(Economy)b –.13 –.12 –0.15

Brand 2 (Upscale)b –.01 –.01 0.04

Brand 3 (Ext stay)b .161 .17 0.19

Top-mgt years with fi rm

.01 .01 0.01

Top-mgt years in industry

–.01* –.01* –0.02*

Top-mgt highest level of education

–.14 –.14 –0.17*

Unionized work- force

–.34* –.34* –0.42*

Percent full-time employees

.00 .00 .00

HPWS programs .02** 0.02** 0.02*

Top-management HPWS values –0.04 –0.06

HPWS programs × HPWS Values

0.08**

R-squared .17* .18 .25**

Change in R-squared

.01 .07**

aN = 112; Unstandarized regression coeffi cients. Variables listed in italics are those that are used to test our hypotheses. bDummy coded, business class hotel is the omitted brand. *p < .05, **p < .01 (two-tailed tests).

426 HUMAN RESOURCE MANAGEMENT, MAY–JUNE 2016

Human Resource Management DOI: 10.1002/hrm

high HPWS values. Under the low HPWS values condition, the relationship between HPWS pro- grams and HPWS perceptions appears to be slightly negative. Our test of the significance of these simple slopes again revealed a significant positive relationship under the high HPWS values condi- tions, but no statistically significant effect under the low HPWS values condition. These results sup- port our arguments that the positive effects of HR cause-effect belief and on HPWS programs as well as the positive relationship between HPWS values on HPWS perceptions are conditional on a high level of top management HPWS values.

Tests of Moderated Mediation

We used PROCESS software (Hayes, 2013) to simultaneously test all of the proposed relation- ships shown in the moderated mediated model in Figure 1. Hayes (2013) developed PROCESS soft- ware to support empirical tests of various forms of moderated mediation (or “conditional indirect effects”) using SAS and SPSS statistical programs.

“low” levels of HPWS values (i.e., one standard deviation above and below the mean). These plots are shown in Figures 2 and 3. Figure 2 shows that under conditions of high HPWS values, there is a positive relationship between the level of HR cause-effect belief and HPWS programs. Under the low HPWS values condition, however, the relationship between HR cause-effect belief and HPWS programs appears to be slightly negative. Using the technique suggested by Aiken and West (1991), we calculated the significance of both of these simple slopes (Hayes, 2013). We found that the positive simple slope under the high HPWS values was statistically significant (p < .05). However, the negative simple slope under the low HPWS values condition was not significantly dif- ferent from zero.

Figure 3 shows the plot of the interaction of HPWS programs and HPWS values on HPWS per- ceptions. Similar to the previous plot, this plot shows a positive relationship between HPWS pro- grams and HPWS perceptions under conditions of

FIGURE 2. Plot of Interaction of HR Cause-Effect Beliefs and HPWS Values on HPWS Programs

20

22

24

26

28

30

32

highmedlow

H P

W S

P ro

g ra

m s

HR Cause-Effect Beliefs

HPWS Values

high

med

low

FIGURE 3. Plot of Interaction of HPWS Programs and HPWS Values on HPWS Perceptions

3

3.1

3.2

3.3

3.4

3.5

3.6

3.7

highmedlow

P er

ce iv

ed H

P W

S

HPWS Programs

HPWS Values

high

med

low

Human Resource Management DOI: 10.1002/hrm

TOP MANAGEMENT HR VALUES 427

We found that a

significant positive

relationship

between the HPWS

program intensity

and employees’

HPWS perceptions

was contingent

on the level of top

management HPWS

values.

Discussion

We began the current study by considering a puz- zle posed by findings from empirical strategic HR research: there appears to be a positive relation- ship between the intensity of HPWS programs adopted by firms and firm-level performance out- comes, yet researchers continue to find that only a relatively small subset of firms report having adopted the full set of HR programs that make up an HPWS (Kaufman & Miller, 2011). What, then, explains this persistent interfirm heterogeneity in the use of HPWS? Applying upper-echelons theory, we proposed and found that heterogeneous top manage- ment characteristics in the form of HR cause-effect beliefs and HPWS value-based beliefs provide part of the solution to this puzzle. First, we found that hotel franchisees differed in the extent to which their top managers believe that investments in HR programs (relative to other potential investments) will result in improved financial performance. These differences were found to be significantly related to the inten- sity of HPWS programs at hotels. Second, we found that the existence of top-management value-based beliefs that focus on management’s concern and responsibility for the well-being of employee stake- holders had a moderating effect on the intensity of HPWS program adoption in hotels. Third, we found that the level of top- management HPWS values affected the quality of HPWS program implementation in organizations. Specifically, we found that a significant positive relationship between the HPWS program inten- sity and employees’ HPWS perceptions was con- tingent on the level of top-management HPWS values. In sum, we found that firms’ ability to translate top managers’ cause-effect beliefs about the economic value of HR investments into adop- tion of HPWS programs, as well as their ability to effectively implement these HPWS programs, was

PROCESS uses a path analysis approach described by Preacher, Rucker, and Hayes (2007) and Edwards and Lambert (2007) and allows for a bootstrap test of indirect effects in mediation (MacKinnon, Lockwood, Hoffman, West, & Sheets, 2002) at various levels of the moderator(s). An important advantage of the bootstrapping technique over the more traditional Sobel (1982) test of mediation is that confidence intervals based on bootstrapping adjust for the nonnormality of the distribution of mediated effects (MacKinnon et al., 2002).

