Arbitration and Conflict Resolution in Labor- Management Bargaining
分类了强制仲裁方案的理论问题，综述已有研究，并指出未来最有前景的研究方向，帮助理解不同仲裁方案对劳资谈判的影响。
Compulsory arbitration is frequently employed to resolve labor-management bargaining disputes when the union is legally prohibited (as are, for example, many public employees' unions) from striking. In this form of arbitration, an arbitrator is empowered to impose a settlement on the bargaining parties if their negotiations break down. Various compulsory-arbitration schemes are now in use in many states, including Alaska, Connecticut, Iowa, Maine, Massachusetts, Michigan, Minnesota, Nebraska, Nevada, New Jersey, New York, Oregon, Pennsylvania, Rhode Island, South Dakota, Washington, Wisconsin, and Wyoming. But there has been little formal analysis of the various schemes that are employed in these states and, as a result, the basis available for choice among them remains incomplete. This paper classifies the theoretical problems that must be resolved before a more careful comparison of these compulsory-arbitration schemes is possible, provides a brief overview of the work that has been done on each of these problems, and indicates what appear to be the most promising directions for future research along these lines. Four kinds of compulsory arbitration are considered here: conventional ccmpulsory arbitration (CCA), in which the arbitrator imposes a settlement of his (unrestricted) choice if negotiations break down; finaloffer arbitration (FOA), in which the arbitrator must choose without compromise between bargainers' final offers if negotiations break down; multiple FOA, a variant of FOA originally suggested by Donn; and, on occasion, issue-by-issue FOA, which is like simple FOA except that the arbitrator is permitted to fashion his settlement from the components of bargainers' final offers. CCA, simple FOA, and issue-by-issue FOA are already in widespread use, while multiple FOA, which is similar but not identical to a scheme used in Eugene, Oregon, has been suggested by Donn and my 1979a article as an improvement on simple FOA. In the literature of industrial labor relations, compulsory-arbitration schemes have been judged primarily by three criteria: the quality of the arbitral settlements they generate when negotiations break down; their freedom from bias, which is usually defined as the distortion of negotiated settlements away from what they would have been in ordinary bargaining, with both strikes and lockouts permitted; and the extent to which they create environments conducive to negotiated settlements. An integrated analysis, in which bargainers choosing their stategies consider the effects of their actions on negotiated and arbitral settlements as well as on the probabilities of these possibilities, would be ideal. But in beginning the study of the effects of arbitration schemes, it is convenient, and probably not misleading, to simplify the problem by dividing it. Thus, I shall propose separate analyses of the quality of arbitral settlements, under noncooperative behavioral assumptions; the bias of negotiated settlements, under cooperative assumptions; and the probability of a negotiated settlement, under a blend of both noncooperative and cooperative assumptions. Each section of this paper in turn discusses existing work that is relevant to judging arbitration schemes by one of the above three criteria. *University of California-San Diego. This research was supported by the National Science Foundation. Many of the observations made here evolved in discussions and correspondence with Clifford Donn, who by no means agrees with all, or even most, of them. I am grateful to him, and to Joel Sobel and participants in a workshop presentation at the University of Chicago, who also made helpful comments.