U9A1-24 - Contrast equity theory with expectancy theory and apply these concepts to your own organization. See Details Below.

HRM SYSTEMS FOR KNOWLEDGE

WORKERS: DIFFERENCES AMONG

TOP MANAGERS, MIDDLE

MANAGERS, AND PROFESSIONAL

EMPLOYEES

A C H I M K R A U S E R T

This theoretical study contributes to the debate in the fi eld of strategic HRM

on whether HRM systems should differ across employee groups within the

fi rm. It explores differences in the effi cacy of two HRM systems—high-in-

volvement HR and internal labor market systems—across top management

teams, middle managers, and professional employees. In addition, it ex-

amines factors that may prevent the adoption of effective HRM systems for

these groups of knowledge workers. A model is proposed depicting how four

job-level moderators infl uence the HRM-performance relationship and how

these moderators apply to each of the three employee groups. Short-term

performance pressures and favoritism in staffi ng are proposed as factors that

may hinder the adoption of effective HRM systems. Finally, HRM systems for

different employee groups are classifi ed based on their gains potential and

barriers to implementation, yielding four categories of strategic HRM activity

metaphorically labeled low-hanging fruits, high-hanging fruits, peanuts, and

barren land. © 2013 Wiley Periodicals, Inc.

Keywords: HRM systems, agency theory, knowledge workers, high- involvement HR, internal labor markets, strategic HRM

Correspondence to: Achim Krausert, Nottingham University Business School China, AB376, 199 Taikang East

Road, Ningbo 315100, China, Phone: +86 150 5885 9967, Fax: +86 574 8818 0125, E-mail: achim.krausert@ nottingham.edu.cn.

R ecently, researchers in strategic HRM have paid increasing atten- tion to differences in the efficacy of high-involvement HR (HIHR) sys- tems in improving the performance

of different employee groups. HIHR systems encompass practices such as employee

discretion, skill-based staffing, socialization, training, and group-level incentives (Batt & Colvin, 2011; McClean & Collins, 2011), and they impact on employee performance by enhancing the employees’ skills, motivation, and opportunities to contribute (Jiang et al., 2012). The research shows that HIHR is

Human Resource Management, January–February 2014, Vol. 53, No. 1. Pp. 67–87

© 2013 Wiley Periodicals, Inc.

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

DOI:10.1002/hrm.21554

68 HUMAN RESOURCE MANAGEMENT, JANUARY–FEBRUARY 2014

Human Resource Management DOI: 10.1002/hrm

ILM systems are

most effective

where employees

require firm-specific

skills and networks

to perform their job.

particularly effective if applied to knowledge workers (Lepak & Snell, 2002; McClean & Collins, 2011; Yan, Peng, & Francesco, 2011). Fewer studies examined differences in the actual adoption of HIHR across employee groups. They indicate that it is more likely to be adopted for knowledge workers than for manual and clerical employees (Kalleberg, Marsden, Reynolds, & Knoke, 2006; Lepak, Taylor, Tekleab, Marrone, & Cohen, 2007) but also that the extent of adoption for knowledge workers is smaller than expected given the evidence for its effectiveness (Lepak & Snell, 2002). This indicates that HIHR is not equally effective across all groups of knowledge workers and/or that other factors,

besides its effectiveness, influence adoption decisions. Research on factors influencing the adoption of HIHR found the HR philosophy of a firm’s top management to have a positive influence on its adoption among less skilled em- ployees (Baron & Hannan, 2002; Lepak et al., 2007). However, there is little knowledge or theory about factors that might explain why HIHR is not adopted for some knowledge workers.

Another recurring focus of research in strategic HRM are internal labor market (ILM) systems, encompassing practices such as internal staffing, employment security, and above market-level pay and benefits (Batt & Colvin, 2011; Shaw, Dineen, Fang, & Vellella, 2009). These systems reduce employee turnover and, thereby, direct costs of turnover as well as disruptions to perfor- mance and team processes (Hausknecht & Trevor, 2011; Kacmar, Andrews, Van Rooy, Steilberg, & Cerrone, 2006). ILM systems are most effective where employees require firm- specific skills and networks to perform their job (Hausknecht & Trevor, 2011; Hom et al., 2009). Despite evidence for their effective- ness, there has been a pronounced decline in the use of ILMs over the past two decades, also in managerial career tracks, where firm- specific human and social capital are arguably needed the most (Hamori & Kakarika, 2009; Kotter, 1982). Research on top managers

yielded conditions under which external staff- ing may be effective despite firm-specific cap- ital requirements, such as turbulent business environments, persistent firm performance problems, the presence of firm-specific capi- tal on the board, and time for external hires to develop the needed firm-specific capital— however, this research also showed that top management positions are staffed externally, too, where it does not seem to be in the inter- est of the company (Karaevli, 2007; Zhang & Rajagopalan, 2003). With regard to managers below the upper echelons of the firm, Hamori and Kakarika (2009) found that employers continue to value their firm-internal expe- rience, despite a generally increased career mobility also among managers. Interpreting their findings, Hamori and Kakarika sug- gested the effectiveness of internal versus external labor market strategies might vary between different groups of knowledge work- ers but did not elaborate on this suggestion. Further knowledge and, in the first instance, theory is required that specifies how and why the effectiveness of ILM systems might vary between different groups of knowledge work- ers and that identifies factors potentially pre- venting the adoption of ILM systems where they would be effective.

Understanding differences in the effec- tiveness of HRM systems across employee groups has been argued to be important for the strategic planning processes of HRM departments, enabling a more targeted allo- cation of effort and resources across HRM sys- tems for different employee groups (Lepak, Liao, Chung, & Harden, 2006; McClean & Collins, 2011). Hence, if there are significant differences in the effectiveness of HRM sys- tems across groups of knowledge workers, it is important to study them. Likewise, if there are any factors that might hinder the implementation of otherwise effective HRM systems for a group of knowledge workers, a better understanding of these factors would be needed to make decisions on whether to invest in the application of these HRM sys- tems to that group and, if they are invested in, to navigate barriers to their successful implementation. The argument is illustrated in Figure 1. If HRM departments spend time

HRM SYSTEMS FOR KNOWLEDGE WORKERS 69

Human Resource Management DOI: 10.1002/hrm

and resources on HRM systems that will likely yield only limited gains while they are diffi- cult to implement, they are, metaphorically speaking, “plowing the barren land.” HRM activities that promise to have a bigger impact may be said to target either the “high-hang- ing“ or the “low-hanging” fruits, depending on the extent of resistance to these activities. A fourth type of HRM activity may generate results that are easy to obtain but also of rela- tively limited value—“peanuts,” metaphori- cally speaking. To enable a strategic resource allocation by HRM departments, it is impor- tant to understand which of these categories different HRM systems fit into.

A perspective that has not been resorted to previously in order to explain differences in the effectiveness or adoption of HRM sys- tems across employee groups is agency theory. As the dominant theoretical perspective in the field of corporate governance (Daily, Dalton, & Cannella, 2003), agency theory is concerned with management choices motivated by pri- vate benefits rather than performance consider- ations (Shleifer & Vishny, 1997), as well as with management choices that bolster short-term performance to the detriment of longer-term performance (Edmans, 2009). One type of private benefit of managers related to HRM is arguably the ability to hire and promote their own social network contacts, which was sug- gested to affect the quality of staffing decisions (Bandiera, Barankay, & Rasul, 2009; Tyler,

2012). HRM practices are most effective if they are adopted as part of systems or bundles of HRM practices (Chadwick, 2010; Jiang et al., 2012; Lepak et al., 2006; Subramony, 2009) while rigorous skill-based staffing is consid- ered a central component of HIHR systems (Jiang et al., 2012; Lepak et al., 2007; McClean & Collins, 2011). Thus, if private benefits of managers did affect the quality of staffing deci- sions, it could be argued that agency problems influence the effectiveness of entire HIHR sys- tems, too. In the same vein, the effectiveness of ILM systems could be reduced if the hir- ing of personal network contacts outside the firm affected the internal staffing principle as a core component of ILM systems (Batt & Colvin, 2011; Osterman, 1987).

Moreover, HIHR was shown to impact on company performance with a time lag of several years (Birdi et al., 2008; Birkinshaw, Hamel, & Mol, 2008). If managers under short-term performance pressure considered the adoption of HIHR to be a long-horizon investment, that might be a factor explaining instances of nonadoption. However, before more definite statements can be made, it is necessary to think through in detail alterna- tive explanations for a partial adoption or nonadoption of seemingly effective HRM sys- tems, including the possibility that HIHR and ILM systems are not equally effective across all groups of knowledge workers.