We used the PROCESS program to test the first- and second-stage moderated mediation model shown in Figure 1 and present our results in Table  IV. Table IV shows confidence intervals for bootstrap tests our model at three values of the HPWS values: (1) low (i.e., one standard deviation below mean), (2) mean level, and (3) high (one standard deviation above the mean). The model we tested includes all control variables shown in Tables II and III. We used a bootstrap test of the indirect effect based on 10,000 samples with replacement (MacKinnon et al., 2002). The 95 percent confidence intervals (CIs) shown in Table IV are considered to be statistically significant if the values between the low and high CIs do not include zero. As can be seen in Table IV, the boot- strap CI for the indirect effect when HPWS Values are low (–01 to .06) includes zero, indicating that there is no significant indirect effect under this condition. The test of the indirect effect of HR cause-effect belief on HPWS perceptions when HWPS values are at the mean is also rejected at the 95 percent confidence level. In contrast, results in Table IV show that under conditions of high HPWS values, there is a significant indirect (i.e., mediated) effect of cause-effect beliefs on employ- ees’ HWPS perceptions through HPWS programs (95 percent CI range = .04 to .23, which does not include zero). These results reinforce those shown in Figures 2 and 3—namely, that only under con- ditions of high top management HPWS values do we find a significant mediated effect. The medi- ated (or indirect) effect is therefore said to be con- ditional on a high level of top management HPWS value-based beliefs, which supports Hypothesis 4.

T A B L E I V Bootstrap Test of Conditional Indirect (CI) Effect of HR Cause-Effect Beliefs on HPWS Practices at  Different Levels of HPWS Values

HPWS Values Boot Indirect

Effect Boot SE Boot Lower CI

(alpha=.95) Boot Upper CI

(alpha=.95)

–1 SD (–.326) .004 .015 –.012 .057

Mean (0) .018 .015 –.000 .066

+1 SD (.326) .111 .0542 .040 .233

428 HUMAN RESOURCE MANAGEMENT, MAY–JUNE 2016

Human Resource Management DOI: 10.1002/hrm

theory-based explanation of why we find persis- tent interfirm heterogeneity in the adoption and use of HPWSs. As noted previously, this finding has been identified as a central issue in strategic HR research—especially as it relates to resource- based view arguments concerning HPWSs as a potential source of competitive advantage (e.g., Kaufman & Miller, 2011; Wright et al., 2001). In focusing attention on the central role played by top managers’ beliefs in affecting a firm’s ability to acquire and utilize an HPWS, this study adds to and enhances previous research that emphasizes the importance of managerial skills and capabili- ties in shaping a firm’s ability to gain and sustain competitive advantage through the use of HR sys- tems (Chadwick & Dabu, 2009).

Implications for Practice

We believe our conceptual framework and find- ings have important implications for management practice. Implications for selection of top man- agers may be the most obvious. The study finds that beyond education and experience, the beliefs held by top managers about the importance of HR investments as well as their employee-centered values can affect their willingness and ability to successfully adopt and implement an HPWS. Since the uses of HPWS programs are positively related to firm-performance outcomes, consideration of these beliefs in selecting members of the top man- agement group appears relevant.

Of course, other managerial skills and char- acteristics such as knowledge of the industry and ability to conceive and execute strategies that don’t involve acquiring and implementing an HPWS are also expected to be critical to firm per- formance. By selecting these other relevant mana- gerial characteristics, it is possible that firms will find themselves in a situation in which members of the top management group do not share the HR cause-effect beliefs and employee-centered HPWS values that we have identified. The findings from our study suggest two possible responses to this situation. First, a firm might decide to pur- sue a strategy that does not require implement- ing an HPWS. Our results suggest that the extent to which HPWS program investments translate into employee HPWS perceptions is limited under conditions of low top-management HPWS val- ues. It might, therefore, be more cost effective for firms in which top managers don’t share these values to invest in an HR system other than an HPWS. Further research is required to support this possibility.

A second possible response to a situation in which top-management HPWS values are low is to increase managerial awareness of the extent to

significantly affected by the value-based beliefs of top managers. Since high levels of top manage- ment HPWS values appear to be relatively rare, these results help to explain why we continue to see differences in the level of HPWS adoption among similar firms, despite widespread knowl- edge regarding their economic benefits.

