Published Articles (selection)
- Vencatachellum, D., and Versaevel, B., 2009, "R&D Delegation in a Duopoly with Spillovers", B.E. Journal of Economic Analysis & Policy (Contributions), 9(1), Art. 55.
- Vencatachellum, D., and Versaevel, B., 2008, "Horizontal R&D Cooperation and Spillovers: Evidence from France"., Economics Bulletin, 15(19), pp. 1-11.
- Billette de Villemeur, E., and Versaevel, B., 2003, "From Private to Public Common Agency", Journal of Economic Theory, 11(2), pp. 305-309.
- Versaevel, B., 2002, "Coordination Costs and Vertical Integration in Production Franchise Networks: A Common Agency Model", Research in Economics, 56(2), pp. 157-186.
Other Contributions
Ruble, R. et Versaevel, B., 2010, "Mise en oeuvre de la collusion et détection : approches actives et passives", Revue Lamy de la Concurrence, 22, pp. 11-17.
Versaevel, B., "La Description des Marchés Concernés", pp. 92-107 in : Ferrier, D. et Ferré, D., 2004, Droit du Contrôle National des Concentrations, collection Référence, Dalloz : Paris.
Working Papers
- Billette de Villemeur, E., R. Ruble, and B. Versaevel, "On the Timing of Vertical Relationships" (this version: April 2011).
We show that the standard analysis of vertical relationships transposes directly to investment timing. Thus, when a firm undertaking a project requires an outside supplier (e.g. an equipment manufacturer) to provide it with a discrete input, and if the supplier has market power, investment occurs too late from an industry standpoint. The distortion in firm decisions is characterized by a Lerner index, which is related to the parameters of a stochastic downstream demand. When feasible, vertical restraints restore efficiency. For instance, the upstream firm can induce entry at the correct investment threshold by selling a call option on the input. Otherwise, competition may substitute for vertical restraints. In particular, if two firms are engaged in a preemption race downstream, the upstream firm sells the input to the first investor at a discount that is chosen in such a way that the race to preempt exactly offsets the vertical externality, and this leader invests at the optimal market threshold.
- Billette de Villemeur, E., L. Flochel, L. and B. Versaevel, "Optimal Collusion with Limited Liability Constraint and Policy Implications" (this version: July 2011).
Collusion sustainability depends on firms’ aptitude to impose sufficiently severe punishments in case of deviation from the collusive rule. We extend results from the literature on optimal collusion by investigating the role of limited liability. We examine all situations in which either structural conditions (demand and technology), financial considerations (a profitability target), or institutional circumstances (a regulation) set a lower bound, possibly negative, to firms’ profits. For a large class of repeated games with discounting, we show that, absent participation and limited liability constraints, there exists a unique optimal penal code. It commands a severe single-period punishment immediately after a firm deviates from the collusive stage-game strategy. When either the participation constraint or the limited liability constraint bind, and the latter is not too strong, there exists an infinity of multi-period optimal penal codes that permit firms to implement the collusive strategy. We establish that a longer punishment is a perfect substitute for more immediate severity only when (i) the participation constraint would bind if there were no limited liability constraint, and (ii) a firm cannot break even by deviating from the most severe admissible punishment. In all other cases the lowest discount factor that permits the implementation of collusion is strictly higher than without the limited liability constraint. Hence limited liability hinders collusion. The policy lesson is that limiting further firms’ liability may actually result in more efficient markets.
- Ruble, R. and B. Versaevel, "Market Shares, R&D Agreements, and EU Competition Policy" (this version: Decembre 2010).
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 shares of participants are low enough. We show that existing theory does not support limiting the exemption to low market shares. Building on standard models and assuming that a subset of firms invests in R&D whereas others imitate, we assess the link between the combined market share of innovating firms and the product market benefits of R&D cooperation. With R&D output choices, the market share criterion, while helping to rule out the most socially harmful R&D cooperation agreements, also hinders the most beneficial ones. With R&D input choices, R&D cooperation is more likely to harm consumers at lower combined market share. In so-called “RJV cartels”, consumers always benefit from R&D cooperation and relative consumer surplus increases monotonically with the number of innovating firms and decreases with industry size when imitation is sufficiently strong..
