Mutual Interdependence between Consumers and Firms

Abstract

In markets where multiple agents coexist, decisions across agents can be interdependent. A consumer's decisions about what product to purchase or how much to consume could be influenced by the decisions his or her peers make (e.g., family members, friends, and neighbors). Likewise, when firms make decisions about actions such as entry, pricing, product development, and location, they also consider how other firms' behaviors could affect their sales. This research explores how to quantify these interdependences across both consumers and firms based on behavioral data (also known as "revealed choice data"), while also accounting for confounding factors using econometrics methods.

The first essay explores the interdependence across consumers. In this study, we extend the previous literature on consumption in various settings by accounting for exogenous factors that could change the peer's behavior (the exogenous peer effect) and whether the peer is present at the time of consumption but does not consume (the peer presence effect) in addition to the typically modeled endogenous peer effect (how one's behavior is influenced directly by the peer's behavior). We develop a simultaneous equation model that allows us to identify all three peer effects simultaneously and apply it to behavioral data from a casino gambling setting.

In the second essay, we extend the context to measuring the interdependence among firms' decisions by focusing on the location choices of retailers. An agglomeration of retailers providing different goods can create positive spillovers by attracting multi-purpose shoppers. An interesting notion is that these multi-purpose shoppers may travel farther to visit a store with other retailers nearby than patronizing a standalone store. In the present study, we quantify the agglomeration effect as the increase in the catchment area for retailers. We develop a multinomial choice model and apply it to a consumer store choice data. We measure the increase in consumers' likelihood of visiting a particular grocery store during peak demand periods for non-grocery stores located in the vicinity of the grocer. Based on this measure, we then infer the increase in the catchment area that a retailer could enjoy by locating next to other types of stores.