Research Implications

Ours is the first study we are aware of to apply an upper-echelons perspective to the study of HPWS. We believe that it has implications for both upper- echelons theory and future strategic HR research. Our model expands the predictions of upper-ech- elons theory beyond the content of a firm’s stra- tegic choices to include implementation issues as well. Strategic management researchers have long argued that the distinction between strategy “con- tent” and “process” is artificial (e.g., Helfat et al., 2007, pp. 30–34). In the case of HPWSs, it does not appear to be meaningful to discuss the value of the content of an HPWS as a strategic asset in the absence of its successful implementation. We augment the implications of upper-echelons theory to strategic HR researchers by arguing that top-management value-based beliefs (and the cor- respondence between leaders’ words and deeds) can affect HPWS implementation by affecting the distinctiveness, consistency, and consensus HPWS programs to both lower-level managers and employees (Bowen & Ostroff, 2004).

Our findings regarding the relationship between HPWS programs and HPWS perceptions reinforce the need for strategic HR researchers to pay closer attention to issues regarding the quality of HPWS implementation. This theme regarding the process through which the intended effects of HR practices by upper management translate into employees’ perceptions and behaviors, and eventually into firm-level outcomes, has been a subject of continued interest among strategic HR researchers (Boxall, Ang, & Bartram, 2011; Guest, 2011). We add to the understanding of the “black box” between intended HR programs and firm performance outcomes by providing additional insights into the sources of unit-level differences in employees’ experience and perceptions of stated HR programs. Specifically, we show that top-management values appear to affect how HR programs are implemented and communicated to lower-level managers and employees. Thus the value-based beliefs of organizational leaders appear to provide an important (and largely over- looked) contingency factor in explaining differ- ences in HPWS implementation effectiveness.

Finally, this study contributes to the stra- tegic HR literature by providing evidence for a

Human Resource Management DOI: 10.1002/hrm

TOP MANAGEMENT HR VALUES 429

In contrast to some

strategic investments

that represent

one-time events

(e.g., an equipment

purchase or firm

acquisition decision),

investments in

HPWS programs

must be continuously

supported and

maintained.

HPWS values to be attracted to firms with existing HPWS programs, we would not expect these pro- grams to continue without the subsequent actions and decisions of managers.

Second, we found significant within-hotel heterogeneity among top managers regarding their HPWS values and cause-effect beliefs. This heterogeneity is not surprising, given that these variables are considered to be relatively stable characteristics of individual managers. We would expect, however, that the group dynamics and decision-making processes may differ for hotels characterized by different levels of value-based het- erogeneity among organization leaders (Cannella & Holcomb, 2005). Examination of the effects of this top-management heterogeneity on firm-level HPWS investments and implementation effec- tiveness is beyond the scope of this study. Future research that examines these poten- tial effects as well as the potential impact of power differences among top-management group members would be especially useful.

Third, the fact that the content of the items used for our HPWS val- ues variable may be seen as socially desirable could also be seen as a limitation. In other words, man- agers may simply be respond- ing in ways that they believe will result in social approval, rather than describing what they really believe. Although the confidential nature of our survey protocol as well as gathering responses from multiple managers from each orga- nization may have mitigated this bias somewhat, we assume that responses to HPWS values ques- tions contain some degree of social desirability bias. This bias is analo- gous to Argyris and Schön’s (1978) distinction between “stated values” and “theories-in-use.” While it is common to view managers’ under- lying theories-in-use as “real” and publicly stated values as “not real,” we believe that in the context of understanding the impact of top- management values on HPWS investments and implementation, the interpretation of these dis- tinctions becomes more complex. As mentioned previously, a distinctive characteristic of values (compared to other types of beliefs) is that val- ues reflect a person’s subjective view of what is socially desirable (Meglino & Ravlin, 1998; Rokeach, 1973). Trying to tease out social desir- ability from the measurement of values would thus appear to be counterproductive. In our

which the firms’ espoused values are perceived by lower-level managers and employees as inconsis- tent with management actions. Because personal values are defined as relatively stable aspects of a person’s psychological makeup, we don’t expect that leaders will be able to change their values (even if they wanted to). However, as Simons (2002) points out, consistency between words and deeds is based on perceptions and attributions by others. It’s possible that leaders may be unaware of when their stated values and behaviors are per- ceived as inconsistent by others. In their qualita- tive study, Cha and Edmonson (2006) found that employees’ perceptions of the leader’s values had expanded beyond what the leader intended. The leader was therefore unaware that certain behav- iors were perceived by employees as violating the values that he espoused. Kouzes and Posner (2007) prescriptively call on organizational lead- ers to spend time clarifying their values to them- selves and then to model these values in their actions with others. Additional research is needed to determine the extent to which activities that increase awareness by executives of their underly- ing values (as well as how those values are per- ceived by lower-level managers and employees) affect a leader’s perceived behavioral integrity as well as HPWS implementation effectiveness.

