National Sun Yat-sen University Institutional Repository:Item 987654321/38584

行政院國家科學委員會專題研究計畫 成果報告

學習型組織與知覺環境不確定性在人力資源管理實務與組

織績效關係間扮演之角色

中 華 民 國 95 年 6 月 23 日

What Matters between Human Resource Management and Firm

Performance? The Roles of Organizational Learning and Perceived

Environmental Uncertainty

Introduction

Human resource has been considered as a source of competitive advantage (Barney, 1991; Barney & Wright, 1998; Wright, Dunford, & Snell, 2001). Empirical studies showed promising results with the direct link between human resource management (HRM) and firm performance (Delery & Doty, 1996; Huselid, Jackson & Schuler, 1997; Delaney & Huselid, 1996; Becker & Gerhart, 1996). However, still very few studies have sought to investigate why HRM effectiveness may foster firm performance. Wright, Dunford, & Snell (2001) and Wright & Gardner (2002) suggested that, instead of investigating the direct effects of HRM on firm performance, researchers should identify those critical resources and capabilities that HRM can support or facilitate to achieve competitive advantage. As thus, it appears worthwhile to identify the capability that may intervene in the HRM-firm performance relationship.

Previous literature suggested that organizational learning (OL) may be a candidate in explaining the relationship of HRM and firm performance. HRM has been argued as a facilitator of OL (Pucik, 1998; Watkins & Marsick, 1996; Ulrich, 1987). OL was also shown

as a critical organizational capability and as a source of competitive advantage (Pucik, 1998; Pitt, 1998). However, very few efforts were made to investigate the associations of HRM, OL and firm performance. Under such circumstances, examinations of OL as a mediator between HRM and firm performance should be rewarding.

On the other hand, although OL may enhance firm performance, its effects in various contexts have not received much attention yet. Baldwin & Danielson (2002) argued that OL might not be imperative for organizations in different environments. Researchers found that environment will define an organization’s learning needs (Wijnhoven, 2001; Gnyawali & Stewart, 2003). Previous studies suggested that environmental uncertainty is a critical contingent factor (Milliken, 1990; Brouthers, Brouthers, & Werner, 2000). Therefore, this study will also investigate how environmental uncertainty may moderate the relationship of OL and firm performance.

In summary, this study endeavors to open a “black box” between HRM and firm performance. In an effort of identifying critical capabilities for a firm, this study adopts the organizational learning perspective and investigates how OL mediates the relationship of HRM and firm performance. Then, it further examines the association of OL and firm performance from contingency perspective. It is expected that findings in this study may fill the research gaps and advance the understandings of HRM as a source of competitive advantage.

Theoretical Background and Hypotheses

HRM and Firm Performance

Resource-based theory (Wernerfelt, 1984; Barney, 1991) proposes that a firm with heterogeneous, rare, valuable, and inimitable resources can gain competitive advantage. Since a well-established HRM system can only be embedded in an organization, and both its causal ambiguity and path dependency make it difficult to be imitated (Becker & Gerhart, 1996), HRM has been viewed as a critical source of competitive advantage (Boxall, 1998; Barney & Wright, 1998; Wright et al. 2001).

However, the reasons why HRM effectiveness may be related with firm performance were still largely overlooked. Baldwin & Danielson (2002) contended that practicing managers often questioned about the direct link of HRM issues and various performance measures. They argued the existence of intermediaries in the HRM-performance relationship. A few recent studies also encouraged researchers to identify and investigate the mechanisms or processes with which HRM can affect firm performance (Becker & Gerhardt, 1996; Wright et al., 2001; Wright & Gardner, 2002). Therefore, it should be valuable to identify critical resources or capabilities that may benefit firm performance, and then examined their roles in the relationship of HRM and firm performance.

In this article, we specifically identify OL as a critical capability and explore its role in the relationship of HRM and firm performance.