This article studies differences in the HRM-performance link between employee groups within the category of knowledge workers. It aims to identify HRM systems the adoption of which would be in the com- pany’s interest—if they were fully imple- mented—while agency problems resulting from private benefits and short-term perfor- mance pressures affect their implementation. The category knowledge workers encompasses a diverse range of employee groups. Salient groups that are commonly distinguished in the management literature include mana- gerial and professional employees (Barker, 2010). The category of managerial employees is often further divided into top management teams (TMTs) and middle managers (MMs) (Raes, Heijltjes, Glunk, & Roe, 2011). TMTs encompass CEOs and executives reporting to

Barriers to Implementation of HRM System

Gains Potential of HRM System

High

Low

Low High

“Barren Land” “High-hanging

Fruits”

“Low-hanging

Fruits” “Peanuts”

FIGURE 1. Framework for Strategic Resource Allocation by HRM Departments

70 HUMAN RESOURCE MANAGEMENT, JANUARY–FEBRUARY 2014

Human Resource Management DOI: 10.1002/hrm

By specifying areas

of HRM that may

have a significant

impact on company

performance and

are, simultaneously,

associated with

agency risks, this

article should

also inform

future research

in corporate

governance.

CEOs (Lin & Shih, 2008). MMs are all manag- ers positioned below top managers and above first-line supervisors (Wooldridge, Schmid, & Floyd, 2008). Professional employees, such as engineers, lawyers, and accountants, perform technically complex tasks requiring a discrete body of specialized expertise that is acquired and certified at professional education insti- tutions (Barker, 2010; Freidson, 1986). Thus, this article will focus on differences in the HRM-performance link between TMTs, MMs, and professional employees as three salient groups of knowledge workers that are com- monly contrasted in the literature.

To identify HRM systems that would have a significant impact on company performance

if fully implemented while they are associated with agency risks, it is necessary to resort to, expand, and integrate theory from the strategic HRM, management, and corporate governance fields. The strategic HRM field has generated theory on HRM systems, how they impact on performance, and job-level mod- erators of the HRM-performance link. Previous research comparing the effectiveness of HRM systems between knowledge workers and less skilled workers has yielded job- level moderators such as the value (Lepak & Snell, 2002; McClean & Collins, 2011) and the firm specific- ity of the job (Lepak & Snell, 2002). This article will propose additional moderators as well as a compre- hensive model depicting how the different moderators influence the HRM-performance relation- ship. The model will be expanded

using knowledge from the management litera- ture about differences between the functions of TMTs, MMs, and professional employees so as to link the various moderators with each of the three employee groups. Finally, the cor- porate governance literature will yield factors that may result in an only partial adoption or nonadoption of effective management prac- tices. Building on the strategic HRM literature again, propositions will be made concerning the applicability of these factors to the HRM

systems proposed to be effective for each of the three employee groups. Thus, by speci- fying the gains potential and barriers to the implementation of HIHR and ILM systems for TMTs, MMs, and professional employees, the various HRM systems can be classified into the earlier introduced four categories for the strategic allocation of resources of the HRM department.

While this article primarily targets researchers and practitioners in strategic HRM, the interweaving of HRM and corpo- rate governance theory will generate insights that may also be of interest to researchers and practitioners in corporate governance. The corporate governance literature has identified the neglect of future-oriented human capi- tal investments as a critical agency problem (Edmans, 2011; Souder & Bromiley, 2012), which may even constitute a threat to the inter- national competitiveness of the U.S. industry (Porter, 1992). However, the corporate gov- ernance literature by itself does not contain the knowledge that would be required to fur- ther specify which HRM practices should be regarded as long-horizon investments while they are critical for sustainable company performance and to what extent their adop- tion is really affected by agency problems. As a result, the argument that human capital investments are affected by agency problems remains a somewhat abstract idea. By specify- ing areas of HRM that may have a significant impact on company performance and are, simultaneously, associated with agency risks, this article should also inform future research in corporate governance.

Effi cacy of HRM Systems Across Employee Groups

The impact of HRM practices on performance is greatest if they are adopted as part of sys- tems or bundles of complementary practices (Chadwick, 2010; Subramony, 2009). In their review of the HRM systems literature, Lepak et al. (2006) called for more research distin- guishing different HRM systems and their re- lationships with performance. This section is concerned with differences in the effective- ness of HIHR systems and ILM systems across

HRM SYSTEMS FOR KNOWLEDGE WORKERS 71

Human Resource Management DOI: 10.1002/hrm

HIHR is considered

an alternative

to the control-

oriented approach

to HRM, whereby

supervisors evaluate

and incentivize

performance

in relation to

performance output

and/or process

standards set

ex ante.

three groups of knowledge workers. The ex- tant literature usually does not distinguish HIHR and ILM systems but treats them as part of so-called high-performance work systems (Lepak et al., 2006). Studies that focus on ei- ther HIHR or ILM systems are part of two largely separate research streams. A limited number of studies have contrasted the effects of the two systems empirically. Batt and Colvin (2011) and Shaw et al. (2009), for ex- ample, compared the effects of what they referred to as high-involvement and invest- ment and inducement systems, the compo- nents of the latter being largely identical to those of ILM systems. A distinction of HIHR and ILM systems is particularly relevant in the context of the current article. While com- plementing each other (Lepak et al., 2006), they impact on company performance via two distinct main effects, entailing different job-level performance moderators and differ- ent agency problems. Treating ILM systems separately is warranted also because firm-in- ternal and -external mobility are more of an option for knowledge workers than for man- ual and clerical workers and, therefore, po- tentially more of an issue in the context of this particular category of employees (Hamori & Kakarika, 2009).

High-Involvement HR (HIHR) Systems

HIHR systems seek to “win the hearts and minds” of employees rather than merely to obtain their compliance with work rules set by their supervisors (Walton, 1985). They seek to direct employee performance behav- iors through employee involvement, identifi- cation with the organization, and informal or “clan” control rather than formal output and process standards (Ouchi, 1979). For the high- involvement approach to be effective, it was argued that HRM must simultaneously en- hance the employees’ abilities, motivation, and opportunities to contribute (Jiang et al., 2012). To that end, HIHR systems encompass employee discretion, problem-solving groups, and self-directed teams as well as rigorous (skill-based) staffing, socialization, training, competency appraisals, and incentives linked

to company, unit, or team performance (Batt & Colvin, 2011; Birdi et al., 2008; McClean & Collins, 2011).

HIHR is considered an alternative to the control-oriented approach to HRM, whereby supervisors evaluate and incentivize perfor- mance in relation to performance output and/or process standards set ex ante (Ouchi, 1977; Walton, 1985). Control-oriented HRM practices include employee monitoring, for- mal performance appraisals, and individual performance-related pay (Batt & Colvin, 2011; Shaw et al., 2009). The high-involve- ment and control-oriented approaches to HRM may be regarded as ends on a contin- uum (Lepak & Snell, 2002; Yan et al., 2011). Researchers often use one scale to evaluate the extent to which HIHR versus control-oriented HR systems are used (e.g., Lepak et al., 2007; Way, 2002). This study follows these researchers in that it examines the effectiveness of HIHR as an alternative to control- oriented HR systems.

To explore whether the effec- tiveness of HIHR systems differs across TMTs, MMs, and profes- sional employees, job complex- ity and task uncertainty will at this point be introduced as job- level moderators of the HIHR- performance link. Subsequently, research on the functions of TMTs, MMs, and professional employ- ees will be resorted to in order to assess to what extent these mod- erators apply to each of the three job categories. The choice of job complexity and task uncertainty as moderators was initially inspired by Tsui, Pearce, Porter, and Tripoli (1997), who stud- ied the effects of an HRM system that com- bined elements of HIHR and ILM systems. Although they did not study moderators of those effects, they mentioned the possibility that their HRM system could be less appropri- ate where “a performance contribution can be clearly defined and measured” and more appropriate given “informational and tech- nical complexity” as well as “a high degree

72 HUMAN RESOURCE MANAGEMENT, JANUARY–FEBRUARY 2014

Human Resource Management DOI: 10.1002/hrm

The

interconnectedness

of the management

function with the

markets and the

dynamics of these

markets lead to

a generally high

degree of task

uncertainty for

managers.

of environmental uncertainty and rapid change” (p. 1092). More recently, Yan et al. (2011) studied differences in the effectiveness of HIHR between knowledge workers and manual workers. Similar to Tsui et al. (1997), they surmised that HIHR is more effective for knowledge workers because knowledge work is less structurable, so that “autonomy seems required by nature” (p. 410).

