Evolution, Learning, and Economic Institutions

研究的基本现象是社会学习，以及社会学习如何内生地产生社会制度和规范。 社会学习发生在大量人群中的个体随机互动，然后根据他们过去的经历和人群中其他人的行为来调整他们的行为。 重点是这样一个过程的动力学，以及随着时间的推移出现的行为模式。 该项目特别感兴趣的情况下，这些聚集模式对应于可识别的经济和社会行为模式，如经济合同的形式，政府市场交易的规则，以及社会交往中的行为规范。 第一个也是主要的目标是将进化博弈论的预测与特定类型的经济契约的发展进行比较。 这项研究的重点是委托人和代理人之间的激励合同，特别是真实的房地产佣金，意外事故费的法律，并在农业作物共享合同，为时间序列和横截面数据存在。 这项研究的出发点是以前的工作的主要研究者的进化特性的分配谈判游戏。 这个框架是通过明确建模代理的机会成本，通过将不确定性纳入输出函数，通过允许代理讨价还价的多个维度的合同，并通过提供一个更详细的处理过程中，他们收集信息。 总的目标是展示如何在这些类型的合同的惯性，变化和区域变化的模式可以通过进化的论点来说明，并显示这些模式与新古典均衡模型预测的模式有何不同。 第二部分研究了进化选择结果在不同的学习规则下的鲁棒性，特别是它们的合理性程度。 第三部分考察了这样一种观点，即社会和经济互动的结构--游戏规则--也可以被解释，至少部分可以用进化论的观点来解释。 有证据表明，有效的制度比无效的制度更稳定（从进化的意义上讲）;此外，它们的收益表现出集中趋势效应--它们倾向于位于可行收益集的中间，而不是极端。 这一结果不仅对经济制度的形式（例如，形式的合同），但更普遍的是不同形式的社会组织的福利属性，并与哲学和博弈论文献中的分配正义的突出思想相联系。

The basic phenomenon studied is social learning, and the ways in which social learning generates social institutions and norms endogeneously. Social learning occurs when individuals in a large population interact randomly, and then adjust their behavior based on information about their past experiences and what others in the population are doing. The focus is on the dynamics of such a process, and on the aggregate patterns of behavior that emerge over time. The project is particularly interested in situations where these aggregate patterns correspond to recognizable economic and social patterns of behavior, such as forms of economic contracts, rules governing government market transactions, and norms of behavior in social interactions. The first and primary objective is to compare the predictions of evolutionary game theory with the development of specific kinds of economic contracts. The study focuses on incentive contracts between principals and agents, particularly real estate commissions, contingency fees in the law, and crop sharing contracts in agriculture, for which both time-series and cross-section data exist. The starting point of this research is previous work of the principal investigator on evolutionary properties of distributive bargaining games. This framework is extended by explicitly modeling agents' opportunity costs, by incorporating uncertainty into the output function, by allowing the agents to bargain over multiple dimensions of the contract, and by giving a more detailed treatment of the process by which they gather information. The general goal is to show how patterns of inertia, change, and regional variations in these kinds of contracts can be illuminated by evolutionary arguments, and to show how these patterns differ from those predicted by neoclassical equilibrium models. The second part examines the robustness of evolutionary selection results under different specifications of the agents' learning rules, and in particular their degree of rationality. The third part examines the idea that the structure of social and economic interactions in general -- the rules of the game -- can also be explained, at least in part by evolutionary arguments. It is conjectured that efficient institutions are more stable (in an evolutionary sense) than inefficient ones; moreover their payoffs exhibit a central tendency effect -- they tend to lie toward the middle of the feasible payoff set as opposed to the extremes. This result has implications not only for the form of economic institutions (e.g., forms of contracts), but more generally for the welfare properties of different forms of social organization, and connects with prominent ideas of distributive justice in the philosophy and game theory literatures.