OrganizationScience Vol. 21, No. 2, March���April 2010, pp. 490���506 issn 1047-7039 eissn 1526-5455 10 2102 0490 informs �� doi 10.1287/orsc.1090.0446 �� 2010 INFORMS A Relational Model of How High-Performance Work Systems Work Jody Hoffer Gittell The Heller School for Social Policy and Management, Brandeis University, Waltham, Massachusetts 02454, jgittell@brandeis.edu Rob Seidner College of Urban Planning and Public Affairs, University of Illinois at Chicago, Chicago, Illinois 60607, seidner@alumni.brandeis.edu Julian Wimbush Health Services and Policy Analysis Program, University of California, Berkeley, Berkeley, California 94270, wimbush@berkeley.edu Ioutcomes. n this paper we explore a causal mechanism through which high-performance work systems contribute to performance We propose that high-performance work systems can improve organizational performance by strengthening relationships among employees who perform distinct functions, a pathway that is expected to be particularly important in settings characterized by highly interdependent work. In a nine-hospital study of patient care, we identify high-performance work practices that positively predict the strength of relational coordination among doctors, nurses, physical therapists, social workers, and case managers, in turn predicting quality and efficiency outcomes for their patients. Relational coor- dination mediates the association between these high-performance work practices and outcomes, suggesting a relational pathway through which high-performance work systems work. Key words: high-performance work systems relational coordination patient care History: Published online in Articles in Advance July 21, 2009. Introduction One of the core principles of strategic human resource management is that organizational performance is influ- enced by the way employees are managed. In support of this argument, certain sets of human resource practices have been found to improve employee effectiveness and to predict higher levels of organizational performance (Bailey et al. 2001, Ramsey et al. 2000 see also reviews by Becker and Gerhart 1996, Ichniowski et al. 1996). Researchers have documented the impact of human resource practices on efficiency outcomes such as worker productivity (Arthur 1994, Bartel 1994, Datta et al. 2005, Koch and McGrath 1996) and equipment reliabil- ity (Ichniowski et al. 1997, Youndt et al. 1996), on qual- ity outcomes such as manufacturing quality (MacDuffie 1995) and patient mortality (West et al. 2002), and on business growth (Bartel 2004) and financial performance (e.g., Collins and Smith 2006, Delery and Doty 1996, Huselid 1995, Wright et al. 2006). Human resource prac- tices have also been found to explain performance dif- ferences among steel-finishing lines (Ichniowski et al. 1997), call centers (Batt 1999), airlines (Gittell 2001), banks (Richard and Johnson 2004), and high-tech firms (Collins and Clark 2003), though some studies have found no performance differences associated with human resource practices (e.g., Cappelli and Neumark 2001). Multiple labels have been applied to this basic argument, including high-performance work systems, high-commitment work systems, high-involvement work systems, and high-performance human resource man- agement. Despite these different labels, their common thread is that organizations can achieve high perfor- mance by adopting practices that recognize and lever- age employees��� ability to create value. Though some disagreement remains among researchers, it is generally agreed that these practices include selection, training, mentoring, incentives, and knowledge-sharing mecha- nisms (Horgan and Muhlau 2006) and that these prac- tices are most effective when they are implemented in bundles because of their combined effects on perfor- mance (Batt 1999, Dunlop and Weil 1996, Ichniowski et al. 1997, Laursen 2002, MacDuffie 1995). There is less agreement, however, regarding the causal mechanisms through which high-performance work sys- tems influence performance outcomes. The two dominant arguments are based on human capital and skill on the one hand, and motivation and commitment on the other. In addition, there is an emerging view that employee- employee relationships constitute a third causal mech- anism through which high-performance work systems influence performance outcomes (Delery and Shaw 2001). Rather than focusing primarily on the knowledge 490

Gittell, Seidner, and Wimbush: Relational Model of How High-Performance Work Systems Work Organization Science 21(2), pp. 490���506, �� 2010 INFORMS 491 and skills of employees or on the commitment of employ- ees to their organization, this third view focuses on relationships between employees as the primary causal mechanism that connects high-performance work sys- tems and performance outcomes (e.g., Collins and Clark 2003, Collins and Smith 2006). In this paper we adopt this third view and propose a model of high-performance work systems in which each component practice reaches across multiple func- tions to engage employees in a coordinated effort. All the high-performance work practices identified in this study are focused on building employee-employee rela- tionships. We argue that these high-performance work practices contribute to performance by supporting the development of relational coordination, a mutually rein- forcing web of communication and relationships carried out for the purpose of task integration (Gittell 2002b). We test our model with multilevel data from a nine- hospital study of patient care that includes administrator interviews to measure work practices, care provider sur- veys to measure relational coordination, and patient sur- veys to measure patient outcomes. We explore the effects of these high-performance work practices on quality and efficiency outcomes for patients and the mediation of these effects through relational coordination among care providers. Hospitals are notorious for operating with well-defined silos that engender turf battles between them. We expect that relational coordination will be crit- ical for achieving desired performance outcomes in this setting due to the high levels of task interdependence, uncertainty, and time constraints found there (Gittell 2000). We expect these high-performance work prac- tices will foster relational coordination, thus bridging the boundaries between the distinct professions that are responsible for carrying out the work. Human Capital and Commitment Models of How High-Performance Work Systems Work Models of high-performance work systems often draw on human capital theory, whose central implication is that human resource practices can improve organiza- tional performance by increasing the knowledge and skills of employees (Becker 1975). To be successful, firms must invest in and maintain the workforce just as they invest in and maintain the capital infrastruc- ture. Researchers have found that companies can achieve sustained performance advantages by leveraging the knowledge of their employees. High-performance work systems can foster the development of human capital in the form of firm-specific idiosyncratic skills (Gibbert 2006), creating a performance advantage for organiza- tions (Fried and Hisrich 1994, MacMillan et al. 1987, Tyebjee and Bruno 1984) through processes such as increased employee problem solving (Snell and Dean 1992) and improved customization by employees in ser- vice industries (Batt 2002). Others have argued that in addition to building the knowledge and skills of employees, high-performance work systems also enhance the motivation and commit- ment of employees. Commitment-based human resource practices create an organizational climate that motivates employees to act in the best interest of the organi- zation, thus enhancing performance (Appelbaum et al. 2000, Arthur 1992, Osterman 1988, Rousseau 1995). A key argument in this literature is that human resource practices build a psychological contract by signaling an employer���s commitment to a long-term relation- ship, in turn yielding a long-term commitment from the employee (Tsui et al. 1997). Consistent with this argument, studies have found that particular work prac- tices are associated with higher levels of commitment (e.g., Tsui et al. 1997, Whitener 2001) and that commit- ment in turn is positively associated with performance. In particular, Bowen and Ostroff (2004) provide argu- ments suggesting that motivation and discretionary effort underlie the association between human resource prac- tices and performance and are triggered by a strong human resource system. Note that the human capital and commitment pathways are not mutually exclusive. Although research and theory often focus on one or the other, some theorists have argued that high-performance work systems can contribute to performance through both pathways (e.g., Appelbaum et al. 2000). Relational Models of How High-Performance Work Systems Work Relationships among employees have also been theo- rized to play a role in achieving high levels of orga- nizational performance. Some scholars have made this argument by drawing on the concept of organizational social capital, a type of social capital that exists in and can be developed by organizations as a distinctive orga- nizational capability and source of competitive advan- tage (Leana and Van Buren 1999, Nahapiet and Ghoshal 1998). Organizational social capital has been shown to improve performance by enabling employees to access the resources that are embedded within a given network and by facilitating the transfer and sharing of knowledge (Levin and Cross 2004, Tsai and Ghoshal 1998). Other theorists have argued that employee-employee relationships are important for coordinating work (Adler et al. 2008, Faraj and Sproull 2000, Gittell 2000), based on the argument that coordination is the management of task interdependence (Malone and Crowston 1994) and therefore fundamentally a relational process (Bechky 2006, Faraj and Sproull 2000, Gittell 2002b, Weick and Roberts 1993). One of these relational perspectives��� relational coordination���identifies specific dimensions of relationships that are integral to the coordination of work. According to the theory of relational coordina- tion, coordination that occurs through frequent, high- quality communication supported by relationships of