The logic implied by Tsui et al. (1997) and Yan et al. (2011) corresponds to organiza- tional control theory (Ouchi, 1977, 1979). The gist of this theory is that the relative effective- ness of clan versus output and process control depends on the structurability of the job (see also Kirsch, Ko, & Haney, 2010). That is, clan

control is argued to be the prefer- able alternative if it is not feasible to define and evaluate compre- hensive performance output and/ or process standards. Ouchi pro- posed two factors that determine whether this is feasible: the avail- ability of crystallized goals (which, he argued, corresponds to the amount of exceptions occurring in a job) and knowledge of cause- effect relationships (correspond- ing to the extent to which the work process can be analyzed). Thus, it can be argued that the reason that Tsui et al.’s (1997) construct of environmental uncertainty may moderate the HIHR-performance link is that it entails task uncer- tainty. Task uncertainty relates to the amount of exceptions occur-

ring in a job and, consequently, the feasibil- ity of specifying needed performance outputs ex ante. Because the present article focuses on job-level differences in HIHR effective- ness, the construct task uncertainty rather than environmental uncertainty will be used in the following (allowing for differences in task uncertainty across jobs at a given level of envi- ronmental uncertainty). The other construct mentioned by Tsui et al. (1997)—informa- tional and technical complexity—can then be argued to moderate the HIHR-performance link because it relates (negatively) to knowl- edge of cause-effect relations (or the extent

to which the work process can be analyzed): It makes it difficult to prescribe and evaluate performance processes.

Differences in Job Complexity Among

TMTs, MMs, and Professional Employees

The jobs of TMTs, MMs, and professional em- ployees can all be characterized as complex. The jobs of the two manager categories are complex due to the interconnectedness of their jobs with choices of actors inside and outside their organization (Raes et al., 2011). They coordinate processes in social systems characterized by ambiguous cause-effect rela- tions (Drucker, 1977). They must process a large amount of information from various sources, where MMs are arguably more inter- nally focused while TMTs must process infor- mation from sources both inside and outside the firm (Mantere, 2008; Wooldridge et al., 2008). The complexity of the professional employee’s job, by contrast, stems not so much from its interconnectedness but from the technical intricacies of a more narrowly defined set of tasks (Barker, 2010). This com- plexity limits the ability of supervisors to di- rectly control professional work processes and it necessitates job autonomy as a central feature of professional work (Freidson, 1986).

Differences in Task Uncertainty Among

TMTs, MMs, and Professional Employees

The interconnectedness of the management function with the markets and the dynamics of these markets lead to a generally high de- gree of task uncertainty for managers (Conner, 2000; Raes et al., 2011). TMTs are more exter- nally and MMs more internally oriented. However, because the functions of TMTs and MMs are connected, both types of manage- ment job are characterized by a high degree of task uncertainty (Wooldridge et al., 2008). Managers are free to set their own agendas (Tengblad, 2006). They have ample discretion in deciding not only how to go about a task but also what tasks to attend to (Gibson & Birkinshaw, 2004). Professional employees also have autonomy with regard to their work processes (Ouchi, 1979). Compared to

HRM SYSTEMS FOR KNOWLEDGE WORKERS 73

Human Resource Management DOI: 10.1002/hrm

managers, however, they have less freedom in setting their own agendas: It is managers who control the number and types of cases they must handle (Freidson, 1986). The professionals’ contributions are finite and limited to a narrower subject area, constitut- ing inputs to the managers’ broader and more fluid function (Barker, 2010). It can be argued that, in comparison to managerial jobs, pro- fessional jobs involve less uncertainty with regard to the immediate job outputs they must deliver.

Differences in the Effectiveness of HIHR

Among TMTs, MMs, and Professional

Employees

HIHR systems yield the largest performance improvements where jobs are characterized by both complexity and task uncertainty, ren- dering jobs less structurable. If a job is charac- terized by complexity and lower levels of task uncertainty, it is more effective to set and evaluate clear, measurable targets in terms of performance outputs (Ouchi, 1977, 1979). Thus, in the absence of HIHR practices that win the employees’ hearts and minds, the work of professional employees may still be fairly effectively directed via output goals. In comparison, the setting and evaluation of (proximal) output goals can be argued to be less effective for the more fluid jobs of man- agers. At best, one can set (distal) outcome goals, which are affected by a number of fac- tors besides the actions of the manager, in- cluding the actions of other managers and employees in the organization as well as the markets and other external factors (Barker, 2010). Group-level incentives linked to distal performance outcomes, however, were ar- gued to be a component of clan control and HIHR systems (Ouchi, 1979; Snell, 1992). It can be proposed then that the gains resulting from HIHR systems are particularly large if they are applied to TMTs and MMs, larger than if applied to professional employees.

This proposition is further supported by literature suggesting that the work of profes- sional employees is guided by professional value systems, which are instilled at profes- sional education institutions (Barker, 2010;

Freidson, 1986; Ouchi, 1979). In other words, professional employees are directed by their own system of clan control, which extends beyond the influence of managers and indi- vidual organizations (Hendry, 2003; McClean & Collins, 2011; Mroczkowski & Pope, 1987; Osterman, 1987). Complementing output control with elements of clan control is fea- sible and may add to performance in complex jobs characterized by moderate levels of task uncertainty (Kirsch et al., 2010). However, the performance difference between (a) out- put control combined with the profession- als’ own value-based control system and (b) firm-level clan control (established through HIHR) may be argued to be relatively small at best—that is, smaller than the performance increment that may be obtained by applying HIHR to managerial employees.

Proposition 1a: High-involvement HR systems have a particularly large positive impact on the performance of TMTs and MMs, larger than their impact on the performance of professional em- ployees.

The Relationship Between Employee and

Company Performance

With regard to the objectives of this study, it is ultimately the impact of HRM on company performance that is of interest. Employee per- formance matters for company perfor- mance—that is a basic premise of HRM, which is also supported by empirical evidence (Kacmar et al., 2006). However, there may be factors moderating the relationship between employee performance and company perfor- mance. Hence, Proposition 1a does not neces- sarily entail that HIHR is also more effective in improving company performance if ap- plied to managerial employees than if applied to professional employees. A factor that may moderate the employee-company perfor- mance link is the strategic value of an em- ployee group. The strategic value of an em- ployee group was previously proposed as a moderator of the link between HIHR and company performance (Lepak & Snell, 2002; McClean & Collins, 2011). I propose that it moderates, more specifically, the relationship

74 HUMAN RESOURCE MANAGEMENT, JANUARY–FEBRUARY 2014

Human Resource Management DOI: 10.1002/hrm

between employee performance and com- pany performance. In other words, HIHR may be equally effective in improving the perfor- mance of individuals in high- and low-value jobs; however, the performance of those in high-value jobs makes a bigger difference with regard to company performance. The strategic value of an employee group, in turn, depends on both the value of the contribu- tions made by each individual employee and on how numerous the employees in the em- ployee group are: Theoretically, HIHR may yield a small but significant improvement to performance in a job that adds only a small but significant amount of value (such as a manufacturing job), but may nevertheless make a large difference for company perfor- mance if this job is performed by a large num- ber of employees who can collectively be subjected to an HIHR system. The argument is illustrated in Figure 2.

Lepak and Snell (2002) surveyed per- ceptions on the strategic value of different employee groups and found that MMs and professional employees were both seen to be of a high strategic value to the competitive advantage of their firm. TMTs were not cov- ered by this survey. TMTs constitute a small employee group in numbers, however, one that is seen to make a large difference to com- pany performance (Shleifer & Vishny, 1997). This issue deserves further research. However,

based on the evidence available to date, all three employee groups shall be assumed to be of high strategic value to the firm. This entails that, if the efficacy of HIHR in improving employee performance is high when applied to TMTs and MMs and relatively low when applied to professional employees, the rela- tive efficacies of HIHR across the employee groups are analogous in impacting on com- pany performance.

Proposition 1b: High-involvement HR systems have a particularly large positive impact on com- pany performance if applied to TMTs and MMs, larger than the impact they have if applied to pro- fessional employees.

Internal Labor Market (ILM) Systems

ILM systems aim to create job embeddedness and establish a social exchange relationship, making it costly for employees to change em- ployers both in material and psychological terms (Hom et al., 2009). The components of ILM systems include internal staffing, inter- nal mobility opportunities, employment se- curity, pay above the market level, pensions and other benefits, and procedural justice (Batt & Colvin, 2011; Shaw et al., 2009). ILM systems relate to company performance “di- rectly” via costs of turnover, including costs of recruitment, selection, and off-the-job

Employee

Performance

High-

involvement

HR

Job Value

Number of

Employees in

Job Category

Strategic Value

of Employee

Group

Job Complexity

Company

Performance

Task

Uncertainty

FIGURE 2. Job-Level Moderators of the Relationship Between High-Involvement HR and Performance

HRM SYSTEMS FOR KNOWLEDGE WORKERS 75

Human Resource Management DOI: 10.1002/hrm

Studies focusing on

the distinctiveness

of TMT and MM

functions provide

ground for arguing

that there may be

subtle differences

between the two

employee groups in

the degree to which

they require firm-

specific human and

social capital.

training, and “indirectly” via employee performance—that is, disruptions to produc- tion processes and team dynamics as a result of turnover (Hausknecht & Trevor, 2011; Kacmar et al., 2006; Ton & Huckman, 2008). The relationship between turnover and per- formance is stronger given high-involvement work organization (Ton & Huckman, 2008). Hence, ILM systems are seen to be comple- mentary to HIHR systems (Lepak et al., 2006).