Limitation and Future Research

Like all studies, this one contains limitations that should be taken into account in interpreting the validity and generalizability of our findings. First, our cross-sectional data do not allow us to explicitly test the proposed causal order linking HR cause-effect belief, HPWS value-based beliefs, HPWS programs, and HPWS perceptions. For example, it’s possible that hotels with high levels of HPWS programs are more attractive to top man- agers with corresponding levels of HR cause-effect and HPWS value-based beliefs. This potential endogeneity of top-management characteristics is a common problem in upper-echelon research (Hambrick, 2007). Although it remains a possi- bility in our study, the nature of HPWS program investments argues against this reverse-causality explanation. In contrast to some strategic invest- ments that represent one-time events (e.g., an equipment purchase or firm acquisition decision), investments in HPWS programs must be continu- ously supported and maintained. For example, one can’t simply put into place an employee involvement or development program and expect that it will continue to operate as intended with- out additional top management support and attention. Thus, although there may be an ini- tial sorting effect that leads managers with high

430 HUMAN RESOURCE MANAGEMENT, MAY–JUNE 2016

Human Resource Management DOI: 10.1002/hrm

perceptions of front desk and housekeeping employees, the possibility that different HR prac- tices and implementation issues exist for managers or other employee groups within these organiza- tions could not be assessed in the study.

Conclusion

Given the positive relationships found in previ- ous studies between the intensity of HPWS pro- grams and firm performance outcomes, both strategic HR researchers and managers have strong interests in understanding what leads to the adoption and implementation of a full range of HPWS programs in some firms and not others. In this study, we take the position that in order to better understand when and how HPWSs are adopted and implemented in firms, we need to pay closer attention to the beliefs and values of key decision makers in the firm—namely, the top management group. We hypothesized and found that, after controlling for a variety of organizational, workforce, and managerial char- acteristics, variation in the intensity of HPWS programs and employee perceptions found in hotel franchises was significantly related to top managers’ beliefs about the relative importance of HR investments as well as their employee- centered values. These results suggest that for those hoping to increase the use of HPWSs in firms, simply providing more information about the economic benefits of HR investments may not be enough. We find that there is a need to directly address the belief structures and values of organizational leaders to ensure that these are compatible with the adoption and effective implementation of HPWS.

Notes

1. We thank an anonymous reviewer for pointing out these connections to us.

2. We identifi ed the top management group at each hotel based on job titles that we expected to be sig- nifi cantly involved in the hotel franchise’s important strategic and operational decisions (cf. Finkelstein et  al., 2009, pp. 126–127). Specifi cally, the small group of managers at the top of each independent hotel franchise was identifi ed as the general man- ager and the heads of each of the hotel divisions (e.g., front desk, housekeeping, maintenance, sales, and human resources). Although our defi nition of top management group overlaps with the more commonly used term top management team (TMT), we have stayed away from using the term TMT when referring to the group of top managers targeted in this study to avoid its connotation of corporate-level executive offi cers.

case, since managers responded to our questions about their level of HPWS values differently, we assume that managers have different perceptions about what they consider to be socially desirable for managing employees in their organization. In addition, the fact that these stated HPWS val- ues were found to be predictive of the presence of HPWS programs suggests to us that (at least in the case of highly visible choices regarding management of people), differences in leaders’ perceptions of what is socially desirable may be as important in shaping strategic choices and behaviors as their less publicly held beliefs.

A fourth limitation of our study is that we don’t directly measure the quality of HPWS imple- mentation by lower-level managers. We have pro- posed what we think is a plausible explanation of how the correspondence between the level of HPWS programs and top management HPWS val- ues affects the level of employees’ HPWS percep- tions. There may, however, be other mechanisms at work that would also be consistent with our empirical findings. Validation of the proposed causal mechanisms that link top management value-based beliefs with perceived HPWS practices will require additional research.

Fifth, the focused nature of our sample may limit the generalizability of our results to other industry settings. For example, it was somewhat surprising to us that none of the organizational control variables such as organization age, size, and brand were significantly related to level of HPWS programs in the regression results in Tables II and III. Some previous industry-level studies have found these variables to be impor- tant predictors (Arthur, 1992; Batt, 2002). One explanation for these findings is that the basic production function and technology at hotels (i.e., checking in guests, cleaning rooms, check- ing out) is similar across hotels. Thus, there may be fewer structural imperatives for these franchi- sees to choose a specific type of HR system than would be found in other settings. In addition, our discussions with the holding company man- agement staff in this case suggested that inde- pendent hotel franchisees felt strongly that they should be allowed freedom to manage employees as they saw fit. We suspect that in other indus- try settings, external institutional and organiza- tion structural features would have a larger role in limiting managerial discretion and interfirm HPWS program variability.

Finally, previous strategic HR literature sug- gests that firms may use different HR practices for different types of employees within the same organization (e.g., Lepak & Snell, 1999). Since we chose to limit our focus to HR practices and

Human Resource Management DOI: 10.1002/hrm

TOP MANAGEMENT HR VALUES 431

the HRM system. Academy of Management Review, 29, 203–221.

