The leverage theory of tying revisited: evidence from newspaper advertising.

• Author: Slade, Margaret E.

1. Introduction

   Requirements tying occurs when a firm that operates in one market, the tying market, stipulates that users of a good that it produces, the tying good, must also purchase their requirements of a second good, the tied good, from the seller of the first.(1) Tying is common in a broad spectrum of industries that range from competitive to monopoly. For example, many franchisors, who operate in relatively competitive retail markets, require that their franchisees purchase not only a business format from the parent company but also some of their inputs from the chain. Tying also occurs at the other end of the market-structure spectrum. For example, it was common for telephone companies to require that purchasers of telephone services also obtain their telephone equipment from the operating company.

   In certain markets, however, casual observation suggests a positive relationship between tying and monopoly power. For example, consider the advertising industry, where the tying product is space in printed media (e.g., magazines) or time in visual media (e.g., television) and the tied product is the creative and administrative services of advertising agents. As a general rule, advertising media that are local in scope have more monopoly power than those that reach national audiences. Moreover, local media are prone to tying, whereas national media are not.

   To illustrate, in Canada, affiliates of telephone companies have a virtual monopoly on Yellow Pages directory-advertising space and all of the directory-publishing companies tie the provision of advertising services to the purchase of space in the directory.(2) The newspaper-advertising industry, in contrast, has an intermediate market structure. Indeed, some cities have several competing daily newspapers, whereas other cities have only one, and some dally newspapers tie, whereas others do not. Finally, with national-advertising media such as magazines and television, all firms have direct competitors, and tying is uncommon.

   There are a number of reasons why a firm with market power might want to engage in tied sales, and many of these go by the name of efficiency motives. These include price discrimination, cost savings, and quality control. The leverage theory, in contrast, is a market-power motive.

   The leverage theory claims that a monopolist can use his market power in the tying market to extract additional surplus from the tied market even when the latter is competitive. Unlike efficiency explanations, the possibility of leveraging from one market to another has been hotly contested by lawyers and economists. For example, Bork (1978, p. 372) rejects the notion and states that this "theory of tying arrangements is merely another example of the discredited transfer of power theory, and perhaps no other variety of that theory has been so thoroughly and repeatedly demolished in the legal and economic literature."(3) On the other side of the controversy, Grimes (1994, p. 268) endorses the notion and notes that "If the tying seller can foreclose a substantial percentage of the tied-product market, competing sellers of the tied product may face increased costs and be driven from the market."

   Given this controversy, it is somewhat surprising that there has been little empirical analysis of tied sales. The leverage theory, however, is consistent with the few empirical regularities that have been uncovered. For example, Hass-Wilson (1987) found that contact-lens prices were significantly higher in states where their sale was tied to the services of ophthalmologists and optometrists. In addition, Slade (1998) compared the prices of U.K. beers sold in tied and free public houses and found that they were higher in the former.(4) Finally, Grimm, Winston, and Evans (1992) found that, contrary to the leverage theory, monopolization of one portion of a railroad route was insufficient to extract all monopoly surplus.(5) These findings suggest that leveraging can be profitable.

   Due to its intermediate market structure, the newspaper-advertising industry provides an ideal laboratory in which to test the leverage theory. For this reason, after assessing the private profitability of tying by a firm with market power under circumstances where the standard efficiency defenses are not apt to hold, data obtained from Canadian newspaper publishers are examined. Nonparametric tests of association demonstrate that, with newspapers, tying and monopoly power go hand in hand. In particular, whereas over 85% of the newspapers that are published in cities where there is no direct competitor tie the provision of advertising services to the purchase of advertising space, only half of the newspapers that are published in multi-newspaper cities impose a similar restriction.

   Prior to the empirical analysis, I develop a theoretical model that is tailored to fit the circumstances that characterize the advertising industry. In particular, since the tied good is a service whose suppliers are paid a commission rather than a fee for service, the model reflects this empirical fact. I show that, under weak regularity conditions on demand, tying is almost always profitable.

   The model is extremely simple, and to many it will seem obvious. Nevertheless, given the contradictory statements that appear in the literature, it seems worthwhile emphasizing that leveraging can be profitable under a wide range of circumstances. The assumptions that underlie the model are typical of the 'Chicago School' vertical-restraints literature. In particular, my argument does not rely on private information (as in Mathewson and Winter [1997]) or economies of scale and/or strategic behavior (as in Whinston [1990] and DeGraba [1996]).(6) Even in this stylized situation, however, leveraging pays.

   The organization of the paper is as follows. The next section, which uses the case of newspapers to illustrate the phenomenon of monopoly tying, discusses the Canadian newspaper-advertising market and argues that efficiency considerations are unlikely to explain tying in this industry. This leaves leverage as the primary unexplored motive. A formal model of leveraging is developed in section 3 that illustrates the circumstances under which it is profitable. Section 4 describes my survey of newspaper publishers and analyzes the data obtained through this survey. Finally, section 5 concludes. I do not examine the welfare consequences of tying, which are usually ambiguous.

2. The Newspaper-Advertising Market

   The Market

   I use the case of newspaper advertising to illustrate the issues involved in monopoly tying. Advertising is an important industry in Canada, with gross revenue of over $10 billion per year, which corresponds to approximately 1.5% of gross domestic product. Newspaper advertising is the largest category within this class and accounts for more than one quarter of the total.(7)

   Most medium-sized cities and even some large cities have only one daily newspaper. While it is true that other media compete for firms' advertising dollars, a lone newspaper nevertheless possesses considerable market power. Indeed, some types of business find that newspapers are essential for promoting their products. For example, many cinemas regularly list their films in the newspaper and grocery stores rely heavily on color supplements to promote their weekly specials. These businesses might have difficulty finding effective substitutes. Furthermore, the very nature of newspaper advertising, much of which is local, implies that newspapers in distant cities are not close substitutes.

   Newspaper advertising can be divided into two products advertising space and advertising services. Advertising space consists of compiling and processing advertising-sales information, verifying layout, typesetting, photo composing, producing hard copy, printing, and delivering the paper. Advertising services consist of establishing new customers and providing all customers with advice concerning cost, content, design, location, and placement of advertisements. In addition, independent advertising agencies often bear the cost and responsibility of delivering the finished product to the publisher and billing and collecting from the client. Finally, some customers are provided with market research that is tailored to their special needs, and some agencies provide advice on how to integrate newspaper advertising with other media.(8)

   The publishers of newspapers control the supply of advertising space in their papers. Advertising services, in contrast, are competitively supplied by advertising agencies. Newspaper publishers can choose to supply advertising services themselves or to deal with unaffiliated advertising agencies. In this market, therefore, the tying product is advertising space, whereas the tied product is advertising services.

   The range of services that are supplied by newspapers is virtually identical to that supplied by independent agencies. In particular, a newspaper advises its clients, designs their ads, and provides a full range of graphic services. The exception is that newspapers do not deal with other media and therefore do not attempt to integrate newspaper advertising with advertising in those media.

   When newspapers tie, advertisers who purchase space in the paper are required to pay for a bundle that includes both space and services, even when they do not use the services of the newspaper. If advertisers wish to employ an independent advertising agency, they must therefore pay for services twice: once to the newspaper as part of the package and once to the agency for services alone.(9)

   Many publishers pay commissions to agencies for the services that they provide, and this commission is almost always 15% of a customer's expenditure on space. The uniformity of the commission rate is somewhat surprising, particularly since the price of space varies widely across publishers. The norm of 15% seems to have been established by the...