On Synergies And Vertical Integration

. Introduction

When trying to explain the reasons for the large number of mergers among both multinationals and small specialized businesses in recent years, the realization of potential synergies among the merging firms is often invoked.1 In particular, it is usually claimed that when integrated a supplier of inputs will adapt his technology in a much higher degree to the needs of his customer than when he is separately owned. In this paper we focus on vertical mergers and analyze the impact of ownership structures on the realization of synergies. Hence, a key question is why two separate firms might be unable to implement the same degree of relationship specificity as they would choose if they were commonly owned.

For example, consider a supplier of certain car parts who has to make investments in order to produce a specific input for an automobile manufacturer. Suppose that there is only one asset used in the production of the input, namely the supplier’s plant. Following the seminal contribution of Grossman, S.J. and Hart, O.D., 1986. The costs and benefits of ownership: A theory of vertical and lateral integration. Journal of Political Economy 94, pp. 691вЂ"719. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (928)Grossman and Hart (1986), ownership is defined as residual rights of control over the use of assets. Provided that complete state-contingent contracts cannot be written, the owner of an asset can threaten to withdraw the asset from a relationship in order to use it otherwise if a contingency arises which was not specified in the initial contract. Asset ownership can hence improve investment incentives of a party that has to make unverifiable investments that are worthless without the asset. In our example, if the automobile manufacturer does not have to make any significant relationship-specific investments, the GrossmanвЂ"HartвЂ"Moore (GHM) analysis recommends that the investing party, i.e. the supplier, should own the plant.2

In this paper we want to emphasize that the investing party’s decision in practice is more complex than simply deciding on the amount of effort invested in production. In particular, it has to be decided exactly how to spend the invested amount and how to design the production technology adopted. One important aspect in this context is the degree of specificity with regard to the needs of the buyer. For example, the supplier might choose highly specific machines or workers’ training adapted to the buyer’s needs. In this sense the supplier has an impact on the degree to which synergies may be realized within the relationship. Of course, if the input is only produced to be used by the buyer, the realization of the complete potential of synergies is optimal and would be accomplished in a first-best world. But the more specificity the seller chooses, the higher his dependency on the buyer and the stronger the threat of being held-up if contracts are incomplete.

In the first part of the paper we follow the GHM framework and assume that only asset ownership can be contractually determined. This is meant to capture the idea that specificity, much like investment in the sense of effort, often may not be verifiable by the courts. It is shown that this might lead to suboptimal adaptation and therefore to a failure to realize the entire potential synergies, given that firms are not integrated. However, if firms are vertically integrated, i.e. if the buyer owns the supplier’s plant, no such effect arises and hence all gains from specificity are realized. Hence, if besides the investment decision the supplier also makes a specificity choice, the optimality of non-integration in the illustrative example is no longer clear. The following trade-off arises: While vertical integration leads to full specificity but underinvestment, non-integration entails a loss in specificity but might provide better investment incentives.

The property rights literature has been motivated by the search for an answer to the question “Why is not all production carried on by one big firm?” which was formulated by Coase (1937, p. 394) and has more recently become known as the “Williamson puzzle”.3 Indeed, one might say that the main focus was on finding a convincing reason for non-integration. Yet, Williamson (1985, p. 105) also emphasizes that “the orthodox theory of the firm has no explanation for why successive stages in the core technology should be under unified ownership rather than each owned autonomously”. In the following sense this paper helps to provide an answer to this problem: It is reasonable to believe that successive stages in the core technology of a production process are characterized by a high degree of asset specificity and mutual adaptation. Here it is argued that non-integrated firms might not want to adapt too well in fear of being held-up, and this makes integration of those stages in one firm superior.4

In the second part of the paper we briefly analyze what happens if both parties can contractually agree on the level of specificity. In this case, asset ownership only serves to provide incentives to exert effort, so that in accordance with the analysis of Hart and Moore (1990) and Hart (1995) it turns out that non-integration will always be preferred given that only the supplier has to make an investment decision. Yet, it will be shown that even though the specificity decision may now be contractually determined and enforced, and even if it is always optimal to produce only for the buyer’s needs and therefore full specificity would be chosen in a first-best world, both parties will deliberately agree on less specificity since this improves the supplier’s investment incentives and hence increases total surplus.

In view of the emphasis that has usually been put by practitioners on synergies as an important factor in determining ownership structure and scope of the firm, it is surprising that this issue has so far largely been neglected in the economic literature on the theory of the firm.5 However, some extensions of the well-known GHM framework are related to our model. Rajan and Zingales (1998) discuss a version of the basic GHM model in which investment increases the value of the traded good within the relationship but decreases the outside value, so that ownership can reduce the incentives to make specific investments. However, their assumption that effort decreases the outside value of the traded good may be hard to justify in general. In this paper the supplier’s decision is split up in two components: the investment and the specificity choice.6 HolmstrÐ"¶m and Tirole (1991) also