معتبرساختن ساختار سیستم منابع انسانی: رویکرد استراتژیکی HRM بر اساس رده بندی

Despite considerable empirical evidence that human resource systems are related to firm performance, significant conceptual and methodological issues remain. We argue that progress along these lines requires basic development of the validity of the HR system construct. To help accomplish this, we draw on previous strategic HRM literature and insights from organizational levels of analysis literature to explicate a levels-based HR systems approach for validating the HR systems structure. We propose that conceptual and levels-based distinctions between HR system structure components can shed light on current methodological debates in the field, and offer an expanded strategic HRM research agenda.

Despite growing empirical evidence that systems of “progressive” or “high performance” human resource practices are related to firm performance outcomes, persistent conceptual and methodological issues remain concerning the size and nature of this relationship (Becker and Gerhart, 1996, Cappelli and Newmark, 2001, Delery, 1998, Gerhart, 1999, Gerhart et al., 2000, Guest et al., 2003, Wall and Wood, 2005, Wright et al., 2005 and Wright et al., 2001). Perhaps the most significant conceptual issue involves understanding how the central construct in this literature, the human resource system, affects firm performance outcomes ( Becker and Huselid, 2006, Bowen and Ostroff, 2004, Evans and Davis, 2005 and Ferris et al., 1999). One controversy involving this issue is whether there exists a set of “best practices” that are universally effective across firm contexts and industries ( Becker and Gerhart, 1996, Datta et al., 2005, Delery and Doty, 1996 and Lepak et al., 2006). Although it is possible to identify a set of high performance work practices ( Pfeffer, 1994), empirical studies have found significant firm performance effects using different combinations of HR practices ( Becker & Gerhart, 1996). This finding of ‘equifinality’ (i.e., different HR practices are associated with similar performance outcomes), led Becker and Gerhart (1996) to propose an HR system structure explanation in which individual HR practices could be seen as embedded within the higher-level components of HR policies and “guiding principles.” In other words, equally effective firms might share certain overarching HR principles or values, but differ in terms of the specific HR programs and practices that they use ( O'Reilly & Pfeffer, 2000). In addition to issues regarding the appropriate level at which to analyze the effect of HR systems on performance, recent empirical studies have raised important methodological issues regarding the relationship between measures of HR systems and performance (Gerhart et al., 2000, Wright et al., 2005 and Wright et al., 2001). First, Gerhart, Wright, and colleagues argued that effect size estimates of the relationship between HR system measures and performance depend on assumptions about measurement reliability which may not be valid. This issue has led to a debate within the literature concerning the significance of inter-rater reliability measures and the need for multiple respondents in measuring HR practices (Becker and Huselid, 2006 and Huselid and Becker, 2000). Second, several recent studies have called into question whether evidence from previous studies could be seen as demonstrating that HR practices caused changes in performance ( Wall and Wood, 2005 and Wright et al., 2005). These results highlight the need for strategic HRM researchers to re-examine some basic assumptions about the validity of the HR system construct in order to move research forward both theoretically and empirically. The first step in establishing construct validity is to address the issue of construct meaning and definition (Schwab, 1980). As part of this definition phase, organizational researchers need to pay special attention to the appropriate level of analysis of the construct. “For example, does the construct represent a structural property of organizations, perceptions of the property (and if so, by whom), or employee affect toward the property? Much confusion has been created because the construct referent has not been made clear in the definition and/or moving from definition to measurement” (Schwab, 1980: 12–13). This observation, made over 25 years ago, applies as well to current strategic HRM research. To move forward in terms of developing more complex and comprehensive theoretical models, the time may be right for strategic HRM researchers to pay closer attention to conceptual and empirical issues related to organizational levels of analysis (Chan, 1998, Klein et al., 1994, Kozlowski and Klein, 2000 and Rousseau, 1985). The purpose of this study is to contribute to the study of the HR system and firm performance relationship by addressing definitional and levels of analyses issues involved in conceptualizing and measuring the HR system construct. We begin by briefly reviewing the strategic HRM literature and identifying five components of the HR system structure: HR principles, policies, programs, practices and climate. Material in this section builds on previous distinctions and emphases in the strategic HRM literature. Next we draw on organizational levels of analysis literature to address current and future issues in strategic HRM research including the debate over inter-rater reliability and the identification of specific measurement and conceptual issues involved in defining and measuring the HR system structure components. From this review of research on levels of analysis, we develop guidelines for researchers to answer questions about who in the organization should provide information about the HR system construct and what types of data should be gathered. Finally, we draw on insights from the framework to discuss understanding sources of variability in measures of HR system structure components and suggest an expanded strategic HRM research agenda.

In this paper, we have sought to contribute to the growing body of conceptual and empirical research linking the HR system construct and firm performance outcomes. We proposed that research progress requires addressing some basic definitional and levels of analyses issues involved in conceptualizing and measuring the HR system construct. Drawing on previous strategic HRM literature, we identified five components of the HR system structure: HR principles, policies, programs, practices, and climate. Concepts and insights from the literature on organizational levels of analysis provide important guidelines to researchers in conceptualizing and measuring these HR system structure components. In addition, distinctions between HR system components may help to shed some light on current methodological debates in the strategic HRM literature and avoid the potential for misattribution across HR system components. Finally, by questioning the within-group homogeneity assumption and exploring sources of systematic as well as random error in measures of HR system structure components, we can begin to develop and test more complex and comprehensive models that promise to enhance our knowledge of the inter-relationship between HR system components and firm performance outcomes.