The central job-level moderator of the rela- tionship between ILM systems and employee performance is the extent to which the job requires firm-specific human and social capi- tal (e.g., Hausknecht & Trevor, 2011; Hom et al., 2009; Lepak & Snell, 2002). That is, ILM systems affect performance at the employee level by reducing turnover-related perfor- mance disruptions, the magnitude of which relates to firm-specific capital requirements of the job. Thus, any differences in firm-specific capital requirements between the functions of TMTs, MMs, and professional employ- ees must be expected to entail differences in the efficacy of ILM systems across the three employee groups.

Firm-specific capital appears to be critical in managerial jobs (Kotter, 1982). Top man- agers studied by Conner (2000), for example, “shared a tacit assumption that, to be success- ful, managers needed to understand how the company really operated” (p. 150). Research on the careers of managers found that those who transfer between employers progress more slowly in their career (Hamori & Kakarika, 2009). Positive effects of external CEO succes- sion in turbulent environments and under- performing firms were found to be offset by lacking firm-specific human, social, and polit- ical capital (Karaevli, 2007). Managers were found to be the employee group most likely to be in ten or more years’ service in the same employment (Doogan, 2001). And Drucker (1977) argued that, in order to shape firm-spe- cific strategies and processes, managers need to understand the existing competencies, pro- cesses, and culture of a firm and they must be integrated in their firm’s network to be able to exercise any influence.

Studies focusing on the distinctiveness of TMT and MM functions provide ground

for arguing that there may be subtle differ- ences between the two employee groups in the degree to which they require firm-specific human and social capital. The TMT job is more focused on managing the entrepreneur- ial process, requiring more industry knowl- edge. The MM job is, in comparison, more focused on internal operations (Wooldridge et al., 2008). MMs need to be more intercon- nected within the organization; TMTs require networks both outside and inside the firm (Raes et al., 2011; Westphal & Zajac, 2001). If industry knowledge can be assumed to be more transferable than knowledge of a com- pany’s internal operations and if industry- wide social network contacts can be seen to be more transferable than intra- firm network contacts, it can be argued that the required mix of general and firm-specific capital leans, relatively speaking, more toward general capital at the TMT level and more toward firm-spe- cific capital at the MM level. It can then be proposed that costs of employee turnover and, conse- quently, opportunity costs of not implementing ILM systems are higher at the MM than at the TMT level.

Compared to the two manager groups, the jobs of professional employees can be argued to require less firm-specific human and social capital. The defining features of professionalism include control by professional institutions (rather than by employers) over a discrete body of professional knowledge and over a professional value sys- tem as well as an active, occupation-wide community (Barker, 2010; Freidson, 1986). As a result, professional expertise is relatively eas- ily transferable across firms (Song, Almeida, & Wu, 2003; Teece, 2003). Professional career paths are more likely to extend across firms than managerial career paths (Hendry, 2003; Osterman, 1987). Professional employees were found to be less likely in at least ten years’ employment with the same firm than managers (Doogan, 2001). And employment

76 HUMAN RESOURCE MANAGEMENT, JANUARY–FEBRUARY 2014

Human Resource Management DOI: 10.1002/hrm

security was found to be less important to them than other aspects of the employment relationship (Mroczkowski & Pope, 1987).

If professional jobs tend to be less firm- specific than managerial jobs, the impact of ILM systems on professional employee per- formance must be expected to be generally lower than their impact on the performance of managerial employees. To be sure, this does not rule out that ILM practices may have some positive effect if applied to professional employees, who may also require (a smaller amount of) firm-specific knowledge to per- form their jobs (Campbell, Coff, & Kryscynski, 2012). The proposition to be made, however, is that the efficacy of ILM systems is largest if applied to MMs, and larger if applied to TMTs than if applied to professional employees.

Proposition 2a: Internal labor market systems have a particularly large positive impact on the performance of MMs and they have a larger im- pact on the performance of TMTs than on the per- formance of professional employees.

As before, for the purposes of this article, it is the impact of HRM systems on company performance that is relevant. It was argued ear- lier that the relationship between employee and company performance is moderated by the strategic value of the employee group. If the strategic value of the three groups of knowledge workers is at a similar, high level (Lepak & Snell, 2002; Shleifer & Vishny, 1997), it can be argued that the relative effi- cacies of ILM systems across the three groups are analogous in impacting on individual and company performance. Besides their effect via employee performance, ILM systems may also impact on company performance via direct costs of turnover (Hausknecht & Trevor, 2011). To the extent that direct turnover costs such as costs of training and socialization cor- relate with firm-specific capital requirements of the job, they should add to the proposed job-level differences in the company perfor- mance effect of ILM systems.

Proposition 2b: Internal labor market systems have a particularly large positive impact on company performance if applied to MMs and they have a

larger positive impact on company performance if applied to TMTs than if applied to professional employees.

Agency Problems Related to HRM for Knowledge Workers

The subject of agency theory is management decisions that serve the managers’ own inter- est and not the interest of the firm. Managers are in a position to make such decisions be- cause, as an interest group in the firm, they understand their firm’s production function better than any other interest group, includ- ing the shareholders of the firm (as their em- ployers), and because their actions are largely not observable (Jensen & Meckling, 1976). To align their interest with the interest of the in- vestors, the compensation of managers is typ- ically linked to company performance through bonus and stock-based compensa- tion systems; investors also have the option to dismiss CEOs (Murphy, 1999, 2003). Despite that, it was shown that managers continue to make decisions deviating from the firm’s best interest, securing private ben- efits for themselves, including perquisites, pet projects, and professional status related to empire building (Shleifer & Vishny, 1997). Unless managers hold the entire stock of their firm, agency problems cannot be completely solved through incentive pay: Murphy (1999) estimated that, given average performance sensitivities of CEO compensation, $10 mil- lion of perquisite consumption costs the av- erage CEO only about $60,000 in incentive pay. A second type of agency problem that persists, and is potentially aggravated through management incentives, is what is referred to as managerial myopia—that is, sacrifice of in- vestments required to sustain performance in the longer term in order to bolster near-term performance (Edmans, 2011; Miller, 2002; Porter, 1992; Souder & Bromiley, 2012; Souder & Shaver, 2010).

The previous section examined differ- ences in the effectiveness of HIHR and ILM systems across TMTs, MMs, and professional employees. This section will explore the pos- sibility that the adoption of otherwise effec- tive HRM systems is hindered by factors in

HRM SYSTEMS FOR KNOWLEDGE WORKERS 77

Human Resource Management DOI: 10.1002/hrm

the domain of agency theory—in particular, short-term performance pressures and private benefits of managers.

Short-Term Performance Pressures

The evidence suggests that HIHR systems incur considerable implementation costs, while the manifestation of benefits lags be- hind their implementation for years. Kossek (1987) suggested that the implementation of HIHR causes temporary performance reduc- tions associated with changes to informal be- havioral routines (see also Lawler, 1986). Kozlowski and Klein (2000) argued the pace at which high-involvement work processes can emerge is additionally constrained by work flow and budget cycles. Longitudinal re- search by Birdi et al. (2008) revealed that em- powerment practices take one to four years and self-directed teamworking six to nine years to have an impact on company perfor- mance. Similarly, Birkinshaw et al. (2008) assumed behavioral effects of changes to management systems take “several years” until they become apparent (p. 837).

The corporate governance literature sug- gests that the neglect of long-horizon invest- ments constitutes an agency problem. Human capital investments are identified as critical long-horizon investments (Edmans, 2011; Souder & Bromiley, 2012), but it is not speci- fied what these human capital investments are. The arguments in this section suggest that the implementation of an HIHR system may represent an investment in the future human capital of a firm. Thus, to the extent that top managers are myopic, the benefits lag associated with HIHR systems can be a factor that reduces management support for them. This is not to state that TMTs generally have too short an investment horizon, prevent- ing the adoption of HIHR across most firms. Short-termism is a consequence of a number of factors, which vary across firms, including top management incentives but also perfor- mance problems threatening the survival of the company and flawed capital budgeting criteria (Miller, 2002). The proposition to be made is then that the time lag between the implementation and manifestation of

benefits of HIHR systems reduces the likeli- hood that these systems are adopted—how- ever, only to the extent that the management of a firm is subject to short-term performance pressures.

