287 12 Price search and obfuscation: an overview of the theory and empirics Sara Fisher Ellison 1 1 INTRODUCTION The cost of price search is an important and salient feature of how many retail markets function. Take the market for new cars as a concrete example. Prices are typically not posted at dealerships. When they are, it is understood by most market participants that they are negotiable, so the posted price really is just a maximum price. Calling a dealer- ship for a price quote is often futile: customers are told to ‘come in, test drive the car, and we can talk price at that point’. There are many ways to upgrade and customize each car, making price comparisons across dealerships difficult. All-in prices may also include additional fees, taxes, destination charges, and so forth, which may neither be reflected in the quote nor be particularly transparent. Also, of course, a visit to a dealership to obtain a price quote can be a very time-consuming activity. It might involve a fair amount of sitting around, perhaps totaling a few hours. Dealerships are typically spread out geographically, so obtaining a number of reliable price quotes could easily be a two- or three-weekend proposition. Clearly, the high cost of price search in that market can have a large effect on consumer behavior (how many customers end up visiting one dealership and buying from them?) and can have a large effect on dealers’ equilibrium prices as well. Furthermore, it can certainly seem as though dealers go out of their way to make price search a very difficult and time-consuming process. 2 In 1961, George Stigler published his seminal article ‘The economics of information’, which started the formal conversation in economics about costly price search. Industrial organization, in an attempt to understand retail pricing behavior, has seen a recent ren- aissance of some of the ideas first mentioned in Stigler (1961). Stigler does not provide a game-theoretic model of price search, where both sellers and consumers behave ration- ally. He does, though, make a number of important observations which have helped frame the conversation on price search. He ties together price dispersion and imperfect information about prices by consumers, quite closely as it turns out: ‘Price dispersion is a manifestation – and, indeed it is the measure – of ignorance in the market’ (Stigler 1961, p. 214). That statement augurs an important branch of the search literature highlighting equilibrium price dispersion as a result of search costs. He points out that price search typically has a decreasing marginal value in the number of quotes already obtained. He observes that search may be more valuable in markets where consumers make repeated purchases (provided there is positive correlation in prices over time). He mentions situa- tions where one-time buyers would be expected to pay higher average prices than repeat purchasers. Much modern empirical work leverages these insights. Stigler also discusses market mechanisms that arise to mitigate market failures or inefficiencies caused by costly price search, such as specialized brokers and centralized market platforms.