Consumer search in the U.S. auto industry: The role of dealership visits

Abstract

In many markets, consumers visit stores and physically inspect products before making purchase decisions. We view the inspection of a product at a retail location as a search for product fit. We quantify the cost and benefit from searching for product fit using a discrete choice model of demand with optimal sequential search. In these models, the benefit of searching is measured by the standard deviation of the product fit and has, heretofore, been fixed to one in estimation. We show that, with an exogenous search cost shifter, both the cost and benefit of searching can be separately estimated. Our empirical setting is the U.S. automotive market. We assemble a unique data set containing individual-level smartphone geolocation data that inform us about dealership visits. We also obtain information on new vehicle purchases from proprietary DMV registration data. Our exogenous cost shifter is the distance a consumer must travel to visit a dealership. Our results show that the benefit provided by dealerships to consumers is substantial. Within our empirical context, failure to estimate the standard deviation of the product fit leads to biased search cost and consumer surplus estimates and to inaccurate predictions regarding consumers' number of searches and effects of at-home test drive programs.

Sequential Search Models: A Pairwise Maximum Rank Approach

“…They point out that if we can reliably estimate the endogenous unobserved product quality as a first step, then their estimator can recover the true consumer preference and search cost parameters. Yavorsky et al (2021) identifies the variance of match value with exogenous cost shifters if the match value follows a normal distribution, whereas this paper considers a larger class of distributions for the match value.…”

“…Other examples of z aj are loading time and information complexity of the web page because consumers incur higher cognitive costs to gather and analyze information when the information is less ordered (Gu and Wang, 2021). In the context of offline search, examples of z aj include consumers' distances to stores (Moraga-González et al, 2018;Yavorsky et al, 2021); how time-constrained consumers are when they search (McDevitt, 2014;Seiler and Pinna, 2017;Chen and Yao, 2017); other consumer characteristics that might affect search cost such as age, education, experience on searching: Hortaçsu and Syverson (2004) suggests that new investors in mutual funds who are younger, less educated and less experienced are likely to have higher information-gathering costs. The search cost parameter γ ∈ R qz is homogeneous across consumers in the baseline model.…”

“…Current sequential search models have to assume that researchers know the true distribution of the match value error term in utility to make the likelihood estimation possible. Yavorsky et al (2021) show that misspecifying the distribution of match value leads to biased estimates of search costs, thus incorrect predictions of search behavior and welfare implication. I will corroborate their argument with monte carlo results in Section 5.…”

“…In many situations where people make choices, they might not have the information about all existing options for free. Acquiring and processing information can be costly in a wide range of empirical contexts: for example, school applications (Hoxby and Turner, 2015), housing and neighborhood (Bergman et al, 2020), hotels (Koulayev, 2014;Gu and Wang, 2021;Chen and Yao, 2017;Ursu, 2018), financial products (Hortaçsu and Syverson, 2004), insurance products (Brown and Goolsbee, 2002;Honka, 2014), credit market (Agarwal et al, 2020), TV channels (Yao et al, 2017), medical devices (Grennan and Swanson, 2020), automobiles (Moraga-González et al, 2018;Yavorsky et al, 2021;Murry and Zhou, 2020), other durable goods (Kim et al, 2010(Kim et al, , 2017Dong et al, 2020), and so forth.…”

“…Moraga-González et al (2018) andElberg et al (2019) derive analytical solutions of the reservation utility for type I extreme value and logistic distributions of match value Yavorsky et al (2021). show that the variance of match value can not be separately identified from the search cost without an exogenous search cost shifter if the match value follows a normal distribution.…”