Price Discrimination Is Good, Part I

Cite this Article
Joshua D. Wright, Price Discrimination Is Good, Part I, Truth on the Market (November 30, 2008),

Price discrimination involves a firm taking advantage of different elasticities of demand for the same goods by charging different prices relative to marginal cost. Price discrimination is ubiquitous in our economy but remains a four letter word in policy and regulation circles. We observe price discrimination in all sorts of product markets, from small and large firms, and in marketing strategies from brick and mortar to Web 2.0. Its economic definition is relatively straightforward and it is a concept, unlike the complex models and explanations for some business practices in the modern economics literature, that is intuitive for everyday consumers. Airlines charge reduced fares for children or require Saturday stayovers in order to exclude business purchasers. We see this type of price discrimination every day in grocery stores, gas stations, movie theaters, online retail websites, and bookstores. It also exists in markets with which every day consumers might be less familiar, e.g. the tying of ink and printers.

The important economic point is that price differentials can be expected to persist in equilibrium in competitive markets characterized by differentiated products. And what are the welfare effects of price discrimination? Economists have spilled a great deal of ink on this subject. I summarize the literature in “Missed Opportunities in Independent Ink,” but here’s the bottom line:

The conventional mantra for those advocating strict antitrust scrutiny of price discrimination is the well-known economic analysis which concludes that the total welfare effects of third-degree price discrimination are ambiguous. From a static perspective, this analysis makes some intuitive sense: some buyers receive lower prices (and purchase higher quantities) while other buyers receive higher prices (and purchase lower quantities) and therefore the net impact of price discrimination on output is ambiguous. Indeed, standard models of price discrimination in the economic literature demonstrate that an increase in aggregate output is a necessary condition for price discrimination to increase welfare under monopoly. This conventional welfare analysis is of limited utility for several reasons. The first is because most aftermarket metering agreements do not involve what Pigou classified as third-degree price discrimination. Recall that third-degree price discrimination refers to the practice of breaking buyers into distinct groups and setting a profit maximizing price for each group, resulting in one price for each group. But aftermarket metering arrangements, like those involved in Independent Ink, do not involve third-degree price discrimination. Rather, Illinois Tool Works’ arrangement involved a single price for printers and ink which allowed it to vary the total package price across each unit sold to each buyer according to the intensity of use in an attempt to capture the maximum surplus. Benjaim Klein and John Wiley point out that aftermarket metering arrangements, the most common example of price discrimination in antitrust analysis, are much more appropriately described as second-degree price discrimination.

Klein and Wiley (70 Antitrust LJ 599 (2003)) show that most price discrimination schemes involving metering, the most common observed class of such arrangements, are much better described as imperfect attempts at first-degree discrimination than third-degree discrimination. Why? Because aftermarket metering, like perfect price discrimination, involves an attempt to collect the maximum for each unit sold to each buyer by metering the intensity of the package demand. High intensity users pay a higher package price and lower intensity users pay a lower package price, with the seller collecting varying levels of consumer surplus from each type of user. Of course, we know that the static welfare effects of perfect price discrimination are unambiguous: producer surplus increases and consumer surplus falls.

But this brings us to a second critical failure of the standard welfare analysis: ignoring dynamic efficiencies. Of course, this shortcoming is common to much of antitrust analysis and is not unique to price discrimination. A few basic points are in order. First, the increase in producer surplus predicted by the conventional price discrimination analysis provides incentives for additional investment in innovation and other competitive investments such as increasing product variety, expanding retail outlets, or research and development. Second, this means that investments to enhance the ability to price discriminate are not socially wasteful, rent-dissipating investents. They are best viewed as part of the competitive process as firms attempt to expand output and attract consumers through the offering of valuable products and services.

A related and third shortcoming of the standard static welfare economics of price discrimination is that it ignores the possibility that price discrimination intensifies competition and therefore increases consumer welfare for all consumers. Recent models analyze the competition-intensifying impact of price discrimination and challenge the result that price discrimination necessarily involves losses to some consumers. The economic intuition behind these models is that price discrimination is a competitive tool that allows firms to compete for all consumers on different segments of the demand curve by offering a menu of prices rather than competing only for the marginal consumer with uniform pricing. Competitive price discrimination therefore leads to lower profits and lower prices.

In sum, while economics provides no single universal welfare theorem for all arrangements involving price discrimination, the lessons above tell us that price discrimination is nothing to fear from a competition and efficiency perspective. When one accounts for both static and dynamic welfare effects, competitive discrimination is likely to result in lower prices, higher output, and increased innovation. You didn’t think those grocery store coupons, ladies night, and student discounts were bad for consumers did you?

Yet, price discrimination gets a bad rap in policy circles. In antitrust, for example, price discrimination is still incorrectly (in my view) used from time to time to infer antitrust-relevant market power. Suspicion of price discrimination looms large in the network neutrality debate. And oh yeah, its also illegal — at least sometimes. Its true that the Antitrust Modernization Committee mustered up 9 votes in favor of repeal of the Robinson-Patman Act, but I don’t know anybody that is betting on that happening anytime soon. Of course, Baumol and others have turned much of the regulatory animosity toward price discrimination on its head by showing that price discrimination is not only consistent with competition, but is caused by it, i.e. competition forces firms to adopt discriminatory prices under competitive conditions without the presence or promise of monopoly profits. And we know that given the ubiquity of price discrimination in competitive markets, even ignoring the dynamic efficiency arguments above and focusing exclusively on static effects, it would be nearly impossible to accurately identify those discriminatory price regimes that result in net consumer welfare losses even if we wanted to! And in the rare case where a discriminatory regime So, why the suspicion for practices that involve price discrimination?

Perhaps the animosity toward price discrimination is right there in the title. Its a form of discrimination. And that’s bad, right? Or at least unfair in some important way? But recall that one important aspect of the economics of price discrimination is that it is almost always the case that it involves reduced prices for the price-sensitive group and higher prices for the price-insensitive group. To the extent that it is true that the lower income groups are the price-sensitive group, price discrimination generally benefits the lower income group at the expense of the higher income group. So lets not pretend that price discrimination opponents havea monopoly over fairness arguments.

Let me conclude with few thoughts:

First, I intend to start a new and occasional series with the title of this blog post, in the spirit of Mankiw’s cross-elasticity of gasoline and Marginal Revolution’s markets in everything posts, where I cite examples of common and not-so-common price discrimination schemes and puzzles. So, readers, send me your ideas!

Second, one lesson of the price discrimination literature that is ignored is exactly how unusual uniform and linear pricing is in the modern economy! The ability and incentive to price discriminate, which requires only the absence of perfect substitutes, is ubiquitous. It is a power held by every restaurant, landlord, corner gas station, supermarket, and small firm in the economy. Because price discrimination is profitable, we can expect firms in our modern economy to invest substantial resources not only in product differentiation, but also in finding methods to price discriminate. These provide no reason for competitive concern. Indeed, I think one could make a plausible argument (though I won’t in this post) that the absence of price discrimination is more worthy of regulators’ attention than its presence.