Supplier Relations and Performance in the Auto Industry: European-Japanese-US Comparisons of the Voice/Exit Choice

This paper analyses data from surveys of 1400 first-tier automotive suppliers in Europe, the United States and Japan, conducted as part of MIT's International Motor Vehicle Program. The paper shows that the nature of supplier relations in the three regions is converging in some respects towards closely-linked voice relationships. The diffusion of voice relationships in Europe and the USA appears to be due to increased competition and globalization rather than to product characteristics (such as technical complexity or asset specificity). The surveys also found that such voice relationships have performance advantages for vehicle manufacturers and their suppliers in all regions. Performance is measured in terms of customer awards to suppliers, Just-in-Time production and delivery, and costs and margins of suppliers. The realised performance advantage was however least marked for European suppliers as compared to US or Japanese suppliers. One reason behind this small impact of voice on performance in Europe is shown to be due to the prevalence of traditional cosy relationships in contrast to more recently formed committed relationships Supplier Relations and Performance in the Car Industry: EuropeanJapanese & US Comparisons of the Voice/Exit Choice I n t r o d u c t i o n In the last decade, the European automotive industry has experienced dramatic changes in its supply structure. Typically, vehicle manufacturers used to have contracts with a few thousand parts suppliers, but having more than halved their supplier base, they now have a core of three or four hundred suppliers (Wolters 1994). This process of rationalisation was in part due to a change in the purchasing philosophy. Whereas once it was thought that the best component prices could be obtained by the use of short-term contracts and hard-nosed bargaining tactics, more recently, vehicle manufacturers are asking their suppliers to work in close partnership in order to speed up the rate of improving their performance to meet the global competition (Lamming 1993). Such close relationships between customers and suppliers have been shown to have beneficial effects on performance in several areas. Clark (1989) found that early supplier involvement in product design was a key part of Japanese car manufacturers' edge in introducing new models both faster and with fewer total labor hours than their US and European counterparts. Noordeweier, John, and Nevin (1990) found that more "relational" purchasing arrangements reduced acquisition costs in the presence of high uncertainty. Heide and John (1988) found that customers and suppliers who were mutually dependent invested more in specific assets. Despite performance advantages of partnership sourcing which have been demonstrated in some parts of the world, mutual suspicion remains between customers and suppliers. For instance, suppliers might be reluctant to commit to investment with a long payback period if they believe that the customer’s commitment would last only until the next round of rationalising its supplier base. Suppliers may also suspect that vehicle manufacturers are asking them to share sensitive information, including cost breakdowns, for the sole purpose of cutting prices. When mistrust is entrenched, a shift from adversarial to more cooperative relations is evidently not easy. How much has really changed in the nature of supplier relations in the European automotive industry? How do the trends in Europe compare with those in Japan and the USA? What factors affect the choice in the type of supplier relationships? What types of relationship are likely to lead to good performance by both suppliers and customers? Do suppliers in close partnerships with vehicle manufacturers perform better in all respects than those without such relationships? In order to address these questions, Helper and Sako conducted postal surveys of first-tier component suppliers in the USA, Japan and Europe during the period 1993 1994. All the three regional surveys were sponsored by the International Motor Vehicle Program (IMVP) at MIT. The survey results are based on responses from 1416 supplier companies all of which traded directly with vehicle manufacturers.1 The respondents were asked about their situation now (at the time of the survey) and four years ago (circa 1989/1990). In this sense, the survey was not able to pick up losers (i.e. those whose contracts were severed) in the process of rationalisation as described above. Nor did it cover supplier relationships at lower tiers in the hierarchy. However, the surveys yielded an unusually comprehensive international database, which enable us to explore not only inter-regional but also intra-regional variations in the link between supplier relations and performance. The surveys show that there has been a considerable convergence in the methods of supplier relations towards those consistent with closer and longer-term relationships in Europe, the USA and Japan. There was some convergence also in the performance of suppliers in the three regions. But the gap between high performing and less well performing suppliers is not so marked in Europe as compared to in the USA and Japan. This paper explores the reasons behind this differential impact of the nature of supplier relations on supplier performance indifferent regions. But first, a conceptual framework of "voice" and "exit" is presented in Section 1. Section 2 examines the survey evidence on the changing nature of supplier relations practices. Section 3 discusses the factors which explain why some suppliers have come to forge "voice" relations while others do not. Section 4 investigates the linkages between the nature of supplier relations and performance. 