Organizational Learning and Firm Performance

OL has been viewed as critical for a firm’s success. A few researchers argued that OL is a source of competitive advantage in turbulent environment (Slater & Narver, 1995; Lei, Slocum, & Pitts, 1999; Pitt, 1998; Pucik, 1998; Vorhies & Morgan, 2005). Ulrich et al. (1993) suggested that learning capabilities would be the premise for a firm to win. Empirical studies have shown that the more an organization learns, the more it can advance innovation capabilities (Stata, 1989; Hargadon, 2002; Calantone, Cavusgil, & Zhao, 2002). Moreover, OL matters much with a firm’s market orientation (Hurley & Hult, 1998) and competitive position (Sloan, Hyland, & Beckett, 2002; Blazevic & Lievens, 2004).

In addition, a few studies showed that OL is correlated with financial performance. For example, Lien, Yang, & Li (2002) found OL is associated with perceptual financial performance in Taiwanese firms. Ellinger, Ellinger, Yang, & Howton (2002) demonstrated OL is related with both perceptual and objective measures of financial performance. As such, Stata (1989: 64) proclaimed that: “the rate at which individuals and organizations learn may become the only sustainable competitive advantage”. Under such circumstances, we propose:

Hypothesis 1: The degree of a firm’s organizational learning will be positively related

HRM and Organizational Learning

HRM has been characterized as a facilitator of OL. HRM was proposed to be the infrastructure of OL (Pucik, 1998) and the key in sculpting OL culture (Watkins & Marsick, 1996; Buyens, Wouters, & Dewettinck, 2001). A few HRM practices were argued as the foundations of OL (Vassalou, 2001). HRM may also enhance a firm’s learning abilities (Buyens et al., 2001). A firm’s commitment to “high-impact learning” is often expressed in HRM practices and policies (Ulrich et al., 1993). Moreover, HRM may motivate employees to successfully implement OL (Appelbaum & Goransson, 1997).

Among very few empirical studies, Hatch & Dyer (2004) showed HRM practices, e.g. human capital selection and training, are related with learning by doing. Thomsen & Hoest (2001) found that HRM practices, e.g. reward system, have significant effects on learning environment. Griego et al. (2000) showed rewards and training are motivators of OL. These primary findings suggested further examinations of the relationship between HRM and OL.

Hypothesis 2: A learning-based HRM system will be positively related to a firm’s degree

of organizational learning.

Organizational Learning as a Mediator

The direct links between HRM and firm performance faced serious challenge recently. Researchers argued that the effects of HRM on firm performance should be through its effects on the firm’s critical resources and capabilities (Wright & McMahan, 1992; Becker &

Gerhardt, 1996; Wright et al., 2001; Wright & Gardner, 2002; Collins & Clark, 2003). Empirical studies have identified some intervening factors. Employees’ turnover rate and productivity (Huselid, 1995), learning by doing (Hatch & Dyer, 2004), market orientation (Harris & Ogbonna, 2001) and TMT social network (Collins & Clark, 2003) were found to mediate the relationship of HRM and firm performance. Since a firm’s competitive advantage can benefit from OL, we expect that one way HRM may affect firm performance is through its effect on OL.

Hypothesis 3: Organizational learning will mediate the relationship of learning-based

HRM system and firm performance.

The Contingency Perspective of OL-Performance Relationship

Contingency theory assumed that there is no best way to organize, and any one way of organizing is not equally effective in all situations (Galbraith, 1977). Environmental uncertainty was considered as an important contingency (Ginsberg & Venkatraman, 1985). Empirical findings supported that alignments of business strategy and uncertainty may lead to better performance (Tan & Lischert, 1994; Venkatraman & Prescott, 1990). As such, it seems meaningful to further investigate how OL-performance relationship may be contingent on environment.

Only few references examined OL from contingency perspective.Baldwin & Danielson (2002) contended that OL may be more effective in uncertain environment than in stable

environment. Previous studies also proclaimed that OL will be relatively useful in highly uncertain environment (Sitkin, Sutcliffe, & Schroeder, 1994; Jabnoun, Khalifah, & Yusuf, 2003). OL mechanism is found to be associated with perceived uncertainty (Ellis & Shpielberg, 2003). Similarly, Miller & Shamsie (1996) found that knowledge-based resources, such as OL, would produce superior financial performance in uncertain environment. Since competition is fluid and market is dynamic in nature, firms that are characterized as corporate-wide learning, knowledge sharing, and continuous transformation becomes advantageous.