Gittell, Seidner, and Wimbush: Relational Model of How High-Performance Work Systems Work 492 Organization Science 21(2), pp. 490���506, �� 2010 INFORMS shared goals, shared knowledge, and mutual respect enables organizations to better achieve their desired out- comes (Gittell 2006). Defined as ���a mutually reinforcing process of interaction between communication and rela- tionships carried out for the purpose of task integration��� (Gittell 2002a, p. 301), relational coordination is a type of employee-employee relationship that is particularly relevant for coordinating work that is highly interdepen- dent, uncertain, and time-constrained. Substantial progress has been made toward identifying the work practices through which organizations influence the development of employee-employee relationships. Leana and Van Buren (1999) argue that stable employ- ment relationships and reciprocity norms facilitate the formation of social capital among employees. Evans and Davis (2005) argue that work practices such as selec- tive staffing, self-managed teams, decentralized deci- sion making, extensive training, flexible job assignments, open communication, and performance-contingent com- pensation influence multiple dimensions of an organi- zation���s social structure, including the development of bridging ties, norms of reciprocity, shared mental mod- els, role making, and organizational citizenship behavior. Gittell (2000) argues that work practices such as cross- functional selection, cross-functional conflict resolution, cross-functional performance measurement, flexible job design, and cross-functional boundary spanner roles can foster the development of relational coordination. These work practices were shown to predict significantly higher levels of relational coordination among airline employees from 12 distinct functions who were engaged in the flight departure process, though their impact on performance was not explored. Similarly, Gant et al. (2002) show that on steel- finishing lines with high-performance work systems, defined as selection, training, incentive pay, job design, problem-solving teams, and extensive labor/management communication, production employees have denser com- munication networks with each other and that these steel- finishing lines also exhibit higher performance, measured in terms of fewer delays and higher yields. They argue that these human resource practices influence perfor- mance outcomes because they influence the social net- works of production employees. Their results suggest that social networks may mediate the link between work practices and outcomes, though mediation was not demonstrated. Collins and Clark (2003) have provided one of the best empirical tests to date of the argument that human resource practices influence outcomes through their impact on relationships among employees. They argue that the social networks of top management teams enhance a firm���s information-processing capabil- ity and that human resource practices, including men- toring, incentives, and performance appraisals, can be designed to encourage the development of these social networks. They then demonstrate that the impact of these high-performance work practices on firm performance is mediated by the strength of firms��� top management team social networks. More recently, Vogus (2006) has argued that high-performance work practices such as selec- tion, training, performance appraisal, performance-based rewards, and job security contribute to high-quality inter- actions and mindfulness by signaling to employees the importance of relationships. Vogus continues by postu- lating that these high-quality interactions contribute to higher-quality outcomes for hospital patients. Empirical tests of this model demonstrated mediation. Though the types of employee-employee relationships explored in these studies are varied, including relational coordination (Gittell 2000), social networks (Collins and Clark 2003, Gant et al. 2002), social capital (Evans and Davis 2005, Leana and Van Buren 1999), and mindful interacting (Vogus 2006), these studies suggest that high-performance work practices can enhance per- formance through the pathway of employee-employee relationships. TheoryBuilding The work practices found in the studies described above resemble in many ways the work practices found in the earlier high-performance work systems literature ���they include selection, training, performance mea- surement, rewards, knowledge-sharing mechanisms, and so on. But they differ in an important way. Although the work practices found in these studies have the potential to influence employee skills and commit- ment, they are focused primarily on strengthening relationships between employees. This understanding of high-performance work practices therefore responds implicitly to an argument by post-bureaucracy theo- rists that traditional work practices often create divi- sions between employees even when relationships are critically important due to the need for coordination (Heckscher 1994, Piore 1993). According to Piore (1993, p. 15), the bureaucratic organizational practices that have become widespread through the rise of Tay- lorism ���have pushed us to restrict communication among the people responsible for the way in which the differ- ent parts are performed.��� Heckscher (1994, p. 24) envi- sions a postbureaucratic, interactive organizational form in which ���everyone takes responsibility for the success of the whole��� and in which ���workers need to understand the key objectives in depth in order to coordinate their actions intelligently ���on the fly���.��� High-Performance Work Practices as Predictors of Relational Coordination. Rather than rejecting the role of formal work practices as the postbureaucratic litera- ture has tended to do, we argue that formal work prac- tices can be redesigned to foster the employee-employee relationships through which work is effectively coordi- nated ���on the fly.��� High-performance work practices can