سهم بازار، موافقت نامه R & D و معافیت بلاک اتحادیه اروپا

Regulation (EC) No 1217/2010 on the application of Article 101(3) of the Treaty on the Functioning of the European Union to categories of R&D agreements exempts horizontal R&D agreements from antitrust concerns when the combined market share of participants is low enough. We examine the theoretical basis for this criterion by extending existing models so that a subset of firms innovates and participates in an R&D cooperation agreement. We show that the incentive to increase innovation depends on a complex set of effects. We identify one, the outsider effect, that can lead firms to increase R&D under cooperation precisely when their combined market share is high. In a general model in which all firms innovate, we also find that R&D agreements can be more beneficial at higher market shares. We argue that existing theory therefore does not support limiting the exemption to low market shares.

Horizontal research and development (R&D) agreements, whereby firms in the same industry coordinate their R&D operations and jointly exploit the results, restrict de facto inter-firm competition in the discovery of new products or processes. However, horizontal R&D agreements can lead to economic benefits, as they permit participating firms to avoid the duplication of budgets, to combine complementary resources, and to internalize knowledge spillovers, possibly toward more innovative goods and lower production costs. The European Commission, 2010 and European Commission, 2011 recognizes that R&D cooperation can result in net economic benefits. In December 2010, it adopted Regulation (EC) No 1217/2010 on the application of Article 101(3) of the Treaty on the Functioning of the European Union (henceforth the Treaty) to certain categories of R&D agreements, together with corresponding guidelines on the assessment of horizontal cooperation agreements under the European Union competition rules (henceforth, the Guidelines). The new legislation updates and replaces Regulation No 2659/2000 and the precedent version of the Guidelines (European Commission, 2000a, European Commission, 2000b and European Commission, 2001). They had been adopted a decade earlier in order to introduce a more economic approach in the assessment of the anti- and pro-competitive effects of inter-firm R&D agreements than the previous formalistic procedures. Although it introduces several amendments, the new regulation maintains the use of a market share criterion to discriminate firms that will be exempted or not from the burden of assessing the compatibility of their contractual relationships with competition law.2 If the firms engaged in R&D cooperation have limited market power, in that their combined ex ante market share does not exceed 25%, their agreement is presumed to have no or negligible harmful anti-competitive effects, in line with a general view that innovation increases with product market competition. The Guidelines clearly assert that, in that case, the parties are more likely to transmit efficiency gains to consumers. However, above the 25% threshold the firms are excluded from the exemption. The Guidelines indicate that the parties to the agreement are then less likely to pass on efficiency gains to consumers. The firms must then proceed to the self-assessment of their agreement, as they bear the burden of proving that the restriction in competition implied by the agreement is outweighed by the R&D results and the related benefits to consumers. They also risk the annulment of the agreement together with financial penalties. A complementary view is due to Galbraith (1952), who observes that R&D activities demand large financial resources, which are not available to small firms in isolation. He sees a direct link between market shares and incentives to invest in the development of new products or processes.3 This view is also consistent with a legal framework that gives preferential treatment to the formation of R&D agreements by small firms. However, the recent theoretical and empirical literature has established that the exact connection between the intensity of market competition and R&D investments is not necessarily monotone (see Gilbert, 2006, Schmutzler, 2009 and Vives, 2008 for recent authoritative surveys). In particular, surveying oligopoly models of R&D cooperation, De Bondt and Vandekerckhove (2010) find that equilibrium R&D levels and consumer surplus can often be inverted U related with the number of cooperating parties. Overall, relatively few theoretical papers address the link between cooperative R&D and policy (Motchenkova & Rus, 2011). The objective of this paper is to assess the market share criterion used in the current European legislation in light of the relevant economic theory of R&D agreements. Using the standard industrial organization approach, we investigate whether this criterion facilitates, or penalizes, the formation of those R&D agreements which are most beneficial to consumers. Our main finding is a negative one, that is to say that the incentive for firms to increase their R&D as a result of a cooperation agreement is not linked univocally to market share. We construct a model that draws from well established contributions (Amir, 2000, Amir et al., 2003, d’Aspremont and Jacquemin, 1988, Hinloopen, 2003, Hinloopen and Vandekerckhove, 2009, Hinloopen and Vandekerckhove, 2011, Kamien et al., 1992 and Qiu, 1997, among others).4 Ex ante symmetric firms, in a Cournot oligopoly, compete on a market for substitutable products. To parsimoniously highlight the link between the market share of participants to an R&D agreement and the possible benefits from R&D, we suppose that a subset of firms in the industry may innovate and coordinate their R&D decisions. R&D activity enhances the quality of their products or reduce the marginal costs of production (satisfying the condition of “economic progress”),5 and firms in the agreement benefit indirectly from the innovative activity of others via R&D input or output spillovers. An important simplification that allows us to derive analytic results is that we assume that outside firms do not conduct R&D, although they benefit at no cost from some fraction of the activity of innovators as may occur in industries in which imitation is important. In our model, we compare R&D decisions, in the context of a cooperative R&D agreement, with those that would arise if the same firms did not coordinate. Cooperation results in more consumer surplus than competition if and only if innovating firms have an incentive to increase their R&D when they are parties to an agreement. We find no simple link between the ex ante market share of innovators and the benefits to consumers that result from coordinated R&D decisions. R&D cooperation agreements can be detrimental to consumer surplus even if the market share of participants is low, and conversely they can lead to higher consumer surplus even if the market share of participants is high. Moreover, in the particular case of so-called R&D input spillovers, we are able to show that the benefits resulting from R&D cooperation are positively related to market share. We argue that existing theory therefore does not support limiting the exemption to low market shares. The model we study is a special case of a more general one, in which all firms conduct R&D regardless of whether they participate in an agreement or not. In the general setting, many more strategic effects arise than those that we have characterized analytically, and the presumption is that the link between market shares and the effect of R&D agreements on consumer welfare should then be even more tenuous. We verify this for a numerical example, which we argue casts further doubt on any theoretical grounding for the market share exemption in the regulation. The remainder of the paper is organized as follows. Section 2 presents the regulation and discusses its implications for firms which have a large market share and contemplate participation in an R&D agreement.6 Section 3 introduces a model where a group of firms with an arbitrary ex ante market share engages in cooperative R&D. Section 4 describes the incentive for firms to conduct more R&D under cooperation, and identifies an effect, the outsider effect, that can result in cooperation agreements that are desirable only when the ex ante market share of innovating firms is large. Section 5 corroborates this finding with a numerical example in a more general model where all firms perform R&D. Section 6 concludes. Proofs are relegated to Appendix A.