Boxall, P., Ang, S. H., & Bartram, T. (2011). Analysing the “black box” of HRM: Uncovering HR goals, mediators, and outcomes in a standardized service environment. Journal of Management Studies, 48, 1504–1532.

Brunsson, N. (1989). The organization of hypocrisy: Talk, deci- sions, and actions in organizations. New York, NY: Wiley.

Cannella, A. A., & Holcomb, T.R. (2005). A multi-level analy- sis of the upper-echelons model. Research in Multi-level Issues, 4, 197–237.

Cha, S. E., & Edmondson, A. C. (2006). How values backfi re: Leadership, attribution and disenchantment in a values- driven organization. Leadership Quarterly, 17, 57–78.

Chadwick, C., & Dabu, A. (2009). Human resources, human resource management, and the competitive advantage of fi rms: Toward a more comprehensive model of causal linkages. Organization Science, 20, 253–272.

Chan, D. (1998). Functional relations among constructs in the same content domain at different levels of analysis: A typology of composition models. Journal of Applied Psychology, 83, 234–246.

Chattopadhyay, P., Glick, W. H., Miller, C. C., & Huber, G. P. (1999). Determinants of executive beliefs: Comparing functional conditioning and social infl uence. Strategic Management Journal, 20, 763–789.

Collins, C. J., & Smith, K. G. (2006). Knowledge exchange and combination: The role of human resource practices in the performance of high-technology fi rms. Academy of Management Journal, 49, 544–560.

Combs, J. G., Yongmei, L., Hall, A., & Ketchen, D. J. (2006). How much do high-performance work practices matter? A meta-analysis of their effects on organizational per- formance. Personnel Psychology, 59, 501–528.

Delery, J. E., & Doty, D. H. (1996). Modes of theorizing in strategic human resource management: Tests of uni- versalistic, contingency, and confi gurational perform- ance predictions. Academy of Management Journal, 39, 802–835.

Dyer, L., & Holder, G. W. (1988). Toward a strategic perspec- tive of human resource management. In L. Dyer (Ed.), Human resource management: Evolving roles and responsibilities. Washington, DC: Bureau of National Affairs.

References

Abrahamson, E. (1991). Managerial fads and fashions: The diffusion and rejection of innovations. Academy of Man- agement Review, 16, 586–612.

Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Newbury Park, CA: Sage.

Argyris, C., & Schön, D. (1978). Organizational learning: A theory of action perspective. Reading, MA: Addison- Wesley.

Arthur, J. B. (1992). The link between business strategy and industrial relations systems in American steel minimills. Industrial and Labor Relations Review, 45, 488–506.

Arthur, J. B., & Boyles, T. (2007). Validating the human resource system structure: A levels-based strategic HRM approach. Human Resource Management Review, 17, 77–92.

Bamberger, P., & Meshoulam, I. (2000). Human resource strategy. Thousand Oaks, CA: Sage.

Batt, R. (2002). Managing customer services: Human resource practices, quit rates, and sales growth. Academy of Management Journal, 45, 587–597.

Becker, B. E., & Gerhart, B. (1996). The impact of human resource management on organizational performance: Progress and prospects. Academy of Management Jour- nal, 39, 779–801.

Becker, B. E., & Huselid, M. A. (2006). Strategic human resource management: Where do we go from here? Jour- nal of Management, 32, 898–925.

Berson, Y., Oreg, S., & Dvir, T. (2008). CEO values, organiza- tional culture and fi rm outcomes. Journal of Organiza- tional Behavior, 29, 615–633.

Bertrand, M., & Schoar, A. (2003). Managing with style: The effect of managers on fi rm policies. Quarterly Journal of Economics, 113, 1169–1208.

Blau, P. M. (1964). Exchange and power in social life. New York, NY: Wiley.

Bleise, P. D. (2000). Within-group agreement, non-independ- ence, and reliability: Implications for data aggregation and analysis. In K. J. Klein & S. W. J. Kozlowski (Eds.), Multi-level theory, research and methods in organizations (pp. 349–381). San Francisco, CA: Jossey-Bass.

Bowen, D. E., & Ostroff, C. (2004). Understanding HRM- fi rm performance linkages: The role of the “strength” of

JEFFREY B. ARTHUR (PhD, Cornell University) is an associate professor of management at the Pamplin College of Business at Virginia Tech. His research focuses on the effect of human resource systems on employees and organizational performance outcomes.

ANDREW O. HERDMAN (PhD, Virginia Tech) is an associate professor of management within the College of Business at East Carolina University. His research is focused in the areas of leadership and strategic human resource management.

JAEWAN YANG (PhD, Virginia Tech) is an assistant professor of human resource manage- ment at the New York Institute of Technology. His research interests include strategic human resource management, group process and leadership, and workplace climate.