In summary, the investigations of moderating effects of environmental uncertainty on the relationship of OL and firm performance appear to be valuable. Therefore, we propose:

Hypothesis 4: Organizational learning will produce better firm performance when

environmental uncertainty is high than when it is low.

Methodology

Data Source and Sample

Data under research were drawn from the 2003 Top 5,000 enterprises in Taiwan, ranked by China Credit Information Service Company (CCIS). CCIS is recognized as a professional institution in industry analyses. Anonymous questionnaire were sent to 500 firms in the database.

invited to help complete the whole questionnaire. In an effort of reducing the common method bias (Podsakoff & Organ, 1986), we also asked a second external-oriented informant, who should be a marketing specialist or CEO assistant (cf. Smith, Grimm, Gannon, and Chen, 1991), helped fill the second part of the questionnaire (i.e. the environmental uncertainty, munificence and firm performance). Since it’s difficult to collect multiple responses from a company in Taiwan (Hempel & Chang, 2002), the external-oriented informant were invited to rate fewer items with the expectation to increase the response rates. However, the response rates of the second part were still relatively low.

In this way, 120 firms participated in the whole survey, with a response rate of 24%. Among the responses, 10 copies are incomplete or invalid. Therefore, 110 usable questionnaires were used for final analyses. As with the second part, all of 38 responses were valid, and the response rate was 7.6%.

Measures

Organizational Learning. OL occurs at various levels in an organization (Dixon, 1993;

Crossan, Lane, & White, 1999; Bapuji & Crossan, 2004). Since OL notion remained elusive (Friedman, Lipshitz, & Popper, 2005; Berends, Boersma, & Weggeman, 2003) and most of previous OL research relied on archival data (Bapuji & Crossan, 2004), we adapted the OL definition of Crossan et al. (1999) and developed an OL scale for this study, which includes measures of individual-based, mutual-based and global-based learning.

Finally, a 19-item, 7-point OL scale was developed. Cronbach’s α showed good reliabilities of the scale (α= .94). Factor analysis also showed the construct validity of the scale is quite good.

Learning-based HRM System. A specific-oriented HRM system can support the capability

that a firm is trying to develop (Wright & Snell, 1998). Accordingly, we develop a scale for learning-based HRM system which may relate with a firm’s degree of OL. This scale referred to that of Jaw & Liu (2003), Lepak & Snell (2002), MacDuffie (1995) and Collins & Clark (2003). Finally, a 26–item, 7-point learning-based HRM system scale was developed. Cronbach’s α showed good reliabilities of the scale (α= .96). Factor analysis of learning-based HRM system showed good construct validity as well.

Environmental Uncertainty. Environmental uncertainty was assessed with the measure

developed by Miller & Droge (1986). Back translation was implemented to ensure the comparability of original and translated versions. Overall, the 4-item, 7-point scale showed good reliability (α= .76).

Firm Performance. Firm performance was measured with a 7-item, 7-point perceptual

performance scale (α= .95). It was measured with the firm’s past three-year performance averagely relative to that of competitors. Previous studies have found that perceptual performance measures are significantly related with objective measures (Delaney & Huselid, 1996).

Control Variables. Firms were classified into manufacturing or service firms to capture the

industry effects. Firm age, measured with years, was included to capture the founding values (Stinchcombe, 1965). Firm size, measured with the logarithm of the number of employees, may capture the effects of slack resources (Singe, 1990). Environmental uncertainty was assessed with the scale developed by Miller & Droge (1986). Munificence scale was adapted from that of Hambrick & Finkelstein (1987). The 4-item, 7-point mesure showed good reliability (α= .78). Back translation was also implemented in the munificence scale.

Analyses

First of all, we conducted several analyses to validate appropriateness of our data. The results of ICC (1) and ICC (2) (Shrout & Fleiss, 1979) showed that responses in the same firm were highly consistent and firm-level means appeared stable. On the other hand, for those firms that only one informant responded our questionnaire, Harman’s one-factor post hoc analysis provided evidences that common method bias was not a serious problem in this study (Podsakoff & Organ, 1986).

Secondly, in order to examine the associations among variables, hierarchical regression analyses were conducted to test the hypotheses.

