Behavioural Extensions in Estimation of Demand and Market Power

我的研究目的是在不完全竞争的市场中估计需求，而不强加传统的优化假设。了解消费者需求本身就很有趣，因为这样做可以深入了解随着价格和产品特性的变化，消费者如何在商品之间进行替代，以及广告和信息披露如何影响消费者行为。然而，我的研究的目的是估计需求作为一种手段来回答一个基本的问题，在实证产业组织：有多少市场力量的公司？从这个意义上说，我的研究与“新经验产业组织”（NEIO）是一致的。由Bresnahan（1989）提出的NEIO通过采用间接方法来解决不可观察的企业成本的数据问题，在该方法中，企业需求函数（更具体地说，需求的价格弹性）的估计被用来估计企业的加价。所述方法的一个显著应用是在竞争政策方面。事实上，联合王国竞争和市场管理局进行了经济计量分析，以评估合并后价格上涨和串通行为的可能性。这些练习借鉴了学术研究，并推动了经验技术的发展，其中包括行为偏离理性。我估计需求在一个离散的选择设置。离散选择模型预测消费者在两个或多个替代品之间的选择。这一方法的范围很广，涵盖了重要的经济决定，如是否加入劳动力队伍或选择哪种运输方式。出于本主题的目的，我集中讨论消费者在给定行业中有限数量的竞争产品之间进行选择的情况。而不是强加不切实际的最优化假设，消费者最大化预期效用，我的目标是建立一个计量经济学框架，结合行为经济学理论的发展，特别侧重于非预期效用理论（一个突出的例子是前景理论）和有限理性的模型。

My research aims to estimate demand in imperfectly competitive markets without imposing traditional optimising assumptions. Understanding consumer demand is interesting in its own right because doing so provides insights into how consumers substitute between goods as prices and product characteristics change and how advertising and information disclosure affect consumer behaviour. However, the purpose of my research is to estimate demand as a means to answer a fundamental question in empirical industrial organisation: how much market power do firms have?In this sense, my research is line with the "New Empirical Industrial Organisation" (NEIO). Coined by Bresnahan (1989), the NEIO tackles the data problem of unobservable firm costs by employing an indirect approach in which estimates of firms' demand functions (more specifically price elasticity of demand) are used to estimate firms' mark-ups. A notable application of the method described is in competition policy. Indeed, the UK Competition and Markets Authority conducts econometric analysis to assess the likelihood of price increases and collusive behaviour following a merger. These exercises draw on academic research and motivate the development of empirical techniques that incorporate behavioural departures from rationality. I estimate demand in a discrete choice setting. Discrete choice models predict consumer choices between two or more alternatives. The reach of this approach is extensive, covering important economic decisions such as whether to enter the labour force or which mode of transport to select. For the purposes of this topic, I concentrate on the case of a consumer choosing between a finite number of competing products in a given industry. Instead of imposing the unrealistic optimisation assumption that the consumer maximises expected utility, I aim to build an econometric framework that incorporates developments in behavioural economic theory, focusing specifically on models of non-expected utility theory (a prominent example is prospect theory) and of bounded rationality.