Dynamics of firm--supplier relationships in a less developed economy: evidence from African manufacturing firms.

• Author: Fisman, Raymond

1. Introduction

   There exists a significant literature on issues relating to outsourcing in both economics and business management. (1) This literature emphasizes that organizations have limited senior managerial time and financial resources to invest in maintaining efficient performance. Therefore, organizations should focus their efforts on those activities that can be most effectively managed internally, and outsource activities that can be more effectively managed externally. Thus the focus has been more on considering, in an incomplete framework setup, the means to efficiently organize the productive activities (Hart 1995 provides an overview of the above issues).

   In this paper we consider the problem of allocating contracts to suppliers after the decision to outsource has been made. The empirical side of the business management literature is largely based on case studies of strategic alliances between organizations and buyer/supplier relationships. Its conclusions tend to be based on anecdotal evidence (such as case studies of outsourcing activities at specific firms) and informal surveys rather than rigorous statistical analysis.

   In this paper, we document concrete evidence on the nature of outsourcing relationships, using data from four African countries. We focus on how firms might deal with suppliers in an asymmetric information setup. More precisely, we find the following results: Given that there is asymmetric information regarding supplier quality, firms screen suppliers through a gradual learning process. For a fixed requirement by the firm each period, this in turn leads to the firm dealing with a smaller number of suppliers over time and thus a higher volume of transactions with those that remain. In a standard model, if the firms were gradually allocating more volume to firms on the basis of initial experimental stages and if those visits indicate the intensity of information gathering, then we would have expected the frequency of interaction between the parties to decrease over time. In contrast we find that the frequency of transactions increases over time. The reason we find this is because of inventory constraints. Inventory constraints might arise because of space limitations and also because of the sheer difficulty of handling and delivering a huge amount of the materials each time. Thus, given such constraints on the suppliers, the firm-supplier relationship transforms into breaking the contract into smaller units and interacting more often in a given time period. It is true that firms do provide more of the contracts to those suppliers that have been screened in the experimental stages, but the interaction is more during the later stages just because of the fact that the firms have to break the contracts and thus meet with the suppliers more in any given year.

   Thus, we study supplier selection through screening in quality provision. By rewarding good performers with future business, the use of past performance information helps develop long-term relationships with suppliers. We report patterns in the data that are consistent with this model: Both frequency and volume of transactions increase with the length of a firm's relationship with its supplier.

   Our framework very much follows in the spirit of Helper's (1990, 1991) studies of the historical development of the U.S. and Japanese automobile industries. Her work suggests that in markets for complex products and services, the behavior of buyers with market power can have a strong influence on the structure and behavior of the supplier industry. Suppliers tailor their investments and develop expertise to fit the tasks that are delegated to them by buyers. In comparing the U.S. and Japanese approaches to buyer/supplier relationships, Helper adapts a framework developed by Hirschman (1970). Helper divides buyer/supplier relationships into two categories on the basis of the buyer's response when problems arise in the relationship. One is "exit," where the buyer's response is to sever the relationship and find a new supplier; and the other is "voice," where the customer works with the supplier until the problem is corrected. In that vein we find evidence in our study such that the "exit" response, that is, tournaments, rather than continuing to work with "difficult" suppliers, is more prevalent in areas that have poor conflict resolution mechanisms. (See Bleeke and Ernst 1995 and Ernst and French 1996 for related discussions of conflict resolution.)

   This relates to the small but burgeoning literature on intermediaries who can collect and provide information about product quality. This has a moral hazard aspect (one side or the other may deliberately shirk or degrade quality to reduce cost) as well as an adverse selection component (quality may be an innate characteristic or type of one side unknown to the other). Dixit (2003) chose to model moral hazard, whereas Lizzeri (1999) focused on adverse selection. We remain agnostic in our model as to what we mean by poor quality, that is, whether firms intentionally provide bad quality or differing quality is an inherent characteristic for the firms. In the empirical section we show that both types of reasons behind bad quality are present in our sample. In any case, in the absence of such private intermediaries as mentioned above, firms might take recourse to screening by themselves to achieve maximum efficiency from the relationship. Last, our work is related to the literature on credit allocation in an asymmetric information setup, which we discuss in a later section.

   The paper is organized as follows. In section 2 we conjecture a theory for our framework. In section 3 we discuss the data for our study. In section 4 we provide our empirical results. In section 5 we discuss specific issues related to our paper and in section 6 we conclude.

2. Theoretical Conjecture

   Consider a scenario where firms need periodic supplies of materials from suppliers but they have very little or no information about the quality of the suppliers that will cater to them. Firms have a prior belief about the quality of their potential suppliers before they start doing business. Our hypothesis is the following. To start with, an individual firm would deal with the suppliers that are above their basic quality cutoff. They would break their contract requirements among the suppliers that they decide to deal with. There is a search cost for firms each time they try to hire new suppliers. Hence they will keep a select group of suppliers as opposed to one supplier at a time so that they can choose from their continuing suppliers rather than hire new suppliers regularly. As they learn gradually about suppliers' true quality with their experience in dealing with them, they will allocate the contract to the suppliers with better previous records over time. A direct consequence of the above-mentioned hypothesis is that the number of suppliers that the firm deals with over time decreases. The interaction frequency with the continuing suppliers increases because given a fixed requirement by the firm each period, a decrease in the number of suppliers means that each of them gets a...