432 HUMAN RESOURCE MANAGEMENT, MAY–JUNE 2016

Human Resource Management DOI: 10.1002/hrm

Edwards, J. R., & Lambert, L. S. (2007). Methods for inte- grating moderation and mediation: A general analytical framework using moderated path analysis. Psychological Methods, 12, 1–22.

Finkelstein, S., Hambrick, D. C., & Cannella, A. A. (2009). Strategic leadership: Theory and research on executives, top management teams, and boards. New York, NY: Oxford University Press.

Gerhart, B., Wright, P. M., McMahan, G. C., & Snell, S. A. (2000). Measurement error in research on human resource decision and fi rm performance: How much error is there and does it infl uence effect size estimates? Per- sonnel Psychology, 53, 803–834.

Gouldner, A. W. (1960). The norm of reciprocity: A pre- liminary statement. American Sociological Review, 25, 161–178.

Greenbaum, R. L., Mawritz, M. B., & Piccolo, R. F. (2012). When leaders fail to “walk the talk”: Supervisor under- mining and perceptions of leader hypocrisy. Journal of Management, Published online before print April 24, 2012. doi: 10.1177/0149206312442386

Guest, D. (2011). Human resource management and per- formance: Still searching for some answers. Human Resource Management Journal, 21, 3–13.

Guthrie, J. P. (2001). High-involvement work practices, turno- ver, and productivity: Evidence from New Zealand. Acad- emy of Management Journal, 44, 180–190.

Hambrick, D. C. (2007). Upper echelons theory: An update. Academy of Management Review, 32, 334–343.

Hambrick, D. C., & Brandon, G. (1988). Executive values. In D. C. Hambrick (Ed.), The executive effect: Concepts and meth- ods for studying top managers. Greenwich, CT: JAI Press.

Hambrick, D. C., & Mason, P. A. (1984). Upper echelons: The organization as a refl ection of its top managers. Academy of Management Review, 9, 193–206.

Hayes, A.F. (2013). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. New York, NY: Guilford Press.

Helfat, C.E., Finkelstein, S., Mitchell, W., Peteraf, M. A., Singh, H., Teece, D. J., & Winter, S. G. (2007). Dynamic capabili- ties: Understanding strategic change in organizations. Oxford, England: Blackwell Publishing.

Hitlin, S., & Piliavin, J. A. (2004). Values: Reviving a dormant concept. Annual Review of Sociology, 30, 359–393.

Hom, P. W., Tsui, A. S., Lee, T. W., Fu, P. P., Wu, J. B., Zhang, A. Y., & Li, L. (2009). Explaining employment rela- tionships with social exchange and job embeddedness. Journal of Applied Psychology, 94, 277–297.

Huselid, M. A. (1995). The impact of human-resource man- agement-practices on turnover, productivity, and corpo- rate fi nancial performance. Academy of Management Journal, 38, 635–672.

Kaufman, B.E. (2010). SHRM theory in the post-Huselid era: Why it is fundamentally misspecifi ed. Industrial Relations, 49, 286–313.

Kaufman, B. E., & Miller, B. I. (2011). The fi rm’s choice of HRM practices: Economics meets strategic human resource management. Industrial and Labor Relations Review, 64, 226–257.

Kelman, H. C., & Hamilton, V. C. 1989. Crimes of obedience: Toward a social psychology of authority and responsibil- ity. New Haven, CT: Yale University Press.

Klein, K. J., Conn, A. B., Smith, D. B., & Sorra, J. S. (2001). Is everyone in agreement? An exploration of within-group agreement in employee perceptions of the work environ- ment. Journal of Applied Psychology, 86, 3–16.

Kochan, T. A., Katz, H. C., & McKersie, R. B. (1986). The trans- formation of American industrial relations. New York, NY: Basic Books.

Korsgaard, M. A., Meglino, B. M., & Lester, S. W. (1997). Beyond helping: Do other-oriented values have broader implications in organizations? Journal of Applied Psy- chology, 82, 160–177.

Kouzes, J. M., & Posner, B. Z. (2007). The leadership chal- lenge (4th ed.). San Francisco, CA: Jossey-Bass.

Lepak, D. P., & Snell, S. A. (1999). The human resource archi- tecture: Toward a theory of human capital allocation and development. Academy of Management Review, 24, 31–48.

Lepak, D. P., Taylor, M. S., Tekleab, A., Marrone, J. A., & Cohen, D. J. (2007). An examination of the use of high- investment human resource systems for core and sup- port employees. Human Resource Management, 46, 223–246.

Liao, H., Toya, K., Lepak, D. P., & Hong, Y. (2009). Do they see eye to eye? Management and employee perspectives of high-performance work systems and infl uence processes on service quality. Journal of Applied Psychology, 94, 371–391.

Ling, Y., Zhao, H., & Baron, R. A. (2007). Infl uence of founder- CEOs’ personal values on fi rm performance: Moderating effects of fi rm age and size. Journal of Management, 33, 673–696.

