R&D and market sharing agreements

Among the several types of collusion (price collusion, bid rigging…), a scant attention has been given to market sharing agreements, whereby partners refrain from entering in each other’s territory. A market sharing agreement might strongly bolster the competitive advantages generated by an R&D alliance. Firms involved in an R&D alliance are more efficient than the firms outside the alliance in terms of production cost[1]. However, the R&D partners continue competing among them, in the absence of a market sharing agreement. By forming a market sharing agreement, the most efficient firms will only compete with the less efficient firms outside the R&D alliance. Therefore, the competitive advantages generated by the R&D alliance are amplified by the existence of the market sharing agreement. Despite these observations, no work has jointly analyzed the formation of R&D alliances and market sharing agreements in an oligopoly environment. Our paper aims at filling this gap by identifying the R&D and market sharing agreements emerging in the long run.

Many empirical evidences support the fact that some firms are members of these two types of agreements. Members from the Vitamin cartel, detected in the late 1990s, are currently involved in different R&D agreements focused on pharmaceutical research. Moreover, several members of the market sharing agreement on the power transformers’ industry participated in R&D agreements on research closely related to power transformers before, during and after the market sharing agreement. Duso, Röller and Seldeslachts (2014) provide statistical evidence showing that the size of the R&D agreements does matter in the formation of collusive agreements.

Understanding the interplays between the formation of R&D alliances and market sharing partnerships will help the antitrust authorities to better assess the efficiency of the policies promoting R&D agreements. Furthermore, this work identifies several critical markets for the existence of market sharing agreements. The results obtained in this paper could guide the antitrust authorities in the detection of market sharing cartels.

Framework

We provide a game theoretic framework where firms can form R&D alliances and market sharing agreements before deciding the quantities they will sell on each market. Each firm is able to reduce its cost by forming an R&D alliance with other competitors. The reduction of the marginal cost of production for one firm is proportional to the number of firms in its alliance. Each firm is initially present on one market (its home market) and can enter all foreign markets. We model market sharing agreements as coalitional agreements preventing its members to compete with each other in their respective home market. However, the members of the market sharing agreement are still competing with each other on the remaining markets. Once the R&D alliance structure and the market sharing agreements have been formed, firms compete in quantities in every market where they are active.

We analyze the R&D alliance structures and the market sharing agreements that one might expect to emerge in the long run by examining the equilibrium requirement that no group of firms benefits from altering the R&D alliance structure (given the market sharing agreements) or the market sharing structure (given the R&D alliance structure). We characterize the set of stable pairs of coalition structures with identical R&D and market sharing structure.

Critical markets

In an industry where all the firms compete with each other in all markets, at most two R&D alliances can emerge at equilibrium because the two smallest alliances in the industry have always incentives to merge.[2] However, as soon as firms can establish market sharing agreements, some firms would no longer have access to all the markets. In that situation, the merger between the two smallest R&D alliances might no longer be profitable. Consider a market only composed of firms from the two smallest R&D alliance. In this market, the merger between the two smallest R&D alliances makes all the competitors equally efficient relative to each other. Thus, the merger of the two smallest R&D alliances could be not profitable for some of its members. Therefore, under some market sharing agreements, more than two R&D alliances can be formed in the industry. Accordingly, our results highlight that market sharing agreements are more likely to exist in industries containing more than two R&D alliances.

We also show that a market sharing agreement of moderate size can only be formed among sufficiently heterogeneous firms in terms of R&D efficiency. Otherwise, similar efficient firms will prefer to exit the market sharing agreement in order to have access to every market of the economy. Therefore, industries with a high heterogeneity among competitors in terms of R&D efficiency are fertile grounds for the emergence of market sharing agreements of moderate size.

Market sharing agreements do stabilize R&D partnerships

We find that, in the absence of market sharing agreements, there is no stable R&D alliance structure in the oligopolistic industry. R&D alliance structures are stable, only in the presence of some market sharing agreement. Therefore, our model highlights that market sharing agreements foster the formation of R&D agreements. Meaning that the presence of stable R&D alliances in an industry signals the existence of some market sharing agreements.

This work could be seen as a first attempt of developing specific models where agents (firms) can be involved in the formation of several types of coalitions (overlapping coalitions). Moreover, the results obtained in this paper could be useful for the antitrust authorities. First, we shed light on one determinant for the stability of R&D alliances in the long run: the presence of some market sharing agreements. Second, we identify industries that are likely to host market sharing agreements in the long run: (i) industries with a high heterogeneity among competitors in terms of R&D efficiency, and (ii) industries with more than two stable R&D alliances.

References

• Bloch, F. (1995). Endogenous Structures of Association in Oligopolies. The RAND Journal of Economics 26, 537-556.
• Cooper, R. W. and Ross, T. W. 2009. Sustaining Cooperation with Joint Ventures. Journal of Law, Economics, & Organization 25, 31–54.
• Duso, T., Röller, L.-H., and Seldeslachts, J. (2014). Collusion through joint R&D : An empirical assessment. Review of Economics and Statistics 96, 349–370.
• Miyagiwa, K. 2009. Collusion and research joint ventures. Journal of Industrial Economics 4, 768–784.

Jérôme Dollinger  is a a PhD candidate in Economics at UCLouvain. He obtains a Master degree in development economics from TSE and a Master degree from UCLouvain. He is actually conducting a thesis on the interrelation and coevolution of coalitions.
Fields of interest: Game theory, coalition theory, group formation, and industrial organization.

Ana Mauleon is Research Director, Fonds de la Recherche Scientifique (FNRS), Belgium; Professor, UCLouvain Saint-Louis, Brussels, Belgium; Faculty Member of CORE, UCLouvain; Affiliated member of the Swiss Center for Data and Network Sciences, Universität Freiburg. 
She is Associate Editor, of Mathematical Social Sciences journal.
Fields of interest: Game theory, bargaining and matching theory, group formation and network theory, and industrial organization.

Vincent Vannetelbosch is Senior Research Associate, Fonds de la Recherche Scientifique (FNRS), Belgium; Professor, UCLouvain, Belgium; Associate Fellow of CEREC, UCLouvain Saint-Louis; Director of the European Doctoral Program in Quantitative Economics; Affiliated member of the Swiss Center for Data and Network Sciences, Universität Freiburg.
He is Associate Editor of Social Choice and Welfare journal.
Fields of interest: Game theory, bargaining and matching theory, group formation and network theory, and industrial organization.