Post by Keith Hylton
In my last contribution to this blog, I discussed the different implications of waiving legal rights in standard one-on-one litigation and class action scenarios. I noted predispute waivers can be socially desirable in both settings, but the danger of welfare-reducing waivers is greater in the class action scenario.
Let me take some time to elaborate here. One of the basic results of the economic theory of litigation is that the private and social incentives to litigate diverge – this point was demonstrated in an article by Steve Shavell. In other words, an individual may have an incentive to file a tort claim in a setting where society’s welfare would be greater if litigation were prohibited. This proposition does not depend on people being uninformed or suffering from various judgment biases; it holds when litigants are rational and fully informed.
However, the incentive divergence proposition depends on the existence of high transaction costs. If transaction costs are low, waiver agreements will be formed whenever the private incentive to sue is socially excessive – a result I demonstrated in an article on the economics of waiver agreements in 2000. Thus, waivers should tend to be observed in areas where they are efficient. This claim remains valid to some degree even in the context of class action waivers, but the danger of inefficient waivers is considerably greater in that setting.
What about arbitration? Do these claims about the social welfare implications of waiver agreements apply to arbitration as well? The answer is yes.
As I noted in my previous post, a predispute waiver agreement is socially efficient whenever the deterrence benefit from litigation is less than the expected cost of litigation. When the deterrence benefit is less than expected litigation costs, the potential injurer and potential victim will have an incentive to enter into a predispute waiver agreement.
The same statement can be applied to arbitration agreements. An arbitration agreement alters the deterrence benefit from litigation. It could enhance the benefit if arbitration punishes wrongdoers more reliably, exonerates the innocent more reliably, or does both more reliably. However, even if arbitration does not enhance the deterrence benefit from litigation (e.g., suppose the arbitrator is the injurer’s brother), it may still benefit the parties jointly if it reduces the costs of litigation – and it is easy to see how this might occur if arbitration relaxes some of the procedural obstacles associated with ordinary litigation.
So the economics of waiver and arbitration agreements are pretty much the same, whether in the one-on-one litigation scenario or in the class action scenario. Whether greater use of arbitration in a specific area of dispute resolution is socially desirable is at bottom an empirical question (see, e.g., Christopher Drahozal’s empirical analyses of arbitration).
Although one might expect arbitration agreements to have swept over the private economy given the costliness of litigation, this has not happened. There are some areas where potential injurers and victims continue to rely on traditional litigation even though they could easily enter into arbitration agreements. For example, banks have been slow to embrace arbitration. Arbitration agreements often carve out some types of legal disputes, such as intellectual property. The reason is that the reduction in dispute resolution costs (from opting for arbitration) is not large enough in these areas to offset the dilution in the deterrence benefit provided by court-monitored law. These are empirical issues, and cannot be resolved on the basis of sweeping claims about the social desirability of predispute arbitration.