Optimal Asymmetric Strategies in Research Joint Venturesby

This paper identiies an overlooked implication of models of research joint ventures initiated by d'Aspremont and Jacquemin (1988). Even though the aggregate R&D cost of identical rms in a research joint venture would be lowest if they invested equally to reduce subsequent production costs, nonetheless members may often enlarge their overall joint proot by making unequal investments. Such a strategy raises costs in the investment stage but may create more than oosetting beneets in the production stage since industry proots are larger there when the rms are of unequal size. Joint-proot maxi-mization is an appropriate objective when sidepayments are costless. We show that it may also be appropriate in the absence of sidepayments if members of a joint venture participate in multiple markets and the optimum in each market requires asymmetric investment. In that case, maximization of joint proot in each market coupled with the appropriate assignment of the asymmetric roles in each market to the cartel members may substitute for sidepayments. In their analysis of the consequences of research joint ventures, Claude d'Aspremont and Alexis Jacquemin (1988) deduce from the hypothesis of joint-proot maximization the behavior of two rms which are allowed to coordinate investments in cost-reducing R&D but must remain noncooperative rivals in the product market. Their paper has stimulated further work by De Bondt and Veugelers Since rms are assumed identical and the cost of investing at each rm is assumed to be strictly convex, all contributors to this literature have restricted their search for the investment combination 1