Helping buyers beware: the need for supervision of big retail.

• Author: Van Loo, Rory
• Position: II. Firms Systematically Charge Anticompetitive Prices B. Potential Market Constraints on Supracompetitive Pricing 3. Protection by Sophisticated Shoppers through Conclusion, with footnotes, p. 1351-1392

3. Protection by Sophisticated Shoppers

   Another possibility, articulated by Alan Schwartz and Louis Wilde in the legal literature, is that a critical mass of sophisticated consumers could shield unsophisticated consumers from harmful decisions. (181) For example, if enough sophisticated shoppers switched stores when prices on low-salient items were raised, the gains from higher prices paid by unsophisticated customers might not justify the losses from sophisticated customers switching stores. (182) This would keep stores from raising prices on low-salient items, thus benefitting even unsophisticated consumers. Meanwhile, if stores offer products for both kinds of consumers, unsophisticated shoppers might accidentally choose the sophisticated product simply because it is on the shelf, limiting the number of potential sales to the sophisticated shoppers who will not purchase unsophisticated goods. (183)

   While historically these dynamics may have protected consumers, they are likely becoming less important for three reasons. First, sophistication is relative, and as discussed above, seller sophistication and market complexity have grown so much over the past several decades that the bar for being a sophisticated shopper is quite high and growing. (184) Many individual products have become more complex. At the same time, sophisticated shopping for more straightforward items would require digital access to price and product information in a variety of stores or consumers would need to spend large amounts of time putting that information together. Once the information is acquired, it would be no simple task, even with a spreadsheet, to calculate the total costs of different configurations of store visits--how many stores to visit and which items to buy at which stores. Transportation costs and time estimates would also need to be considered. As such, the bar for sophistication is high, and there is no evidence that such sophistication has materialized within consumer goods markets in a sufficiently substantial mass.

   Second, refined segmentation of product categories makes it less likely that consumers will accidentally make the best choice. Markets of only a few segments in the early 1980s, each with many consumers, have given way to highly segmented modern markets--in some instances moving toward segments of a single consumer. (185) In a market with ten product segments, the likelihood of an unsophisticated consumer stumbling on the sophisticated choice is innately less than in a market with three segments. (186)

   Finally, even if markets had fewer segments, the nature of behavioral pricing practices prevents unsophisticated shoppers from serendipitously selecting the best deals. Unsophisticated shoppers will not accidentally purchase the higher-priced printer that corresponds to the lower costs of ink. They will instead choose the cheap printer with (low-salient) expensive ink and pay more overall. Also, retailers leverage their scientific studies of consumers' eye movement sequences to place the sophisticated choices where consumers would be least likely to look--whether on a computer screen or on a store shelf. They can further direct shoppers toward anticompetitively priced items through in-store advertising.

   It is important to note, however, that a critical mass of sophisticated shoppers providing protection for less sophisticated shoppers has the potential to play a more meaningful role than it currently does in moving goods markets toward competitive pricing. This could occur if retailers were required to provide information intermediaries with digital price and product information, thus arming a critical mass of shoppers. This policy recommendation is discussed further in Part V.

   Ultimately, the question of whether market forces such as reputational constraints, consumer learning, and sophisticated shoppers correct anticompetitive practices is an empirical one. The persistence of "lemon" used car markets and the widespread predatory lending practices prior to the 2008 financial crisis support the notion that markets are often unable to self-correct for even highly visible, questionable practices. (187) The most convincing evidence that these market mechanisms do not keep anticompetitive practices in check within consumer goods markets is the systemic presence of those anticompetitive practices and the empirical evidence that they cause consumers to misperceive prices. It is difficult to reconcile such empirics with a rationality-based theory of consumer behavior in the retail goods sector.