- Ruble, R., Versaevel, B., 2008, "On the Tacit Collusion Equilibria of an Investment Timing Game".
This note further characterizes the tacit collusion equilibria in the investment timing game of Boyer, Lasserre and Moreaux (2007). Tacit collusion equilibria may or may not exist, and when they do may involve either finite time investments (type 1) or infinite delay (type 2). The relationship between equilibria and common demand forms is not immediately apparent. We provide the full necessary and sufficient conditions for existence. A simple condition on demand primitives is derived that determines the type of equilibria. Common demand forms are then shown to illustrate both finite-time and infinite-delay tacit collusion.
- Versaevel, B., 2009, "Cumulative Leadership and Entry Dynamics".
This paper investigates the combined impact of a first-mover advantage and of firms' limited mobility on the equilibrium outcomes of a continuous-time model adapted from by Boyer, Lasserre, and Moreaux (2007). Two firms face market development uncertainty and may enter by investing in lumpy capacity units. With perfect mobility, when the first entrant plays as a Stackelberg leader, in equilibrium it invests earlier, and the follower later, than in the Cournot benchmark scenario. There is rent equalization, and the two firms' equilibrium value is lower. This result is not robust to the introduction of firm-specific limited mobility constraints. If one firm is sufficiently less able than its rival to mobilize resources at early stages of the market development process, there is less rent dissipation,and no equalization, in a constrained preemption equilibrium. The first-mover advantage on the product market then results in more value for the less constrained firm, and in less value for the follower than when they play in the Cournot manner with perfect mobility. The leading firm maximizes value by entering immediately before its constrained rival, though later than made possible by its superior mobility. Greater uncertainty reduces the value differential to the benefit of the follower. It also increases the distance between the firms' respective investment triggers. The specifications and results are discussed in light of recent developments in the market for music downloads.
- Dequiedt, V. and B. Versaevel, 2008, "Patent Pools and the Dynamic Incentives to R&D".
Patent pools are cooperative agreements between several patent owners to bundle the sale of their respective licenses. In this paper we analyze their consequences on the speed of the innovation process. We adopt an ex ante perspective and study the impact of possible pool formation on the incentives to innovate. Because participation in the creation of a pool acts as a bonus reward on R&D activity, we show that a firm's investment pattern is upward sloping over time before pool formation. The smaller the set of initial contributors, the higher this effect. The ex ante perspective leads to policy implications. It reveals that patent pools encourage high R&D efforts, in comparison to a joint profit-maximizing benchmark, by early innovators. Patent pools can thus be used as a stimulating instrument for the generation of the first patents when innovation is sequential in nature. We also find that, when solicited by firms in advance of the pool foundation, a forward-looking regulator has an incentive to delay the clearance date. This can result in a suboptimal pool size, as compared with the social welfare optimum. This questions the relevance of a review process by competition authorities before a pool forms.
- Billette de Villemeur, E., and Versaevel, B., 2006, “Efficient R&D Delegation”.