Proposition 3: Short-term performance pressures are negatively related to investments in high-in- volvement HR systems.

ILM systems do not effect changes in work processes and ingrained behaviors of employ- ees. Instead, they seek to embed employees in the firm, making it costly for employees in material and psychological terms to quit (Hom et al., 2009). Given high pay as well as pros- pects of internal mobility and a career within the firm, employees stand to gain less by turn- ing to the external labor market. Procedural justice (constituting another component of ILM systems) reduces the employees’ likeli- hood of quitting due to conflicts. The longer the employees’ tenure with a firm, the more internal network contacts and attachment they will develop, which further raises their cost of quitting (Hom et al., 2009). The effect may be further enhanced by social exchange perceptions: If employees perceive their employer to support their interests (for exam- ple, through high pay and the facilitation of a career in the firm), they may feel morally obliged to reciprocate by being loyal to their employer (Blau, 1964; Shaw et al., 2009).

To explore whether the adoption of ILM systems may be affected by short-term per- formance pressures, it is necessary to evalu- ate (a) the time lag between their adoption and their effect on turnover and (b) the time lag between the occurrence of turnover and its impact on company performance. First, ILM practices may be argued to impact on employee turnover in relatively immedi- ate terms after they are adopted. Once firms raise the pay of employees above the market level or create prospects of a career within the firm, they reduce the employees’ incentives to quit. Further, if employees receive what they perceive to be supportive treatment by their organization, they will likely not post- pone any reciprocating attitudes and behav- iors until years later but start displaying them

78 HUMAN RESOURCE MANAGEMENT, JANUARY–FEBRUARY 2014

Human Resource Management DOI: 10.1002/hrm

soon after they experienced the supportive treatment. These arguments are consistent with the (albeit limited) empirical evidence. Hom et al. (2009) conducted a longitudinal study in China suggesting that ILM systems have a same-year effect on quit intentions via job embeddedness and social exchange per- ceptions (where the effect via social exchange perceptions attenuated over time). Batt and Colvin (2011) also reported evidence consis- tent with a same-year effect of ILM practices on quit rates.

Second, concerning the time lag between the occurrence of turnover and its impact on company performance, Kacmar et al. (2006) found that turnover impacted on unit financial performance via (disruptions of) performance processes within one year. Besides near-term turnover effects via employee performance, direct costs of turnover (such as expenses related to recruitment, training, and social- ization) should also be reflected in the com- pany’s financial performance shortly after the occurrence of turnover (and beyond, depend- ing on the duration of recruitment, training, and socialization processes). Consistent with these arguments, Ton and Huckman (2008) found that turnover impacted on profit mar- gins within a period of three months. Hence, if ILM systems impact on employee turnover in the near term, they may also be argued to impact on company performance in relatively immediate ways. The temporal dynamics between the implementation of ILM systems and the manifestation of performance effects are likely such that they do not add to any agency risks related to the adoption of respec- tive HRM practices.

Proposition 4: Short-term performance pressures do not affect decisions to adopt HRM practices as- sociated with internal labor market systems.

Private Benefi ts

A second source of agency problems besides short-term performance pressures is private benefits of managers (Jensen & Meckling, 1976). Managers may make decisions moti- vated by private benefits despite stock-based compensation and other incentives if the

value of the private benefits exceeds the man- agers’ share in the value they create for the firm (Murphy, 1999). The research has been mostly focused on private benefits related to management compensation and perquisites (Shleifer & Vishny, 1997). However, recent contributions to agency theory in the eco- nomics literature suggest that favoritism in hiring and promotion decisions should also be regarded as a source of private benefits. Being in a position to favorably hire their own network contacts was argued to enable managers to engage in favor exchange (Ponzo & Scoppa, 2011). Favor exchange related to staffing was shown to help the careers of the individuals involved but have an overall neg- ative impact on the performance of the orga- nization (Bandiera et al., 2009; Ponzo & Scoppa, 2010). It was further argued that some managers may have a personal interest in hiring candidates of a lower ability so as to protect their own position and career pros- pects in the firm (Friebel & Raith, 2004). And it was argued that managers may derive pri- vate benefits from favorably evaluating and supporting the progress of those they have so- cial connections with (Prendergast & Topel, 1996). There is evidence to suggest that the influence of personal social networks on staff- ing decisions is stronger in less developed re- gions (Ponzo & Scoppa, 2010). However, a recent publication in the HRM practitioner lit- erature also claimed “rampant favoritism” in promotion decisions in the United States (Tyler, 2012, p. 79).

Rigorous staffing based on ability, includ- ing hiring and promotion decisions, is a central element of HIHR systems (Jiang et al., 2012; Lepak et al., 2007; McClean & Collins, 2011; Snell, 1992) while the effec- tiveness of HRM systems depends on the implementation of the various component practices (Chadwick, 2010; Lepak et al., 2006; Subramony, 2009). To the extent that favorit- ism in staffing affects the quality of staffing decisions—and the available evidence sug- gests it does (Bandiera et al., 2009)—it can then be argued that it may also affect the functioning of HIHR systems. The preva- lence of favoritism in staffing is influenced by factors above the firm level, such as the

HRM SYSTEMS FOR KNOWLEDGE WORKERS 79

Human Resource Management DOI: 10.1002/hrm

state of economic development, but there are also firm-level differences in norms and rules related to favoritism (Ponzo & Scoppa, 2010). Thus, it shall be proposed that HIHR systems tend to be implemented more com- prehensively and, therefore, tend to be more effective in firms in which favoritism is not tolerated.

Beyond rigorous staffing, HIHR may cause private costs in that its implementa- tion requires changes in behavioral routines and power relations. Employees were found to display resistance to such changes (Kossek, 1987; Lawler, 1986). Kossek (1989) found that employees in the category officers and manag- ers are more supportive of HRM innovations if they are applied to other employee groups than if they are applied to themselves. Hence, the continuity of existing work processes and (formal and informal) organizational struc- tures may, in addition to favoritism in staff- ing, also be regarded as a private benefit that may hinder the comprehensive implemen- tation of HIHR systems and, consequently, their effectiveness.

Proposition 5a: Private benefi ts are negatively related to the comprehensive implementation of high-involvement HR systems (and, consequently, to their effectiveness).

As the dominant coalition of the firm (Carpenter, Geletkanycz, & Sanders, 2004), TMTs are ultimately responsible for HRM practices applicable to themselves and other employee groups. It is the trade-off between their performance rewards and their private costs that may create agency problems related to HRM. Thus, private costs associated with HIHR systems, including those related to rigorous staffing and organizational change, may entail a nonadoption of HIHR sys- tems for TMTs. Beyond that, Edmans (2011) argued that TMTs may derive private benefits from raising other employees’ compensa- tion and incur private costs from bargaining employees down to their reservation wage. Extending this logic to other areas of HRM besides compensation, it is conceivable that enforcing the application of HIHR systems to other employee groups—especially to MMs, who are the most interconnected with TMTs (Raes et al., 2011)—may also incur a smaller private cost to TMTs in terms of confronta- tions, reduced loyalty, and reciprocation by the respective employee groups. However, it can be argued that this (smaller) cost is less likely to affect the adoption of HIHR for MMs and professional employees compared to the likelihood of an agency effect on the adop- tion of HIHR for TMTs (see Figure 3).

Company

Performance

Short-term

Performance

Pressures

HIHR System

ILM System

Private Benefits

TMT vs. Other

Employee

Groups

P3 (−)

P4 (0)

P5a (−)

P6a (−)

P5b (+)

P6b (+)

FIGURE 3. Agency Problems Related to HRM for Knowledge Workers

80 HUMAN RESOURCE MANAGEMENT, JANUARY–FEBRUARY 2014

Human Resource Management DOI: 10.1002/hrm

Proposition 5b: The (negative) relationship be- tween private benefi ts and the comprehensive im- plementation of HIHR systems is stronger at the TMT level than at the levels of MMs and profes- sional employees.

Internal mobility opportunities and hir- ing predominantly at the bottom of the job ladder are central components of ILM sys- tems (Batt & Colvin, 2011; Osterman, 1987). If employees derive private benefits from exchanging favors with social network con- tacts within as well as outside their firm, favoritism in staffing may relate negatively also to the comprehensive implementa- tion of ILM systems. This does not necessar- ily imply that the internal staffing principle must be enforced in absolute terms. Recent research on optimal levels of turnover found that some level of external staffing may be beneficial for the firm, although Hausknecht and Trevor (2011) concluded in their review of the literature on employee turnover that “a ‘lower turnover is better’ guideline will typi- cally serve both researchers and practitioners well” (p. 366). The optimal level of exter- nal staffing may differ between employee groups: Earlier, it was proposed that the TMT func tion is less firm-specific than the MM function and that the efficacy of ILM systems is lower at the TMT than at the MM level, although the differences were argued to be subtle (Proposition 2b). If there is some ben- efit to be obtained through external staffing, such as in dynamic business environments (Karaevli, 2007), if optimal levels of external staffing are a function of both the benefits and the costs of external staffing, and if the costs are related to the firm specificity of the job, then it might be argued that optimal ratios of external over internal staffing are higher at the TMT than at the MM level. Thus, it shall be proposed that optimal levels of external staffing, which may differ between employee groups, are more likely to be exceeded where favoritism in staffing decisions is tolerated, hindering the effectiveness of ILM systems.