Results of correlation analysis showed preliminary bivariate relationships for the variables under study.

Regression results showed HRM was significantly related with firm performance (β= .300, P< .01). OL was also positively related with performance (β= .303, P< .01). Hypothesis 1 was supported. As with the relationship of HRM and OL, regression analysis also showed that HRM was significantly related with OL (β= .920, P< .01). The explained variances of OL by HRM are 63.40%. Hypothesis 2 was supported.

As with the mediating effects, when OL was incorporated into the model, HRM became insignificant, which implied the relationship of HRM and firm performance was fully mediated by OL. Hypothesis 3 was supported. Finally, statistical analysis showed the interactions of OL and environmental uncertainty was significantly related with performance. This finding supported our proposition in Hypothesis 4 about the contingent effect of OL on firm performance.

Discussions

Implication

In response to the urge of uncovering the black box between HRM and firm performance (Wright et al., 2001; Wright & Gardner, 2002), this study investigated the role of OL as a mediator in the HRM-performance relationship. Research findings have shown that OL may affect firm performance and HRM may facilitate a firm’s OL. Moreover, OL fully

mediated the relationship of HRM and firm performance, a finding similar to that of Collins & Clark (2003). In addition, contingent investigations showed that OL is more effective for firms in relatively uncertain environment than in stable environment. These results supported our hypotheses and identified HRM and OL as sources of competitive advantage.

This study has implications for academics. Examinations of mediating effects may help identify the direct, indirect and total effects of HRM on firm performance. At the same time, this study found one mechanism (i.e. OL) with which HRM may affect firm performance. As such, HRM system demonstrated its significant contributions to support critical organizational capability and competitive ability as suggested in previous studies (Barney, 1991; Wright et al., 2001). This study also identified HRM as a facilitator of OL, which may provide insightful implications for the OL research. Quantitative studies in OL were still quite few (Bapuji & Crossan, 2004). This study has filled this research gap as well. Moreover, scholars encouraged investigating the boundaries of the OL-performance relationship (Bapuji & Crossan, 2004). Research findings in the moderating effects of environmental uncertainty clearly echoed their concerns. Finally, the measures of OL and learning-based HRM system were developed, which may establish a foundation for further analyses.

This study may also provide insights to business practices. This study suggested that OL consists of various learning issues at individual, group and firm level, which may provide a reference framework to organizations about how to develop an OL culture. Specifically, an

effective OL may not only occur through individuals, but also occur at the social levels. Therefore, it suggests an avenue for organizations to learn from micro vs. macro and internal vs. external sources. Further, HRM practices may be the facilitator of OL. This means that HRM can not only be internal-focused, but also external-focused. As a critical strategic capability, HRM may play the role of gatekeeper to search and browse external information and thus stimulate internal learning and organizational transformation. In this way, a learning-based HRM system became critical in building up a firm’s competitive advantage. Moreover, since many firms are obliged to face dramatically change environment, contingent effects of OL appears to be relatively meaningful than ever before. As Stata (1989) argued, the rate at which a firm learns should not be undervalued since it may be the “only” source of competitive advantage in the fussy and turbulent times. Accordingly, becoming a learning organization should be considered as a critical strategic issue in the strategy formulation and implementation process.

Limitations and Future Directions

Although this study has looked into a critical topic about competitive advantage, some relevant issues were not fully addressed at this moment. Our sample came only from Taiwan and was not randomly selected, which may restrict the generalizabilities of research findings. Meanwhile, since most data came only from one informant in a company, the likelihood of common method bias may not be fully eliminated.

In addition, the effects of OL on performance might not be conclusive. It is very likely that a few factors, e.g. knowledge-based value (Morgan & Turnell, 2003) and innovation capability (Calantone, Cavusgil, & Zhao, 2002), may intervene in the relationship of OL and firm performance. Also, some other antecedents of OL, e.g. leadership (Vera & Crossan, 2004; Noat, Lipshitz, & Popper, 2004), need to be addressed in future research.

In summary, examinations of HRM as a source of competitive advantage from OL perspective have provided a few insights into the line of SHRM research. Although some issues need to be addressed further, it has shown one critical capability that firms can develop and support in order to sustain in rivalry competition.