MacKinnon, D. P., Lockwood, C. M., Hoffman, J. M., West, S. G., & Sheets, V. (2002). A comparison of methods to test mediation and other intervening variable effects. Psycho- logical Methods, 7, 83–104.

March, J. G., & Simon, H. A. (1958). Organizations. New York, NY: Wiley.

Markoczy, L. (1997). Measuring beliefs: Accept no substi- tutes. Academy of Management Journal, 40, 1228–1242.

Meglino, B. M., & Ravlin, E. C. (1998). Individual values in organizations: Concepts, controversies and research. Journal of Management, 24, 351–389.

Muller, D., Judd, C. M., & Yzerbyt, V. V. (2005). When modera- tion is mediated and mediation is moderated. Journal of Personality and Social Psychology, 89, 852–863.

Nishii, L. H., Lepak, D. P., & Schneider, B. (2008). Employee attributions of the “why” of HR practices: Their effects on employee attitudes and behaviors, and customer satis- faction. Personnel Psychology, 61, 503–545.

Nishii, L. H., & Wright, P. M. (2008). Variability within organi- zations: Implications for strategic human resource man- agement. In D. B. Smith (Ed.), The people make the place (pp. 225–248). Mahwah, NJ: Erlbaum.

O’Reilly, C., & Pfeffer, J. (2000). Hidden value: How great companies achieve extraordinary results with ordinary people. Boston, MA: Harvard Business School Press.

Osterman, P. (1994). How common is workplace transforma- tion and who adopts it? Industrial and Labor Relations Review, 47, 173–188.

Peccei, R., Van de Voorde, K., & Van Veldhoven, M. (2012). HRM, well-being, and performance: A theoretical and empirical review. In J. Paauwe, D. E. Guest, & P. M.

Human Resource Management DOI: 10.1002/hrm

TOP MANAGEMENT HR VALUES 433

Wright (Eds.), HRM and performance: Achievements and challenges (pp. 15–45). West Sussex, England: Wiley.

Pfeffer, J. (1998). The human equation: Building profi ts by putting people fi rst. Boston, MA: Harvard Business School Press.

Pfeffer, J., & Viega, J.F. (1999). Putting people fi rst for organi- zational success. Academy of Management Executive, 13, 37–48.

Preacher, K. J., Rucker, D. D., & Hayes, A. F. (2007). Address- ing moderating mediation hypotheses: Theory, methods, and prescriptions. Multivariate Behavioral Research, 42, 185–227.

Ravlin, E. C., & Meglino, B. M. (1987). Effect of values on percep- tion and decision making: A study of alternative work value measures. Journal of Applied Psychology, 72, 666–673.

Riordan, C. M., Vandenberg, R. J., & Richardson, H. A. (2005). Employee involvement climate and organiza- tional effectiveness. Human Resource Management, 44, 471–488.

Rokeach, M. (1973). The nature of human values. New York, NY: Free Press.

Schuler, R. S. (1992). Strategic human resources: Linking the people with the strategic needs of the business. Organi- zation Dynamics, 21, 18–32.

Schwartz, S. H. (1992). Universals in the content and structure of values: Theoretical advances and empirical tests in 20 countries. In M. P. Zanna (Ed.), Advances in experimental social psychology (pp. 1–65). New York, NY: Academic Press.

Simons, T. (2002). Behavioral integrity: The perceived align- ment between managers’ words and deeds as a research focus. Organization Science, 13, 18–35.

Sobel, M. E. (1982). Asymptotic intervals for indirect effects in structural equations models. In S. Leinhard (Ed.), Sociological methodology (pp. 290–312). San Francisco, CA: Jossey-Bass.

Subramony, M. (2009). A meta-analytic investigation of the relationship between HRM bundles and fi rm perform- ance. Human Resource Management, 48, 745–768.

Takeuchi, R., Lepak, D. P., Wang, H., & Takeuchi, K. (2007). An examination of the mechanisms mediating between high-performance work systems and performance in Japanese organizations. Journal of Applied Psychology, 92, 1069–1083.

Thomas, J. B., Clark, S. M., & Gioa, D. A. (1993). Strategic sensemaking and organizatio nal performance: Linkages among scanning, interpretation, action, and outcomes. Academy of Management Journal, 36, 239–270.

Tsui, A. S., Pearce, J. L., Porter, L. W., & Tripoli, A. M. (1997). Alternative approaches to the employee-organization relationship: Does investment in employees pay off? Academy of Management Journal, 40, 1089–1121.

Walton, R. E. (1985). From control to commitment in the workplace. Harvard Business Review, 63, 77–84.

White, M., & Bryson, A. (2013). Positive employee attitudes: How much human resource management do you need? Human Relations, 66(3), 385–406.

Wright, P. M., & Boswell, W. R. (2002). Disagregating HRM: A review and synthesis of micro and macro human resource management research. Journal of Management, 28, 247–267.