Proposition 6a: Private benefi ts are negatively re- lated to the comprehensive implementation of ILM systems (and, consequently, their effectiveness).

The final proposition is that excessive external hiring is more likely to occur at the TMT level than at the MM or professional employee levels. TMTs are part of a close-knit social network that extends beyond firms and guards the interests of its members via strict norms and social sanctions (Westphal & Zajac, 2001). MM networks are, in com- parison, more internally oriented (Raes et al., 2011). Their private incentive to hire exter- nally in order to provide jobs to members of their social networks may be smaller than that of top managers. Additionally, as the dominant faction in the firm, only TMTs are in a position to insist on practices that are in their but not the other stakeholders’ interest (Carpenter et al., 2004). These arguments are supported by the empirical research insofar as top management positions were found to be staffed externally not only when it is in the interest of the firm (Hamori & Kakarika, 2009; Karaevli, 2007; Zhang & Rajagopalan, 2003). Further empirical research will be needed comparing optimal and actual levels of external staffing among different employee groups. However, based on these arguments, it shall be proposed that excessive external staffing motivated by private benefits is more likely to be observed at the TMT level than at the MM and professional employee levels.

Proposition 6b: The (negative) relationship be- tween private benefi ts and the comprehensive implementation of ILM systems is stronger at the TMT level than at the levels of MMs and profes- sional employees.

Discussion

The propositions of this article are summa- rized in Table I. The table provides a basis for the strategic allocation of attention and re- sources of HRM departments. It shows HRM systems that can be thought of as “low-hang- ing fruits” in the sense that they are effective and not associated with agency problems hindering their implementation—for exam- ple, ILM systems for MMs. Vice versa, some HRM systems are associated with agency risks while the opportunity costs of not (or not ef- fectively) implementing them are relatively

HRM SYSTEMS FOR KNOWLEDGE WORKERS 81

Human Resource Management DOI: 10.1002/hrm

low—for example, HIHR systems for profes- sional employees. HRM departments that di- rect their activities at such systems may be said to be “plowing the barren land.” HIHR systems for TMTs represent an example of “high-hanging fruits.” They were proposed to have a high gains potential that is difficult to realize due to agency problems. And ILM systems for professional employees are pre- dicted to yield smaller gains relatively easily, fitting in the category that was metaphori- cally labeled “peanuts” earlier.

This article offers a multifaceted expla- nation for the finding that seemingly effec- tive HRM systems are not as widely adopted for knowledge workers as one would expect (Lepak & Snell, 2002). First, both HRM sys- tems can be argued to make less of a differ- ence when applied to professional than when applied to managerial employees. A limited application of HRM systems to professional employees (McClean & Collins, 2011) may then be the result of HRM departments focus- ing their finite resources on those activities that have the greatest impact. Second, a non- adoption of HIHR systems for TMTs may be

the result of agency problems. Agency prob- lems may also affect the effectiveness of HIHR systems if they are adopted for TMTs or MMs (which may further reduce the incentive to dedicate resources to them). Third, an only partial adoption of ILM systems for TMTs (Hamori & Kakarika, 2009) may be explained by a combination of agency problems and moderate levels of effectiveness.

Some of the factors that may contrib- ute to the nonadoption of effective HRM practices are situated above the level of the firm—in particular, stock-based CEO incen- tives fostering short-termism. Corporate gov- ernance scholars argue that, for stock-based CEO incentives to be effective, the stock price needs to reflect long-horizon investments, including investments in human capital (Souder & Bromiley, 2012; Souder & Shaver, 2010). To that end, they suggest that securities analysts, as central actors in the stock market, should attend to such investments, improv- ing the accuracy of their long-term earnings forecasts and guiding investors more effec- tively (Daily et al., 2003; Edmans, 2011). The present article proposed HRM activities that

T A B L E 1 Summary of Propositions

High-Involvement HR Systems Internal Labor Market Systems

Top Management

Teams

Gains Potential Gains Potential

Effi cacy of HRM Practices (P1) *** Effi cacy of HRM Practices (P2) **

Risk of Agency Problems Risk of Agency Problems

Time Lag of Benefi ts (P3) YES Time Lag of Benefi ts (P4) NO

Private Costs (P5) *** Private Costs (P6) ***

Middle Managers Gains Potential Gains Potential

Effi cacy of HRM Practices (P1) *** Effi cacy of HRM Practices (P2) ***

Risk of Agency Problems Risk of Agency Problems

Time Lag of Benefi ts (P3) YES Time Lag of Benefi ts (P4) NO

Private Costs (P5) * Private Costs (P6) *

Professional

Employees

Gains Potential Gains Potential

Effi cacy of HRM Practices (P1) * Effi cacy of HRM Practices (P2) *

Risk of Agency Problems Risk of Agency Problems

Time Lag of Benefi ts (P3) YES Time Lag of Benefi ts (P4) NO

Private Costs (P5) * Private Costs (P6) *

*denotes “small in comparison”; **denotes “moderate”; and ***denotes “large in comparison”

82 HUMAN RESOURCE MANAGEMENT, JANUARY–FEBRUARY 2014

Human Resource Management DOI: 10.1002/hrm

Although difficult

to obtain, it might

be argued that the

high-hanging fruits,

if obtained, yield a

more sustainable

competitive

advantage to the

firm than the low-

hanging fruits—that

is, an advantage that

is simultaneously

valuable, rare, and

difficult to imitate

or substitute for

competitors.

have a significant impact on company per- formance and are simultaneously associated with agency risks. It is those kinds of activi- ties that securities analysts should attend to for stock-based CEO compensation to incen- tivize sustainable performance.

In terms of firm-level normative impli- cations of this article, one may argue that firms and their HR managers should priori- tize activities targeted at “low-hanging fruits” over activities targeted at the “peanuts” cat- egory, while they will usually want to avoid “plowing the barren land.” Directing atten-

tion and resources to HRM systems associated with both a high gains potential and agency problems— the “high-hanging fruits”—may be perceived as risky. Although dif- ficult to obtain, it might be argued that the high-hanging fruits, if obtained, yield a more sustainable competitive advantage to the firm than the low-hanging fruits—that is, an advantage that is simultane- ously valuable, rare, and difficult to imitate or substitute for competitors (Lado & Wilson, 1994). Whether and under what circumstances a firm should make those potentially risky investments is a subject for further theorizing and research. However, one more implication of this article is that, if resources are dedicated to HRM systems associated with both a high gains potential and agency risks, it is critical that the respec- tive activities are championed by the firm’s TMT. In particular, given investments in HIHR systems, TMTs should support rigorous, skill-based staffing—for example, by foster- ing norms and sanctions against favoritism in staffing. Given invest- ments in ILM systems, they should

limit experienced hiring to specific, defined circumstances.

Contributions to Theory

This study made four contributions to the body of theory on HRM systems and

their relationship with performance. In the first instance, it contributed theory on job- level moderators of the HRM-performance re- lationship. Few of the existing studies on HRM differences across employee groups sys- tematically examined job-level moderators beyond distinctions of job categories such as those of knowledge workers and manual workers (Tsui et al., 1997; Yan et al., 2011). McClean and Collins (2011), for example, fo- cused on the strategic value of an employee group as a single moderating factor. A semi- nal study by Lepak and Snell (2002) distin- guished two job-level moderators—the strategic value and the firm specificity of human capital—which they related to the ef- fectiveness of a high-performance HR system (combining elements of HIHR and ILM sys- tems). The present study distinguished four job-level moderators of the HRM-performance relationship, including job complexity, task uncertainty, firm-specific human and social capital requirements, and the strategic value of the employee group, as well as two factors at the firm level that may, depending on the job, hinder the adoption of individual HRM practices (and thereby the effectiveness of HRM systems). It suggested detailed explana- tions of how these factors differentially influ- ence the effectiveness of two related but dif- ferent HRM systems. Thereby, it also heeded to Lepak et al.’s (2006) call for more research on different HRM systems and their relation- ships with performance.