   3. AGGREGATE HARMS ARE POTENTIALLY SIGNIFICANT

      Sellers' anticompetitive practices have the potential to cost households hundreds of dollars or more annually. The alternative choice--spending additional time shopping to outsmart pricing practices--is impractical and itself harmful. The additional costs likely reach all consumers but are particularly troubling for low-income families, for whom even a few hundred dollars can have meaningful health and educational implications. (188)

      1. The Costs of Market Failure May Be Significant

         Though more comprehensive studies are needed, a number of empirical studies indicate how much extra consumers may pay in specific contexts due to anticompetitive pricing. From these figures, it is possible to gain a sense of the total monetary cost to households annually. Related costs, such as any resulting overconsumption, are not readily quantifiable.

         In the most empirically robust study available--the only one to have access to a firm's internal cost information--economists Glenn Ellison and Sara Fisher Ellison found that firms selling computer accessories to consumers online earned margins six to nine percentage points higher due to pricing obfuscation. (189) These findings were all the more striking because the markets otherwise demonstrated conditions that would be associated with fully competitive markets--ease of entry, minimal product differentiation, and low search costs. (190)

         Other field studies are less comprehensive because they measure only one anticompetitive practice rather than the total amount possible from an array of such practices. For example, purchasers of clothing items in a mail-order catalog paid 23% more when the price ended in "9" compared to prices ending in other numbers. (191) Of course, retailers should be able to end their prices with whatever numbers they choose. And clothing may be unrepresentative of many goods that consumers purchase more regularly. But this study provides another data point on the potential magnitude of behavioral pricing, especially given that it measured only the effect of the last number in the price. And it does so in a context in which consumers were typically purchasing a very small number of items; thus, presumably the prices of those few items were salient.

         Indeed, given the underlying psychology, low-price items may be just as anticompetitively priced as more substantial purchases such as clothing. Low-priced items may be purchased more frequently, but they also have lower salience. An anticompetitive 15% price increase on a $2 item would amounts to $0.30 extra, which provides less incentive to spend extra time or pay extra attention than the $39 item of clothing. (192) In the case of the imperceptible downsizing of packaging by Skippy, Dial, and others, although the equilibrium anticompetitive price increase is unknown, the changes typically amounted to a 10% price increase for an array of products. (193)

         Laboratory studies have also estimated consumer willingness to pay higher prices. Framing effects were found to make consumers pay on average 11% more for beer, (194) over 100% more for electronics items such as cameras and DVDs, (195) and 20% more for clothing. (196) Field studies typically find a smaller magnitude than laboratory studies. However, by focusing on one behavioral lever for increasing prices, many studies likely underestimate the magnitude of possible higher prices in their respective settings.

         Reaching an aggregate annual figure for expenditures requires extrapolating from these existing studies to markets that exhibit anticompetitive practices but have not been directly studied. Assuming a 5% price increase--one percentage point below the low end of Ellison and Ellison's estimates and one-half to one-third of the magnitude found in other studies--would provide a conservative perspective based on the empirical data. A 10% figure is a reasonable high end of the range, as it is closer to the other studies finding double-digit price increases and is only one percentage point above the high end of Ellison and Ellison, which studied a setting favorable to comparison shopping. (197)

         Assuming expenditures on anticompetitively priced goods are made in proportion to general household expenditures, (198) a family earning between $20,000 and $29,999--which covers the poverty line for a family of four--would pay an extra $370 annually at the 5% rate due to market failures. (199) At the 10% rate, families in this range would spend an additional $740 as a result of anticompetitive practices. For a family earning $50,000 per year--about the median U.S. salary--this overcharge would amount to $600 to $1200 annually. (200)

         In different families, these several hundred dollars annually would be spent differently. However, economic development research suggests that many low-income families would spend additional income on health, better food, and education. (201) Studies also suggest that additional money available to low-income families can increase the quality of parenting, which has lifelong implications, especially for children's psychological health. (202) Because health and education are two important predictors of upward mobility and contribute to well-being, especially among low-income families, lowering the impact of supracompetitive pricing could contribute to a reduction of poverty and an improvement in upward...