Wright, P. M., Dunford, B. B., & Snell, S. A. (2001). Human resources and the resource based view of the fi rm. Jour- nal of Management, 27, 701–721.

Wright, P. M., Gardner, T. M., Moynihan, L. M., & Allen, M. R. (2005). The relationship between HR practices and fi rm performance: Examining causal ordering. Personnel Psy- chology, 58, 409–446.

Zacharatos, A., Barling, J., & Iverson, R.D. (2005). High per- formance work systems and occupational safety. Journal of Applied Psychology, 90(1), 77–93.

434 HUMAN RESOURCE MANAGEMENT, MAY–JUNE 2016

Human Resource Management DOI: 10.1002/hrm

A P P E N D I X A HPWS Program Itemsa

HPWS Program Survey Items Used to Assess Each Program Area

1. Autonomous job design

Employees in this area are given lots of opportunity to decide how to do their work. Employees perform simple and repetitive tasks as a major part of their work.(R) The actual job duties are shaped more by the employee than by a specifi c job description.

2. Internal promotions

Qualifi ed employees in this area have the opportunity to be promoted to posi- tions of greater responsibility within the company. Management vacancies in this area are fi lled with people already employed here. Internal candidates are given consideration over external candidates for job openings in this department.

3. Performance appraisals

Performance appraisals are used to plan skill development and training for future advancement with the hotel. Multiple employees provide input to the performance appraisal evaluation of employees in this area. Employees in these jobs regularly (at least once per year) receive a formal eval- uation of their performance.

4. Information sharing

Information about how the hotel is performing fi nancially is shared with these employees. Employees in this area are part of regular information sharing meetings. Strategy and goals are communicated primarily to managers in this area.(R)

5. Performance- based pay

Pay raises for employees in these jobs are based on performance. Employees in these jobs have the opportunity to earn individual bonuses for productivity, performance or other individual outcomes. Employee bonuses or incentive plans are based primarily on the performance of the organization.

6. Employee participation

Employees in these areas are involved in formal participation programs such as quality improvement programs, roundtable discussions and suggestion systems. We have formalized programs to encourage employee participation in this area. Mechanisms exist to consult employees in this area on the hotel’s business plan.

7. Employee training and development

There are formal training programs to teach new hires the skills they need to perform the job. We have a mentoring system to help develop these employees. Apart from on-the-job training, the hotel provides these employees with formal job training, either on or off the premises.

8 Employment security

Providing employment security for our employees is a priority. Job security is almost guaranteed to these employees. If the organization faces economic problems, employees in this area would be the last to get cut.

aItem response scale: “0=does not apply; 1=applies to few employees; 2=applies to half of the employees; 3=applies to most employees; 4=applies to all employees”; (R) indicates reversed scoring for item. Total score for this variable is the average of all items.

Human Resource Management DOI: 10.1002/hrm

TOP MANAGEMENT HR VALUES 435

A P P E N D I X A HPWS Value-based Beliefsa (Continued) Item response scale (1=strongly disagree…5=strongly agree) Average (S.D.)

1. My primary responsibility is to create an environment that maximizes employee productivity and contribution level.

4.40 (0.88)

2. As long as employees continue to do a good job, they can expect to stay here as long as they want.

4.27 (0.98)

3. One of my core responsibilities as a leader is to prepare others for future promotion. 4.14 (0.95)

4. Providing job security to employees is a key management responsibility. 3.81 (1.03)

5. I believe that it is management’s obligation to provide all employees with long- term career opportunities.

3.54 (1.12)

6. When making key decisions, I always seek the opinion and suggestions of the employees who will be affected by the decisions.

3.99 (0.97)

aRespondents are the group of top managers at each hotel.

A P P E N D I X B HPWS Perceptionsa

Item Factor 1 Factor 2

1. It is easy for employees to share their ideas with management. .306 .697

2. Employees are encouraged to participate in decisions that affect their work. .267 .718

3. Employees are not afraid to speak up if there is a problem. .217 .701

4. If a decision is being made in my department, everyone is involved. .294 .606

5. Employees have enough information to do their jobs well. .196 .711

6. Employees are given information about hotel goals and performance. .179 .704

7. This hotel provides job security. .692 .227

8. Employees feel secure in their positions at the hotel. .695 .159

9. This hotel hires people who will fi t in over the long-term. .637 .384

10. This hotel trains people to get them promoted. .734 .282

11. This hotel helps people with their careers. .757 .232

12. This hotel recruits people within the hotel for management positions. .653 .122

13. The pay system encourages employees to work hard. .702 .302

14. Employees at this hotel will earn more money if they do a good job. .665 .370

15. The hotel often takes the time to measure how well employees are doing their jobs.

.621 .316

aSurvey items completed by employees (n=1,355); Response scale—“1=strongly disagree…5=strongly agree” Items in bold are used to defi ne the Factor.