مشخصات بازار، هماهنگی درون شرکت، و انتخاب سیستم های مدیریت منابع انسانی: نظریه و شواهد

This paper begins by constructing a team-theoretical model of organizational adaptation and coordination with three distinct task coordination modes: vertical control, horizontal coordination, and hybrid coordination. The model is then used to provide fresh insights on complementarities involving team work organization, communication channels, training and hiring, and other human resource management practices, and illustrate how such choice of practices is affected by the firm's output market conditions. Our econometric analysis of new data from Japan which provide up-to-date information on the adoption of new team-based instruments for a horizontal coordination system (cross-functional problem solving project teams and Self-Managed Teams) yields results that are broadly consistent with the theory. First, new team-based instruments are more likely to be adopted by firms with well-established formal shop-floor-based communication channels (such as shopfloor committees), while they are much less likely to be adopted by firms with well-established information sharing institutions such as joint labor-management committees, which presumably enhance the efficiency of the vertical control system by minimizing labor-management communication errors. Finally, firms in more competitive markets and those with a higher concentration of sales among a small number of customers are more likely to adopt both types of team, whereas firms facing more erratic price movement tend not to adopt Self-Managed Teams.

Much of the literature on innovative HRMPs (Human Resource Management Practices) focuses on their effects on enterprise performance. It generally finds that HRM systems with complementary practices such as teams, joint labor-management committees, and incentive pay raise productivity or other firm performance measures.1 However, relatively few attempts have been made to identify under what circumstances firms adopt these new practices and how they select a particular combination of HRMPs. Prior studies are largely empirical explorations of the determinants of the adoption of HRMPs, and different authors stress different possible determinants. For instance, Arthur (1992) views the firm's selection of particular HRMPs as a part of its overall business strategy, and presents evidence from minimills in the U.S. that a firm is more likely to choose HRMPs promoting employee commitment and involvement in concert with business strategies focusing on product and customer differentiation (as opposed to low cost strategies). Pil and MacDuffie (1996) use panel data of automobile assembly plants worldwide and show that employee involvement programs are more likely to be adopted in plants with other complementary personnel policies and practices such as hiring and training policies aimed at employee ability to learn new skills and teamwork. Another econometric case study of steel finishing lines by Ichniowski and Shaw (1995) demonstrates that innovative and productivity-enhancing HRMPs are more apt to be adopted by “young” plants with a relatively low cost of switching from the existing old system to the new system. Finally, Osterman (1994) uses nationally representative samples of U.S. establishments encompassing diverse industries to study general patterns of the adoption and diffusion of innovative work practices.2 Among other things, the study reports evidence on a relationship between the adoption of HRMPs and exposure to global competition, which is closely related to one of the key findings of this paper. Evidence on the incidence of new innovative work practices is even more limited outside of the U.S. Jirjahn (2002) applies a similar methodology used by Ichniowski and Shaw (1995) to his German establishment-level data and identifies a number of key determinants of the adoption of innovative HRMPs by German firms. One of Jirjahn's key findings is that innovative HRMPs are more apt to be adopted by firms that rely on international markets for their products, yet less likely to be adopted by firms that consider domestic market share a primary goal. Brown et al. (2007) use Australian establishment-level data (Australian Workplace Industrial Relations Survey) and report evidence pointing to the importance of long-term employment relationships between the firm and workers as a significant complement to employee involvement programs. For cross-country comparative studies, Poutsma et al. (2003) use a survey of firms in ten EU members and stress the importance of country specific factors as determinants of the adoption and diffusion of participatory employment systems. Finally, on Japanese firms, Jones and Kato (1993) find that firms are more likely to adopt employee stock ownership plans when recent business performance is below average, the capital/labor ratio is relatively low, and employment growth is relatively fast. More recently Kato (2006), using firm-level data from Japan, shows that the employment system consisting of complementary HRMPs evolves significantly over time, pointing to the importance of a more dynamic view of HRMPs, which we will take up in this paper by examining why some Japanese firms add new forms of team-based HRMPs to the existing participatory employment system while others do not. By reviewing the literature on the adoption of HRMPs, we are struck by the relative scarcity of coherent and testable theories on the firm's choice of HRMPs. Milgrom and Roberts (1990) demonstrate how we can apply Topkis's framework of monotone comparative statics to a study of complementarities among the firm's choice of technology and practices, using a very important application to flexible production system vs. mass production system. Their approach, however, neglects the information-processing nature of the organization and the resultant need for coordination, which we believe are the key mechanism that explains many differences in HRMPs among firms. Information-processing models of organizational architecture based on team theory (Marschak and Radner, 1972) have been employed by Aoki, 1986 and Aoki, 1990 and Carter (1995). Both works illustrate important tradeoffs among organizational modes: bounded rational control (prior planning) vs. imperfect horizontal coordination (ad hoc adaptation based on posterior information) in Aoki, 1986 and Aoki, 1990 and among seven organizational forms (routine, marketing dominated, production dominated, marketing-led, production-led, centrally-managed, pooled information) which differ in the degrees of information aggregation and their information sharing schemes in Carter (1995). Their implications for the evolution of organizational architecture are rather limited, however, because the relative efficiency of each organizational mode depends on unknown parameters which are hard to observe or interpret: rate of learning-by-doing and interdependency among units in Aoki, 1986 and Aoki, 1990 and organizational costs treated as black-box in Carter (1995). We believe that the literature has been mostly silent on the potentially important interplay between modes of task coordination within the organization and its choice of bundling of HRMPs. Moreover, as contingency theory in organizational behavior (developed in the 1960s) stresses, a firm's technological and output market conditions may play a crucial role in determining the firm's choice of a specific mode of task coordination within the organization, and hence its selection of a specific HRM system (provided there is an important link between task coordination modes and HRM systems).3 This paper begins by constructing a coherent team-theoretical model building on the work of Dessein and Santos (2006) and Aoki, 1986 and Aoki, 1990. Aoki's works focus on comparing the efficiency of prior planning under vertical control to that of decentralized adaptation based on on-site information. The model structure of Dessein and Santos allows us to examine the basic tradeoffs between adaptation and coordination, which are inseparable in Aoki. By combining the elements of the two models, we can evaluate how a change in adaptation requirement affects the relative efficiency of the two organizational modes studied in Aoki. More specifically, we derive three distinct modes of task coordination within the organization: vertical control, horizontal coordination, and hybrid coordination.4 The model is then used to generate implications about complementarities involving work organization, communication channels, training and hiring and other management practices, and illustrate how such choice of HRMPs is affected by the firm's output market conditions. Guided by our theoretical exploration, we analyze unique data from a new survey of Japanese firms that provides comprehensive data on evolving HRMPs in Japan in recent years. Our probit estimates confirm two key insights from our theoretical explorations: (i) complementarity between team-based work organization and communication channels; and (ii) interplay between a firm's output market characteristics and its choice of HRMPs. Specifically, the probit estimates suggest that: (i) firms with well-established formal shop-floor-based communication channels (such as Shop Floor Committees) are more likely to adopt new team-based instruments for information sharing and problem-solving (such as cross-functional problem solving project teams and self-directed teams), for Shop Floor Committees are complementary to Cross-Functional Teams and self-directed teams in horizontal coordination systems; and (ii) firms with well-established information sharing institutions such as Joint Labor-Management Committees (JLMCs) are less likely to adopt Cross-Functional Teams and self-directed teams, since JLMCs tend to enhance the efficiency of vertical control systems by minimizing labor-management communication errors, and hence reduce incentives to introduce new team-based instruments for horizontal coordination systems. Concerning the interplay between product market competition and HRMPs, our probit estimates suggest that firms in more competitive markets and those with a higher concentration of sales among a small number of customers are more likely to adopt both types of teams while firms facing more erratic price movement tend not to adopt Self-Managed Teams. In the next section, we present our theoretical exploration, followed by an empirical analysis of new Japanese data with a brief introduction to the new survey in Section 3. Concluding remarks are offered in Section 4.