As a second contribution, literature on the functions of TMTs, MMs, and professional employees was integrated with this model of job-level moderators, enabling propositions on the job-specific effectiveness of HRM sys- tems. Previously, Lepak and Snell (2002) sur- veyed perceptions on the strategic value and firm specificity of a wider range of jobs, pro- viding an empirical classification. The pres- ent study explored theoretically how three job categories relate to a range of moderators of the HRM-performance relationship. This constitutes an advancement that should add to the practical relevance and applicability of this strand of research.

As a third contribution, this article integrated agency theory with HRM systems

HRM SYSTEMS FOR KNOWLEDGE WORKERS 83

Human Resource Management DOI: 10.1002/hrm

This article

proposed barriers

to implementation

as a second

dimension besides

the efficacy of

HRM systems to be

taken into account

in prioritizing HRM

activities.

theory, suggesting factors that may not only hinder the effectiveness of HRM systems but also result in their nonadoption. The strate- gic HRM literature, to date, has largely been concerned with the effectiveness of HRM systems. Factors that may influence adop- tion decisions have been studied less com- monly, such as the HR philosophy of the firm’s founder and the HR philosophy of the firm’s current TMT (Baron & Hannan, 2002; Lepak et al., 2007). The present article stud- ied the influence of factors in the domain of agency theory. Agency theory has previously received little attention in the strategic HRM literature. A recent exception is Caza’s (2011) research on the effect of MM job discretion on MM performance in order to evaluate compet- ing hypotheses derived from agency theory, ecology theory, and strategic choice theory. His agency-theoretical hypothesis was that job dis- cretion has a negative effect on performance, indicating opportunism, which was rejected by the data. The present article, by contrast, examined how agency problems may affect the implementation of HRM systems, includ- ing HIHR systems, of which job discretion is a central feature. According to this perspective, a positive effect of MM job discretion on MM performance would be expected to the extent that an HIHR system is implemented.

The article’s fourth contribution to theory is a framework for strategically prioritizing the activities of HRM departments. McClean and Collins (2011) and Lepak et al. (2006) suggested that HRM departments should take into account differences in the efficacy of HRM systems across employee groups in directing their efforts and resources. This article proposed barriers to implementation as a second dimension besides the efficacy of HRM systems to be taken into account in pri- oritizing HRM activities. Integrating the two dimensions, it proposed four categories, clas- sifying HRM systems according to their gains potential and agency risks that may hinder their implementation.

Future Research

Future empirical research may evaluate the effects of the proposed moderators, including

effects on the relationships between the two HRM systems and employee performance, be- tween the HRM systems and company perfor- mance, and between employee and company performance. It may test the propositions made with regard to differences in the effec- tiveness of the two HRM systems across the three groups of knowledge workers and it may study the job characteristics that were proposed to explain differences in effective- ness. One could study whether tolerance of favoritism in staffing moderates the effective- ness of the two HRM systems across the dif- ferent employee groups and whether it affects adoption decisions. One could study whether short-term performance pressures reduce the effectiveness and the likelihood of adoption of the two HRM systems. Empirical researchers might also study optimal levels of external staffing: If optimal levels of exter- nal staffing are greater than zero, it will be important to understand what the optimal levels are for dif- ferent employee groups, firms, and industries and to what extent deviations from them have an im- pact on ILM systems and com- pany performance.

Future research should also evaluate the arguments concern- ing the prioritization of HRM activities. One could empirically classify various targets of HRM activity as high- and low-hanging fruits, peanuts, and barren land. One could study whether the allo- cation of resources of the HRM department across the four categories relates to company performance. And one could explore fur- ther whether and under what circumstances the targeting of high-hanging fruits pro- vides a sustainable competitive advantage to the firm. Finally, an issue that could not be included in the already broad scope of this article is whether the differential treatment of employee groups in a firm has any negative effects, offsetting positive effects of a more targeted use of resources. This issue also rep- resents a subject for further theoretical and empirical study.

84 HUMAN RESOURCE MANAGEMENT, JANUARY–FEBRUARY 2014

Human Resource Management DOI: 10.1002/hrm

Concluding Comments

The intention of this article was to develop theory of practical relevance, both for HR managers and actors in corporate governance. Ideally, this theory will help HR managers to make better use of the resources available to them. It offers them ways of thinking about HRM systems that will hopefully enable them to contribute to a more sustainable perfor- mance of their organizations. To the extent that the article informs actors in corporate

governance, it may also facilitate control over firms that is both more effective and makes better use of the time of everyone involved in the corporate governance process.

Acknowledgments

I gratefully acknowledge the excellent sup- port of Clint Chadwick throughout the revi- sion process. I am equally grateful for the in- sightful comments and suggestions of two anonymous reviewers.

ACHIM KRAUSERT is an assistant professor of human resource management at the Nottingham University Business School China. He earned his PhD from the University

of Mannheim in Germany. His current research focuses on strategic human resource

management, with an emphasis on HRM systems and HRM differences across employee

groups, HRM agency problems, performance management and evaluation, and com-

parative employment relations.

References Bandiera, O., Barankay, I., & Rasul, I. (2009). Social

connections and incentives in the workplace:

Evidence from personnel data. Econometrica, 77,

1047–1094.

Barker, R. (2010). No, management is not a profession.

Harvard Business Review, 88(7–8), 52–60.

Baron, J. N., & Hannan, M. T. (2002). Organizational

blueprints for success in high-tech start-ups:

Lessons from the Stanford project on emerging

companies. California Management Review, 44(3),

8–36.

Batt, R., & Colvin, A. J. S. (2011). An employment sys-

tems approach to turnover: Human resources prac-

tices, quits, dismissals, and performance. Academy

of Management Journal, 54, 695–717.

Birdi, K., Clegg, C., Patterson, M., Robinson, A., Stride,

C. B., Wall, T. D., & Wood, S. J. (2008). The impact

of human resource and operational management

practices on company productivity: A longitudinal

study. Personnel Psychology, 61, 467–501.

Birkinshaw, J., Hamel, G., & Mol, M. J. (2008).

Management innovation. Academy of Management

Review, 33, 825–845.

Blau, P. (1964). Exchange and power in social life. New

York, NY: Wiley.

Campbell, B., Coff, R., & Kryscynski, D. (2012).

Rethinking sustained competitive advantage from

human capital. Academy of Management Review,

37, 376–395.

Carpenter, M. A., Geletkanycz, M. A., & Sanders,

G. W. (2004). Upper echelons research revisited:

Antecedents, elements, and consequences of

top management team composition. Journal of

Management, 30, 749–778.

Caza, A. (2011). Testing alternate predictions for the

performance consequences of middle managers’

discretion. Human Resource Management, 50, 9–28.

Chadwick, C. (2010). Theoretic insights on the nature of

performance synergies in human resource sys-

tems: Toward greater precision. Human Resource

Management Review, 20(2), 85–101.

Conner, J. (2000). Developing the global leaders of

tomorrow. Human Resource Management, 39,

147–157.

Daily, C. M., Dalton, D. R., & Cannella, A. A. (2003).

Corporate governance: Decades of dialogue

and data. Academy of Management Review, 28,

371–382.

Doogan, K. (2001). Insecurity and long-term unem-

ployment. Work, Employment and Society, 15,

419–441.

HRM SYSTEMS FOR KNOWLEDGE WORKERS 85

Human Resource Management DOI: 10.1002/hrm

Drucker, P. F. (1977). Management. London, UK: Pan.

Edmans, A. (2009). Blockholder trading, market effi -

ciency, and managerial myopia. Journal of Finance,

64, 2481–2513.

Edmans, A. (2011). Does the stock market fully value

intangibles? Employee satisfaction and equity pric-

es. Journal of Financial Economics, 101, 621–640.

Freidson, E. (1986). Professional powers: A study

of the institutionalization of formal knowledge.

Chicago, IL: The University of Chicago Press.

Friebel, G., & Raith, M. (2004). Abuse of authority

and hierarchical communication. Rand Journal of

Economics, 35, 224–244.

Gibson, C. B., & Birkinshaw, J. (2004). The anteced-

ents, consequences, and mediating role of organi-

zational ambidexterity. Academy of Management

Journal, 47, 209–226.

Hamori, M., & Kakarika, M. (2009). External labor

market strategy and career success: CEO careers

in Europe and the United States. Human Resource

Management, 48, 355–378.

Hausknecht, J. P., & Trevor, C. O. (2011). Collective

turnover at the group, unit, and organizational

levels: Evidence, issues, and implications. Journal

of Management, 37, 352–388.

Hendry, C. (2003). Applying employment systems

theory to the analysis of national models of

HRM. International Journal of Human Resource

Management, 14, 1430–1442.

Hom, P. W., Tsui, A. S., Wu, J. B., Lee, T. W., Zhang, A.

