Inflation Expectations: Firm-Level Experiments

企业的通胀预期对于理解通胀的动态至关重要，特别是在像大流行后的经济体这样的情况下，在这些经济体中，大幅冲击是通胀螺旋式上升的触发因素。公司制定价格，在市场上竞争，并选择决定其边际成本的投入。他们的通胀预期进入并影响了所有这些决定。我们的目标是通过调查实验对企业的通胀预期进行因果推断。具体地说，我们希望测试三个可能影响企业通胀预期的渠道。首先是成本渠道。这包括劳动力成本，在大流行后的情况下，劳动力成本可能会上升，这也是由于劳动力供应减少，以及随之而来的工会权力的增加。相反，自动化可能会帮助企业降低劳动力成本，从而抑制通胀。第二个问题与竞争对手的角色有关，因为在竞争较弱的环境中，公司在提价方面有更大的灵活性。反过来，公司本身也是基于他们对竞争对手定价策略的预期来形成通胀预期的。第三个问题是消费者需求的作用：企业制定价格时，将需求弹性视为已知。大多数情况下，由于担心消费者的反对，他们会避免提价。调查实验将由四个模块组成。首先，我们将提出问题，看看四种渠道中哪一种更有利于企业预期的形成。在其他三个模块中，我们将重点介绍每个渠道。每个模块都有相似的结构：征求公司关于相关变量的先验信息，随机向他们提供信息(以工会要求、竞争对手的期望或消费者的期望的形式)，然后征求随机对照试验公司的后验期望。

Firms’ inflation expectations are crucial for understanding the dynamics of inflation, particularly in situation like the post-pandemic economies in which large shocks act as trigger of inflation spirals. Firms set prices, compete in markets and choose inputs that determine their marginal costs. Their inflation expectations enter and affect all of those decisions. We aim at a causal inference of firms’ inflation expectations through a survey experiment. Specifically, we wish to test three possible channels that may affect firms’ inflation expectations. The first is the cost channel. This includes the cost of labour, which is likely to raise in the post-pandemic scenario also due to reduced labour supply, and the ensuing raise in unions' power. On reverse instead automation may help firms reduce the cost of labour, hence tame inflation. The second is linked to the role of competitors, as firms have more flexibility in raising prices in less competitive environments. In turn firms themselves form their inflation expectations based on what their expectations of competitors' pricing strategies. The third is the role of consumers’ demand: firms setting prices, do so by taking demand elasticity as given. Most often they refrain from raising prices when they fear consumers’ antagonism.The survey experiment will be composed of four modules. In the first we will ask questions to see which of the four channels is more prevalent for firms’ expectation formation. In the other three modules we focus specifically on each of the channels. Each module has similar structures: firms’ priors about relevant variables are solicited, information is provided to them randomly (in the form of unions’ requests, competitors’ expectations or consumers’ expectations), following the randomized controlled trial firms’ posterior expectations are solicited.