This paper constructs a model of R&D delegation, in which several firms may outsource new technology from the same for-profit laboratory. When R&D costs are uncertain at the contracting stage, necessary and sufficient conditions are provided that characterize Nash equilibria in the laboratory's outputs and the firms' non-negative transfer payments. A subset of these equilibria is identified in which firms' payments truthfully reflect their valuation of all possible alternatives vis-´a-vis the expected net profits. These truthful Nash equilibria implement the first-best outcome when the support of the stochastic parameter is not too large, in which case firms' net payoffs do not depend on the stochastic component and are undominated. When externalities impact the nature of competition among firms, both on the intermediate market for R&D services and the final market for goods, sufficient conditions are given for the laboratory to earn positive benefits or not, and for firms to choose either to cooperate horizontally or to acquire the laboratory. The latter two options are shown not to impact the R&D outputs supplied by the laboratory, nor the joint profits earnt by firms. A first policy implication is that no new regulatory tool is needed for aligning firms' interests with a social welfare objective. The results also support legal environments which do not prevent firms to include discriminatory clauses in their R&D contracts on the intermediate market for new knowledge, including in highly concentrated industries. This laissez faire message does not extend to final-market antitrust considerations.
- Vencatachellum, D., and B., Versaevel, 2004, "R&D Delegation in a Bertrand Duopoly with Spillovers".
We provide an extensive comparison of three cost-reducing R&D games where duopolistic firms (i) cooperatively conduct in-house R&D, (ii) non-cooperatively choose in-house R&D, and (iii) delegate R&D to an independent profit-maximizing laboratory. Firms are assumed to behave à la Bertrand on the final market. We establish that delegating R&D to an independent laboratory is Pareto optimal when between-firm technological spillovers are not too high and the production of R&D services by the laboratory, for the two firms, is complementary.
- Billette de Villemeur, E., and Versaevel, B., 2003, “Conflict and Cooperation on R&D Markets”, IDEI Working Papers series.
We investigate the distribution of profits between a laboratory and two firms on the intermediate market for cost-reducing or/and demand-enhancing technology, and infer implications for the governance of R&D. The laboratory supplies tailor-made multi-dimensional R&D services at some costs, and maximizes its individual profits. Firms are interested in delegating the production of R&D services from a common laboratory, to whom they offer contingent payment offers. On the final market for goods, firms compete in prices or in quantities. We unveil mild sufficient conditions for the (non) appropriation of innovation profits by the laboratory. Anti-complementarities in the dimensions of R&D services imply that the laboratory appropriates some non-zero share of joint profits in all equilibria. In this case, we show that each firm has strategic incentives to shift to a more integrated structure by merging horizontally or by acquiring the laboratory. Complementarities in the development dimensions imply that the laboratory exactly breaks even in all equilibria. In that case, firms have no incentive to shift to a more integrated governance structure. A series of specific algebraic forms, as borrowed from the literature, illustrate the broad applicability of the results, and uncovers common features in seemingly unrelated representations of postinnovation cost and demand conditions.
- Gagné, R., Van Norden, S., and Versaevel, B., 2007, “Testing Optimal Punishment Mechanisms Under Price Regulation : the Case of the Retail Market for Gasoline”.
Following a severe price war in the retail market for gasoline in 1996, the Quebec provincial government introduced a price floor. We analyse the effects of this new regulation on the pricing behaviour of a sample of gas stations in Montreal over the 1994-2002 period. We use a Markov Switching Model with two latent states to simultaneously identify the periods of low/high margins and estimate the parameters characterizing each state. We find that the net effect of the price floor on average margins is near zero, as the impact on retail prices in low margin periods is offset by the rise in the average duration of these periods. The increase of the duration of low margin periods lends supports to the presumption of collusive pricing behavior.
- Versaevel, B., 2006, "The Hierarchical Firm Resolves Empty Cores: A Coasean Model and Two Examples".
This note proposes a clarification of the definition of the Coasean hierarchical firm. A simple model in the theory of the core is constructed that formalizes the optimization problem underlying the industrial organization à la Coase. Two simple specific examples illustrate the proposition that empty core problems are more fundamental determinants of the real-world organization of production than the various phenomena that can lead to them, including "transaction costs" and "social costs". In the two cases, the Coasean firm is rationalized as an institutional remedy that resolves an empty core problem by eliminating a selection of potential market transactions. It does not form if the core is non empty, if an empty core is not resolved, or if it is resolved by some alternative device that does not resort to authority of command.