اهمیت موقعیت: شواهدی از آزمایش فیلدی خرید آنلاین Location Still Matters: Evidence from an Online Shopping Field Experiment

I. Introduction A visit to Amazon, eBay, or a host of price comparison sites will quickly confirm Hal Varian’s famous quip (1980) that “the law of one price is no law at all.” Sellers offer apparently identical items for wildly varying prices, with ranges often exceeding 100 percent. At first blush, this seems puzzling—why would some firms charge higher prices than others? The practical (though facile) answer is that firms charge more because (some) consumers will pay more. Yet this is hardly satisfying. Why do consumers pay more? How many consumers pay more? If consumers do pay more, then why don’t all firms charge the higher price? Before the rise of online markets, economists argued that such price dispersion was the product of search frictions. Some consumers were simply unaware of the lower prices and, at least in expectation, it was not worth their while to find out. Stigler (1961) was first to formalize this idea, and such approaches dominated economists’ thinking until the late 1990s. Arguably, the internet revolutionized consumer search. When a consumer could simply scan down a price list, or even sort the list with the click of a mouse, appeals to search frictions became less plausible. While a number of models, starting with Baye and Morgan (2001) offered rational explanations of price dispersion through sophisticated strategic behavior, simpler and more direct explanations ran along two main lines. The first line asserted that, although products appear identical, there are, in fact, myriad differences across sellers. An online consumer is not simply buying the product, but consuming the entire purchase experience, from ease of use of the seller’s website and design of the shop webpage, to after sales service features like warranties, return policies, and restocking fees. From this perspective, price dispersion merely reflects the differing costs and willingness to pay for various types of purchase experiences. Sellers charging high prices simply offer a better experience than those charging lower prices. Moreover, some consumers value these aspects and are willing to pay for them. A second line of argument reflected the growing incorporation of behavioral factors into economic models—in this case, limited attention. By this rationale, scarce consumer attention causes lower prices to sometimes go unnoticed. While it might seem unlikely that a consumer will overlook a low price in a setting where sorting by prices is simple, in practice, seemingly small costs, like clicking a mouse or scanning a page, prove large. In a structural model of search costs, Hong and Shum (2006) estimate that the (mainly psychological) cost of evaluating each price listing on an internet price comparison site exceeds $3. Market forces bear this out—competition is fierce to be the top listing following search queries on Google. The same is true at the price comparison site Nextag, in spite of the fact that the default search order is easily sortable (by price among other things) with a single mouse click. Likewise, featured listings on eBay constitute an important revenue stream despite a variety of easily used search and sorting tools that remove their privileged position. In effect, limited attention models represent a return to traditional search frictions explanations of price dispersion, albeit with psychological microfoundations now added. Moreover, as we show, even seemingly price sensitive consumers who rely on sorting tools do not always transact at the lowest price.