Experiments in Majoritarian Bargaining

立法谈判的Baron-Ferejohn模型(BF)是立法政治研究中应用最广泛的形式框架之一。该模型的变体已被用于研究立法投票规则、委员会权力、猪肉桶计划、政府组建、多党选举和众议院之间的谈判。在考虑一个简单版本的模型中，其中有三个政党，没有一个政党在立法机构中拥有多数选票，BF模型预测，拥有提案权力的政党将提出一个最小的获胜联盟。这将由它和另一个成员组成，第三方没有任何成员。正如在鲁宾斯坦模型中一样，提议的政党将给予其提议的联盟伙伴足够的金额，以确保获得接受。如果提议被拒绝，讨价还价继续进行，金额(或延续价值)等于联盟伙伴的预期回报。因此，提案总是在第一轮中被接受。调查人员指出，提议的政党总是选择延续价值最低的政党作为其联盟伙伴。一般来说，战利品的分配将是高度不平等的，特别是如果各方的贴现因素很低的话。研究小组在受控的实验室环境中进行了一系列实验，测试高炉模型的预测。初步研究表明，BF模型的预测很少得到支持。特别是，提出者经常分配给所有三个参与者，在两人联盟中，分配比例比预期的更高。在这个项目中，研究人员从理论和经验两方面调查了预测失败的原因。该项目的第一个目标是检查研究结果对游戏形式变化的稳健性。在初步实验中，使用了五个和四个周期模型。例如，有人可能会反对，认为这些模型对受试者要求太高，无法完全理解。因此，当务之急是用更少的时间，在一个更简单的游戏中复制他们的发现。因此，研究人员对游戏的两个和三个时期版本进行了新的治疗。下一步，研究人员打算调查可能的理论解释，以解释早期的平等分裂发现，然后在实验室中进行测试。在对这些理论账户的调查中，他们进行了额外的实验，在某些情况下，还进行了计算机模拟和理论研究。他们侧重于对结果的三种可能的解释：1)存在非平凡错误的均衡概念，2)公平或利他主义准则，3)惯例和焦点。由于Baron-Ferejohn模型是理解立法多数派讨价还价的一种广泛使用的方法，拟议的研究应该有助于理解该模型的局限性，从而有助于理解立法多数派讨价还价。

The Baron-Ferejohn model (BF) of legislative bargaining is one of the most widely used formal frameworks in the study of legislative politics. Variants of the model have been used in the study of legislative voting rules, committee power, pork-barrel programs, government formation, multi-party elections, and inter-chamber bargaining. In considering a simple version of the model where there are three political parties with no party having a majority of the votes in the legislature, the BF model predicts that the party with proposal power will propose a minimal winning coalition. This will consist of it and one other member, leaving the third party with zero. As in the Rubinstein model the proposing party will give its proposed coalition partner just the amount necessary to secure an acceptance. The amount (or continuation value) equals the coalition partner's expected pay-off if the proposal was rejected and the bargaining continued. Proposals are thus always accepted in the first round. The investigators note that the proposing party will always choose as its coalition partner the party with the lowest continuation value. The division of spoils will in general be highly unequal, especially if the parties' discount factors are low. The research team has conducted a series of experiments testing the predictions of the BF model in a controlled laboratory environment. The preliminary research shows that the BF model's predictions are rarely supported. In particular, proposers often allocate to all three players and, in two-player coalitions, allocate more proportionally than predicted. In this project the researchers investigate why the predictions fail, both theoretically and empirically. The first goal in the project is to examine the robustness of the findings to changes in the game form. In the preliminary experiments five and four period models used. One may object, for instance, that these models are too demanding for subjects to fully comprehend. the first priority is thus to replicate their findings in a simpler game with fewer periods. Thus, the researchers conduct new treatments with two and three period versions of the game. Next, the researchers intend to investigate possible theoretical accounts for earlier findings of equal splits and then test them in the laboratory. In their investigation of these theoretical accounts they conduct additional experiments and, in some cases, computer simulations, as well as the theoretical research. They focus on three possible explanations for the results: 1) equilibrium concepts with non-trivial error, 2) norms of fairness or altruism, and, 3) conventions and focal points. Since the Baron-Ferejohn model is a widely used approach to understanding legislative majoritarian bargaining, the proposed research should help in understanding the limits of the model and, legislative majoritarian bargaining as a consequence.