Holding Idle Capacity to Deter Entry

Abstract

In this JOURNAL, Dixit (I980) carefully analysed the extent to which an established firm facing a prospective entrant might make an investment in capacity beyond that which is optimal for its pre-entry output. In his model, both the established firm and prospective entrants understand that a quantity setting Nash equilibrium would be established in any post-entry game. Dixit showed that Spence (I977) had studied an imperfect equilibrium (that is, an equilibrium that depended upon a threat that was not credible) to obtain his result that firms might hold excess capacity to deter entry. In a perfect equilibrium, a firm would not wish to install capacity that would be left idle if entry did not occur. Dixit's conclusion, however, depends on his assumption that each firm's marginal revenue is always decreasing in the other's output.' This is quite a restrictive assumption. For example it is never satisfied in the relevant range for economists' second-favourite demand curve constant elasticity demand. The basic point is unchanged if the firms anticipate price competition with differentiated products rather thanquantity competition in any post-entry game: Under quite plausible conditions, rational firms facing rational potential entrants may install capacity that will certainly be left idle. We are thus able to rehabilitate Spence's original proposition in a perfect equilibrium setting.