High-performance manufacturing: evidence from the automotive components industry

This paper examines the hypothesis that lean production practices necessarily lead to high manufacturing performance and contributes to the emergent literature that attempts to establish the empirical relationship between organizational practices and performance (Arthur 1992; Huselid 1993; 1995). Specifically, the paper reports on research into the performance and management practices of 71 'first tier' automotive component plants (33 seat plants, 20 brake-caliper plants and 18 exhausts plants) based in Europe, North America and Japan. The research represents, both in scale and approach, an opportunity to test the prescriptions of the lean production model put forward by the influential International Motor Vehicle Program (IMVP).

The term 'lean production' arose from the IMVP's study of 62 automotive assembly plants and the findings were popularized by the The Machine That Changed the Worm (Womack et al. 1990). The theoretical basis of lean production and its link to high performance has been advanced by MacDuffie (1991, 1995). MacDuffie (1995) argues that an 'organizational logic' of lean production underpins three generic collections or bundles of organizational practices. These are: factory practices (related to the minimization of buffers, for example through the reduction of inventory levels and minimization of repair space), work systems (related to teamwork and the development and application of employee knowledge and skill on the shopfloor) and human resource management (HRM) practices (concerning the encouragement of high commitment and motivation of the workforce). A central claim made by 'lean' proponents is that manufacturing performance in the form of labour productivity and product quality is dramatically improved by the synergistic pursuit of these 'lean' production practices in conjunction with flexible multi-skilling work systems and high commitment human resource policies (Womack et al. 1990; MacDuffie 1995). Furthermore, lean-production management practices are advanced as a universal set of best practices which yield performance benefits at the establishment level, regardless of context and environment.

The following issues are raised by this analysis. First, are the IMVP findings applicable to other industry sectors? If lean production is universally applicable then we would expect it to apply to a closely related sector such as auto components where the car assemblers themselves have 'encouraged' their suppliers to improve their performance (Turnbull et al. 1993). A further factor which suggests the appropriateness of examining this sector is provided by Abo (1994) who suggests that lean or Japanese-style management and production systems have been more enthusiastically adopted by the car-parts makers than the assemblers themselves.

Second, the IMVP research suggests that management practices at the plant level are the most important drivers of manufacturing performance. This raises questions about the importance of more traditional explanators of performance such as the scale of operations and regarding the plant's environmental context. Furthermore, the influence of a plant's suppliers and customers on manufacturing performance remain under-explored in the IMVP analysis.

To test the lean production thesis and assess its applicability outside the car assembler sector, we undertook a detailed survey of 71 autocomponents plants between September 1993 to August 1994. This included taking measures of manufacturing performance and management practices such as the factory practices, work systems and HRM policies. To ensure that questions were answered in a consistent fashion, two members of the research team returned to review each question with plant management (see Delbridge et al. 1995a). A sub-set of 13 high performing plants were then identified and the performance and practices of these top performers were contrasted with the other plants in the sample. In addition, our method extends the work of the IMVP by incorporating inter-organizational relationships between these 'first tier' component suppliers with their respective customers (the car assemblers) and suppliers (the 'second tier') and assesses the impact of important plant characteristics such as production scale, capacity utilization and automation.

Overall, the findings provide some support for certain aspects of the lean production model. In particular, the high performing plants exhibit 'process discipline and control' i.e. they employed lean factory practices such as the use of low buffers that include an emphasis on right first-time production and logistical arrangements that require components and products to be delivered and produced just-in-time. However, the data do not show a necessary relationship between high performance and the pursuit of lean work systems and HRM policies such as team working, high involvement in problem-solving activities, extensive induction and the use of sophisticated selection techniques. Rather, the sample of high-performing plants form two distinctive sub-sets: a set of Japanese located and owned high performers demonstrate a close fit to lean work systems and HRM policies, while a set of Western high performers pursue a variety of work systems and HRM practices to equally good effect. Furthermore, the data show that high-performing plants exist within high-performing supply chains where, for example, suppliers provide them with better quality and on-time deliveries and customers provide them with more stable production schedules. Lastly, the plant's characteristics also provide a partial explanation for differences in performance, with the high-performing plants appearing to derive performance benefits from economies of scale and by operating near to their capacity.