Y., Fu, P. P., & Li, L. (2009). Explaining employment

relationships with social exchange and job em-

beddedness. Journal of Applied Psychology, 94,

277–297.

Jensen, M. C., & Meckling, W. H. (1976). Theory of the

fi rm: Managerial behavior, agency costs and own-

ership structure. Journal of Financial Economics, 3,

305–360.

Jiang, K., Lepak, D. P., Han, K., Hong, Y., Kim, A., &

Winkler, A. (2012). Clarifying the construct of hu-

man resource systems: Relating human resource

management to employee performance. Human

Resource Management Review, 22(2), 73–85.

Kacmar, K. M., Andrews, M. C., Van Rooy, D. L.,

Steilberg, K. C., & Cerrone, S. (2006). Sure every-

one can be replaced . . . but at what cost? Turnover

as a predictor of unit-level performance. Academy

of Management Journal, 49, 133–144.

Kalleberg, A. L., Marsden, P. V., Reynolds, J., & Knoke,

D. (2006). Beyond profi t? Sectoral differences

in high performance work practices. Work and

Occupations, 33, 271–302.

Karaevli, A. (2007). Performance consequences of new

CEO “outsiderness”: Moderating effects of pre- and

post-succession contexts. Strategic Management

Journal, 28, 681–706.

Kirsch, L. J., Ko, D., & Haney, M. H. (2010).

Investigating the antecedents of team-based

clan control: Adding social capital as a predictor.

Organization Science, 21, 469–489.

Kossek, E. E. (1987). Human resources management in-

novation. Human Resource Management, 26, 71–91.

Kossek, E. E. (1989). The acceptance of human

resource innovation by multiple constituencies.

Personnel Psychology, 42, 263–281.

Kotter, J. (1982). What effective general managers re-

ally do. Harvard Business Review, 60(6), 156–167.

Kozlowski, S. W. J., & Klein, K. J. (2000). A multilevel

approach to theory and research in organizations:

Contextual, temporal, and emergent processes. In

K. J. Klein & S. W. J. Kozlowski (Eds.), Multilevel

theory, research, and methods in organizations:

Foundations, extensions, and new directions (pp.

3–90). San Francisco, CA: Jossey-Bass.

Lado, A. A., & Wilson, M. C. (1994). Human resource

systems and sustained competitive advantage:

A competency-based perspective. Academy of

Management Review, 19, 699–727.

Lawler, E. E. (1986). High involvement management.

San Francisco, CA: Jossey-Bass.

Lepak, D. P., Liao, H., Chung, Y., & Harden, E. E. (2006).

A conceptual review of human resource man-

agement systems in strategic human resource

management research. Research in Personnel and

Human Resources Management, 25, 217–271.

Lepak, D. P., & Snell., S. A. (2002). Examining the

human resource architecture: The relationships

among human capital, employment, and human

resource confi gurations. Journal of Management,

28, 517–533.

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 support employees. Human Resource

Management, 46, 223–246.

Lin, H., & Shih, C. (2008). How executive SHRM sys-

tem links to fi rm performance: The perspectives of

upper echelon and competitive dynamics. Journal

of Management, 34, 853–881.

Mantere, S. (2008). Role expectations and middle

manager strategic agency. Journal of Management

Studies, 45, 294–316.

McClean, E., & Collins, C. J. (2011). High-commitment

HR practices, employee effort, and fi rm

86 HUMAN RESOURCE MANAGEMENT, JANUARY–FEBRUARY 2014

Human Resource Management DOI: 10.1002/hrm

performance: Investigating the effects of HR prac-

tices across employee groups within professional

service fi rms. Human Resource Management, 50,

341–363.

Miller, K. D. (2002). Knowledge inventories and mana-

gerial myopia. Strategic Management Journal, 23,

689–706.

Mroczkowski, T., & Pope, J. A. (1987). Professionals

in two employment systems. Industrial Relations

Journal, 18, 284–297.

Murphy, K. J. (1999). Executive compensation. In O.

Ashenfelter & C. Card (Eds.), Handbook of labor

economics (Vol. 3b, pp. 2485–2563). Amsterdam,

the Netherlands: North Holland Publishers.

Murphy, K. J. (2003). Stock-based pay in new econo-

my fi rms. Journal of Accounting and Economics,

34(1–3), 129–147.

Osterman, P. (1987). Choice of employment systems in

internal labor markets. Industrial Relations, 26(1),

46–67.

Ouchi, W. G. (1977). The relationship between or-

ganizational structure and organizational control.

Administrative Science Quarterly, 22, 95–113.

Ouchi, W. G. (1979). A conceptual framework for the

design of organizational control mechanisms.

Management Science, 25, 833–848.

Ponzo, M., & Scoppa, V. (2010). The use of informal

networks in Italy: Effi ciency or favoritism? Journal

of Socio-Economics, 39, 89–99.

Ponzo, M., & Scoppa, V. (2011). A simple model of

favouritism in recruitment. Research in Economics,

65(2), 78–88.

Porter, M. (1992). Capital disadvantage: America’s fail-

ing capital investment system. Harvard Business

Review, 70(5), 65–82.

Prendergast, C., & Topel, R. H. (1996). Favoritism in

organizations. Journal of Political Economy, 104,

958–978.

Raes, A. M. L., Heijltjes, M. G., Glunk, U., & Roe, R. A.

(2011). The interface of the top management team

and middle managers: A process model. Academy

of Management Review, 36, 102–126.

Shaw, J. D., Dineen, B. R., Fang, R., & Vellella, R. F.

(2009). Employee-organization exchange rela-

tionships, HRM practices, and quit rates of good

and poor performers. Academy of Management

Journal, 52, 1016–1033.

Shleifer, A., & Vishny, R. W. (1997). A survey of corpo-

rate governance. Journal of Finance, 52, 737–783.

Snell, S. A. (1992). Control theory in strategic human

resource management: The mediating effect of ad-

ministrative information. Academy of Management

Journal, 35, 292–327.

Song, J., Almeida, P., & Wu, G. (2003). Learning-by-

hiring: When is mobility more likely to facilitate in-

terfi rm knowledge transfer? Management Science,

49, 351–365.

Souder, D., & Bromiley, P. (2012). Explaining tempo-

ral orientation: Evidence from the durability of

fi rms’ capital investments. Strategic Management

Journal, 33, 550–569.

Souder, D., & Shaver, J. M. (2010). Constraints and

incentives for making long horizon corporate

investments. Strategic Management Journal, 31,

1316–1336.

Subramony, M. (2009). A meta-analytic investigation

of the relationship between HRM bundles and fi rm

performance. Human Resource Management, 48,

745–768.

Teece, D. J. (2003). Expert talent and the design

of (professional services) fi rms. Industrial and

Corporate Change, 12, 895–916.

Tengblad, S. (2006). Is there a “new managerial

work”? A comparison with Henry Mintzberg’s clas-

sic study 30 years later. Journal of Management

Studies, 43, 1437–1461.

Ton, Z., & Huckman, R. S. (2008). Managing the impact

of employee turnover on performance: The role of

process conformance. Organization Science, 19,

56–68.

Tsui, A. S., Pearce, J. L., Porter, L. W., & Tripoli, A. M.

(1997). Alternative approaches to the employee-

organization relationship: Does investment in em-

ployees pay off? Academy of Management Journal,

40, 1089–1121.

Tyler, K. (2012, June). Undeserved promotions: A re-

searcher uncovers evidence of rampant favoritism

in promotion decisions. HR Magazine, p. 79.

Walton, R. E. (1985). From control to commitment in

the workplace. Harvard Business Review, 63(2),

77–84.

Way, S. A. (2002). High performance work systems

and intermediate indicators of fi rm performance

within the U.S. small business sector. Journal of

Management, 28, 765–785.

Westphal, J. D., & Zajac, E. J. (2001). Decoupling

policy from practice: The case of stock repurchase

programs. Administrative Science Quarterly, 46,

202–228.

HRM SYSTEMS FOR KNOWLEDGE WORKERS 87

Human Resource Management DOI: 10.1002/hrm

Wooldridge, B., Schmid, T., & Floyd, S. W. (2008). The

middle management perspective on strategy pro-

cess: Contributions, synthesis, and future research.

Journal of Management, 34, 1190–1221.

Yan, M., Peng, K. Z., & Francesco, A. M. (2011). The

differential effects of job design on knowledge

workers and manual workers: A quasi-experi-

mental fi eld study in China. Human Resource

Management, 50, 407–424.

Zhang, Y., & Rajagopalan, N. (2003). Explaining new

CEO origin: Firm versus industry antecedents.

Academy of Management Journal, 46, 327–338.

Copyright of Human Resource Management is the property of John Wiley & Sons, Inc. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.