There is a striking paradox in research on high-involvement work practices. A growing body of literature establishes strong empirical links between such practices and improved economic performance, and reveals that these practices are most effective as part of a larger “bundle,” or system, that includes complementary human resource (HR) practices (Arthur, 1992; Batt, 1995; Huselid, 1995; Ichniowski & Kochan, 1995; Ichniowski et al., 1994; Kochan & Osterman, 1990; MacDuffie, 1995a). Thus, from the perspective of economic rationality, one should expect high-involvement work practices to be widely used. Yet many argue that the imitation, learning, and diffusion of these practices have been slow and sporadic (Ichniowski & Shaw, 1995; Kochan & Osterman, 1994; Osterman, 1994; Pfeffer, 1994). The goal of this chapter is to explain why individual work practices (as well as bundles of high-involvement work practices and complementary HR practices) are adopted more rapidly by some establishments than by others.

We develop a theoretical framework that draws on evolutionary economics, innovation, and strategy literature to identify three factors driving adoption: (1) the presence of complementary HR practices and technology; (2) low levels of economic performance achieved with current work practices; and (3) organizational characteristics and behaviors that reduce the cost of introducing new work practices. Longitudinal data from the international auto industry enable us to test several hypotheses about these factors.

We examine the adoption of high-involvement work practices by analyzing data gathered in two rounds of surveys (1989 and 1993–1994) from 43 automobile assembly plants located around the world (MacDuffie & Pil, 1996). Although the use of such practices increased considerably over this five-year period, the rate of increase varied dramatically from plant to plant.