Regulation (EC) No 1217/2010, and the Guidelines, reflect the laudable objective to make consumers benefit from the virtues of R&D cooperation, while filtering out arrangements that may not pass on benefits to consumers. By introducing a market share criterion, so that only firms with limited sales may collaborate in R&D within “safe harbours”, the regulation penalizes the technological collaboration of larger suppliers. When firms have a high combined market share, horizontal R&D agreements take place in a more uncertain legal environment. The parties bear the burden of self-assessing the compatibility of their contractual relationships vis-à-vis the legislation. They face the risk of incurring pecuniary sanctions, in addition to unrecoverable costs in case of infringement and subsequent automatic annulment of the agreement. Our message is that, by focusing too closely on the competition dimension via market shares, the regulation loses sight of the raison d’être of the R&D block exemption. In order to test the regulation from a theoretical economic viewpoint, we have extended standard models of R&D cooperation to a framework with cooperating innovators and imitators. For several specifications of the impact of R&D on existing products and/or processes, we have established that the connection between the combined market shares of cooperating firms and the benefit of R&D to consumers is not univocal. In fact, it can be the opposite of what the regulation seems to assume. More specifically, under R&D cooperation agreements, participating firms internalize a negative externality that we have identified as the outsider effect. In our model, this negative externality – which leads to reduced R&D under cooperation – is lower when the cooperating firms have higher ex ante market shares. When spillovers are in R&D inputs, and imitation is sufficiently strong, the likelihood that cooperation benefits consumers increases with the innovators’ combined market share. The latter is thus not a good predictor of the social desirability of agreements. Moreover, in a general model in which all firms conduct R&D, we have also found no strong support for the market share criterion. Another possible motivation for the use of a market share criterion, that we have not studied in this paper, relates to the restriction of competition on the final market (and not only in R&D efforts), as made easier by R&D cooperation. Martin, 1995 and Martin, 1997 and Cabral (2000) demonstrate that R&D joint ventures can facilitate collusion in the marketing of goods in the formal context of two-firm models. With more than two firms, if the impact on the economy is more negative when the colluding parties have high market power ex ante, then the 25% threshold could be rationalized as a precautionary mechanism that systematically puts the most harmful potential colluders under scrutiny. However, the exact connection between the ex ante combined market share of firms engaged in R&D cooperation and the likelihood that they participate simultaneously in a price fixing or quantity reducing agreement remains to be investigated. A recent empirical study by Goeree and Helland (2010) in the US environment has found that a shift in antitrust policy (the revision of the leniency program by the Department of Justice in 1993), which specifically aimed at deterring final market collusion, reduced significantly the probability that a firm joins a R&D joint venture in the telecommunications industry. This result suggests that some past R&D agreements were indeed motivated by non-technological considerations. It also indicates that collusive behavior can be dealt with successfully with dedicated instruments. In terms of policy implications, our analysis thus calls for both simplification and specialization. The simplification of the R&D block exemption by withdrawing the market share criterion would eliminate the disincentives to R&D cooperation faced by large firms. In that case the regulation would be specialized,35 with the objective of making consumers benefit from R&D cooperation, leaving aside the detection and prosecution of welfare-reducing behavior to other specialized competition rules.