This paper has explored theoretically and empirically a potentially important yet oft-neglected interplay between task coordination within an organization and the structure of the organization and HRMP bundling. In so doing, we have also provided fresh insights on linkages between a firm's output market characteristics and its choice of HRMP system, for a firm's market conditions affect how the firm coordinates various tasks within the organization. Furthermore, such output market conditions also influence the firm's decision about its hiring and training policy, as well as its market strategy. In sum, we have explored complementarity among strategies, human capital, and organizational choices as suggested by Bresnahan et al. (2002). Our model, which was built on works by Dessein and Santos (2006) and Aoki (1986), has two important implications: (1) as the value of adaptation vs. coordination increases, a firm is more likely to rely on horizontal coordination rather than vertical control, and utilize teams and invest in horizontal communication channels; (2) as the market environment becomes more disruptive, the firm is more likely to adopt a vertical control system, and hence develop labor-management communication channels. A rise in the value of coordination may or may not induce more usage of teams and horizontal coordination. Guided by our theoretical exploration, we have conducted an empirical analysis of new data from Japan, which provide for the first time information on newer forms of organizational innovation, such as self-managed online teams and cross-functional offline teams, in addition to data on longer-established practices, including joint labor-management committees and shopfloor committees. One novel finding is that new team-based instruments are more likely to be adopted by firms with well-established formal shop-floor-based communication channels (such as shop-floor committees), while they are less apt to be adopted by firms with well-established information sharing institutions such as joint labor-management committees. This finding is consistent with the view that: (1) firms where activities or decisions are coordinated differently tend to adopt different types of work organization; and (2) communication channels such as labor-management committees and shop-floor committees play an important role in intra-firm coordination. The finding largely supports path-dependency in the evolution of organizational architecture. We have also confirmed empirical patterns that are consistent with the implications from our theory about the relationships between market structure/environment and the adoption of team organizations. Specifically, firms in more competitive markets and those having a higher concentration in sales among a small number of customers in long-term buyer–supplier relationships are more likely to adopt both types of teams, while firms facing more erratic price movement tend not to adopt self-managed online teams. Finally, some of our findings can be meaningfully compared to prior studies (although different data and methodologies used by prior studies make strict comparisons rather difficult). First, our finding of a positive and significant association between exposure to global competition and innovative work practices is consistent with some prior studies based on data from other countries, such as Osterman (1994) on U.S. establishments, and Jirjahn (2002) on German establishments. Second, some prior studies report a significant linkage between the incidence of innovative HRMPs and firm age (notably Ichniowski and Shaw, 1995). We failed to find such linkage. The lack of a significant linkage of the incidence of newer forms of HRMPs to firm age is, however, not too surprising in the Japanese context, in which most firms had already adopted a participatory work system some time ago, and what we are studying is in essence whether these already “participatory” Japanese firms add newer forms of HRMPs. The “birth effects” (Ichniowski and Shaw, 1995) are probably less relevant to our study. Third, prior evidence on the relationship between firm size (measured by number of employees) and the adoption of innovative HRMPs tends to be inconclusive. For instance, Osterman (1994) and Jirjahn (2002) report mixed evidence whereas Brown et al. (2007) find more consistently positive associations between firm size and employee involvement schemes. Though the estimated coefficients on firm size in our study are mostly positive, as in Brown et al. (2007), they are imprecisely estimated, and hence no definitive conclusion can be drawn.