1In the USA, 675 responses came from independent US-owned firms, Japanese transplants and vertically-integrated divisions of US automakers, representing a response rate of 55%. In Japan, 472 responses were received from verticallyintegrated divisions of Japanese automakers and a few foreign-owned companies as well as independent Japanese-owned firms, representing a response rate of 30%. In Europe, 269 completed questionnaires were received, a response rate of 17%. (See Appendix 1 for a description of the survey methodology. 1. A Conceptual Framework: Voice and Exit in Supplier R e l a t i o n s h i p s Traditional studies of purchasing have emphasized the distinction between "make" and "buy" (see for example, Kaserman (1978), Williamson (1975)). However, in order to analyze different options within the "buy" alternative, another framework is necessary. Here, we employ the exit-voice framework, adapted from Hirschman (1970) by Helper (1990, 1991) to classify supplier relations according to the methods used to resolve problems which arise between the parties. In an "exit" relationship, a customer who has a problem with a supplier finds a new supplier. In a "voice" relationship, the customer works with the original supplier to resolve the problem. In most cases a voice relationship is more efficient, since a rich flow of information between the parties makes possible effective use of techniques such as value analysis and value engineering. But the resulting enhancement in productivity must be distributed between the customer and the supplier, and bargaining over distribution may well detract parties from the business of enlarging the size of the pie. Moreover, a customer who wants to have a voice relationship with its suppliers must make a commitment to them. Commitment refers to the supplier's degree of certainty that the customer will continue to buy its products for some length of time. This assurance can be provided by any mechanism that makes it harder for the customer to exit from the relationship, such as vertical integration, asset specificity (Williamson 1985), long-term contracts, or desire to retain suppliers' trust (Sako, 1992). Commitment is necessary both to obtain suggestions for improvement (which may be based on proprietary information) and to make investments that respond to these suggestions. Below, this framework is used to investigate first, the trends in the diffusion of ‘voice’ based customer-supplier relationships, second the factors affecting the diffusion of voice relationships, and third whether voice relationships are indeed associated with good performance. Unless otherwise noted, all comparisons reported are statistically significant at the .05 level or better according to the Kruskal-Wallis test (a non-parametric version of the t-test.) 2. Trends in Supplier Relations Practices: A Limited C o n v e r g e n c e 'Voice' in supplier relations may be operationalised in a number of ways. For the purpose of this study, it was specified that a voice-based customersupplier relationship exists when the following three conditions hold: (A) suppliers provide customers with a detailed breakdown of their process steps; (B) suppliers believe there is a high probability that they will continue to provide products to the same customer for more than three years; and (C) If a competitor offers a lower price for a product of similar quality, suppliers expect the customer to help match the competitor's effort rather than switch to the competitor as soon as possible. How widely spread is the practice of voice from the suppliers’ viewpoint? In this section, the extent of change in each of the three dimensions of voice as defined above will be examined. Sharing of Suppliers’ Process Information with their Customers Figure 1 shows that in the last four years, the proportion of suppliers in Europe which provide their customers with a detailed breakdown of the steps they use in their production process increased from 45% to 77%. This information should help vehicle manufacturers ensure that their component designs are compatible with suppliers' processes, thus improving productivity and quality. The practice of information disclosure has also spread in the USA, from 50% in 1989 to 79% in 1993. This diffusion in Europe and the USA is in stark contrast to the stagnant situation in Japan at around 80% for both 1989 and 1993. Overall, therefore, there has been a marked convergence in the proportion of suppliers disclosing such information to customers towards around 80%.