Doctoral Dissertation Research in Economics: Do Firms Indeed Pursue Profit Maximization? An Experiment

利润最大化是经济学家用来模拟企业行为的一个简单工具；这种利润最大化的假设在经济学中很普遍。然而，利润最大化是否是描述企业行为的唯一方法值得怀疑。在现代经济中，除了家族企业外，许多企业的所有权和管理权是分开的。所有者通常是股东，他们从自己的公司获得股息。经理，如CEO和工人，根据他们的工作业绩从公司获得工资，这些业绩表现为利润和销售额、收入、市场份额或其他量化指标。例如，假设一个经理的工资是由他的销售额决定的。然后，他会专注于增加销售，这与经理的工资是根据公司的利润来决定的情况不同。换句话说，企业的行为取决于经理人的激励是如何决定的。管理者（由所有者委派）做出与公司运营相关的决策，这并不一定会导致利润最大化的行为。虽然之前也有一些关于战略委托的研究，但大多数经济学研究都假设利润最大化，没有考虑战略委托。因此，本研究将考察企业在战略委托下的行为与纯利润最大化下的行为有何不同。本实验研究采用两阶段古诺双寡头博弈来描述市场中的战略委托（Vickers, 1985; Fershtman and Judd, 1987; Jansen et al., 2007）。在实验中，市场上有两家公司。每个公司由一个所有者和一个经理组成。在第一阶段，两个企业所有者都选择了一个决定管理者激励的契约：利润授权（只有利润）、收入授权（利润和加权收入）和市场份额授权（利润和加权市场份额）。在第二阶段，管理者选择是否串通。在战略委托下，所有者和管理者的选择将决定市场产出。理论上，战略委托下的市场结果比利润最大化下的市场结果更具侵略性。因此，本研究将探讨理论结果是否在战略委托下成立。这项研究将有助于理解企业的行为比之前分析的更具/更少的竞争力。例如，Spierdijk和Zaouras（2017）表明，如果企业不实现利润最大化，由利润最大化假设获得的勒纳指数不适合描述市场力量。作为上述研究的延伸，本研究将使用实验来识别企业的非利润最大化行为。此外，该项目将有助于改进结构模型。自从BLP模型（Berry, Levinsohn, and Pakes, 1995）被提出以来，需求估计已经得到了提升，但该模型中的“供应”部分仍然假设利润最大化，从而导致低估边际成本。因此，实验研究结果可能有助于改进结构模型。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Profit maximization is a simple tool that economists use to model firms' behaviors; and this assumption of profit maximization is pervasive in economics. However, it is doubtful that profit maximization is the only way to describe firms’ behaviors. In a modern economy, ownership and management are separated for many firms, except for family-owned businesses. Owners are usually shareholders, and they get dividends from their own companies. Managers, such as a CEO and workers, get salaries from companies based on their working performances, indicated as profit and sales, revenue, market shares, or other quantitative indices. For example, suppose a manager’s salary is decided by his sales. Then, he will focus on increasing sales, which leads to a different outcome compared to the case where the manager’s salary is decided based on the firm’s profit. In other words, firms’ behaviors depend on how managers’ incentives are determined. Managers (delegated by owners) make decisions related to firms’ operations, and this does not necessarily result in profit-maximizing behavior. Though there was some research about strategic delegation before, most economic research has assumed profit maximization, without considering strategic delegation. Therefore, this research will examine how firms’ behaviors under strategic delegation are different from those obtained under pure profit maximization. This experimental study uses a two-stage Cournot duopoly game to describe strategic delegation in a market (Vickers, 1985; Fershtman and Judd, 1987; Jansen et al., 2007). In the experiment, there are two firms in a market. Each firm consists of one owner and one manager. In Stage 1, both firm owners choose a contract determining the managers’ incentives: profit delegation (only profit), revenue delegation (profit and weighted revenue), and market share delegation (profit and weighted market share). In Stage 2, managers choose whether or not to collude. Owners’ and managers’ selections will determine market output under strategic delegation. Theoretically, the market outcome under strategic delegation is more aggressive than that under profit maximization. Thus, this study will investigate whether the theoretical results are held under strategic delegation. This research will help understand firms’ behaviors are more/less competitive than analyzed before. For example, Spierdijk and Zaouras (2017) show that the Lerner index obtained by profit maximization assumption is not appropriate to describe market power if firms do not maximize profit. As an extension of such research, this study will use experiments to identify firms’ behaviors that are not profit-maximizing. Also, this project will contribute to improving structural models. Since the BLP model (Berry, Levinsohn, and Pakes, 1995) was proposed, demand estimation has been advanced, but the ‘supply’ part from the model still assumes profit maximization that leads to underestimating the marginal cost. Thus, the results of the experimental study will potentially help improve structural models.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.