|
Article Excerpt
Improved consumer information about horizontal aspects of products of similar quality leads to better consumer matching but also to higher prices, so consumer surplus can go up or down, while profits rise. With enough quality asymmetry, though, the higher-quality (and hence larger) firm's price falls with more information, so both effects benefit consumers. This occurs when comparative advertising is used against a large firm by a small one. Comparative advertising, as it imparts more information, therefore helps consumers. Although it also improves the profitability of the small firm, overall welfare goes down because of the large loss to the attacked firm.
1. Introduction
* Until the late 1990s, mentioning a competitor's brand in an ad was illegal in many EU countries. This situation was ended by a 1997 EU directive that made "comparative advertising" legal subject to the restriction that it should not be misleading. This brought the European approach closer to that of the Federal Trade Commission in the United States. In other countries, comparative advertising remains illegal, or little used (see Donthu, 1998, for a cross-country comparison). The rationale for a favorable attitude toward comparative advertising on the part of competition authorities is that it improves consumers' information about available products and prices (see Barigozzi and Peitz, 2006, for details and a wealth of examples and discussion). This raises a number of questions for the economic analysis of informative advertising. What is the scope of a practice that involves disclosing information that the product's supplier would choose not to reveal? Is the benefit to consumers from improved information mitigated by a welfare loss for competitors who are (presumably) hurt by comparative advertising about their products? Are consumers hurt by higher prices because product differentiation rises due to comparative advertising about product attributes?
The FTC's position is admirably clear: "Comparative advertising, when truthful and nondeceptive, is a source of important information to consumers and assists them in making rational purchase decisions.'' (1) This view underlies our modelling approach. The FTC also expects performance benefits: "Comparative advertising encourages product improvement and innovation, and can lead to lower prices in the marketplace." To a very large extent we corroborate these conclusions.
Here we consider a game between rival firms and their incentives to provide information. Consumers do not know the characteristics of a firm's product unless they are revealed through advertising, although consumers have (correct) priors about their evaluations. Firms are fully aware of each other's product attributes. If comparative advertising is illegal, then firms can only inform consumers about their own goods. Comparative advertising allows a firm to also inform consumers about rival product attributes that the other firm might not wish to communicate. We also address the welfare economics of comparative advertising.
Evaluating the impact of comparative advertising requires identifying when it changes the information available to consumers. That is, there must be some information that firms would not disclose if restricted to direct advertising, and that will be brought out if comparative advertising is allowed. In much of the literature on informative advertising, it is the cost of advertising that limits the information transmission by firms (see the seminal articles of Butters, 1977 and Grossman and Shapiro, 1984, and the review coverage in Bagwell, 2007). Anderson and Renault (2006) show that a monopoly firm might limit information about its product attributes even if advertising has no cost. This result is a starting point for the present article because it identifies situations where a firm is hurt by information disclosure about its own product, so that there might be some incentives for competitors to provide that information through comparative ads.
The article contributes to the economics of asymmetric oligopolistic competition by first indicating how quality-cost advantages feed into equilibrium prices and sales. Second, it provides results on the impact of increased product information on market outcomes: whereas more information tends to increase profits, and welfare when qualities are similar, welfare can be harmed with more information due to price distortions when qualities are quite different. Third, it provides some predictions on when comparative advertising might be used--by smaller firms with cost or quality disadvantages--and welfare implications.
There is curiously little economics literature on comparative advertising, although in marketing there is quite a lot of documentation of the phenomenon and discussion of its effectiveness (see Grewal et al., 1997, for a comprehensive survey). (2) Barigozzi and Peitz (2006) give a survey and some background modelling of alternative approaches.
Barigozzi, Garella, and Peitz (2006) take a signalling approach. An entrant with uncertain quality confronts an incumbent whose quality is known. The entrant chooses between "generic" advertising, which is standard money burning to signal quality (as in Nelson, 1974), and "comparative" advertising, which involves a claim comparing the two firms' qualities. Firms may have favorable or unfavorable information about the entrant's quality but do not observe it perfectly. If comparative advertising is used, the incumbent may litigate in the hope of obtaining damages if the court, which observes quality perfectly, finds that it is low. Comparative advertising may credibly signal favorable information about entrant quality when a firm with such information expects it unlikely a court will find its quality is low.
Aluf and Shy (2001) model comparative advertising as shifting the transport cost to the rival's product in a Hotelling-type model of product differentiation. Although this is an interesting angle in its own right, the modelling approach does not capture the informative aspect of comparison advertising and is not micro-founded in information revelation. Instead, it seems more like a model of (negative) "persuasive" advertising. In a Hotelling-Downs model of political competition, Harrington and Hess (1996) model negative political advertisements as moving the rival's perceived location away from the center, and positive advertisements as moving one's own perceived location toward the center (and hence increasing the chances of being elected). They show that a candidate with a quality (valence) disadvantage will use more negative advertising, in line with observed facts. Insofar as comparative advertising in our model is information about a rival that the rival would rather not see revealed, this result has an interesting parallel to our findings that a quality disadvantage is necessary for comparative advertising.
We consider the disclosure of horizontally differentiated attributes (valued differently by different consumers), assuming that product qualities are known. Here, product qualities are a device to indicate large or small firms in terms of their equilibrium market shares, and we shall refer to the firms as strong (to be thought of as one with a quality advantage) or weak (quality-disadvantaged). If market sizes are very different (product qualities are sufficiently different), the equilibrium to the disclosure game has only the weak firm disclosing horizontal attributes and the strong one not. If comparative advertising is allowed, then the weak firm will disclose the horizontal attributes of both products (and so it is truly comparative).
To see how the model works, it is first useful to describe some background results which nonetheless hold independent interest for the economics of product differentiation and information. First, under firm symmetry, or close to it, more match information makes for more product differentiation and therefore raises prices (as expected). As we show, consumers may be better or worse off, as their improved ability to select the better match may be swamped by the price hike. Nonetheless, firms are better off. Results are surprisingly different when products are asymmetric, a case rarely treated by the literature. With zero horizontal match information, Bertrand competition leads the weak firm to price at cost whereas the strong one takes all the market by pricing at its quality advantage. Assume for the sequel this quality advantage is large. If the match information for just one firm is known (the weak say), the strong firm has to actually price lower to retain the whole market because it must attract the consumer who likes the weak firm most. With full information, the strong firm must set an even lower price if it is to retain (almost) all customers, because it must now attract the customer who likes it least and likes its rival most. (3) This means that not only are prices lower when there is more information (more product differentiation) but also the strong firm's profits are lower the more information there is. However, the weak firm retains a foothold under full information, so it prefers this.
This leads us into the advertising game analysis. Throughout the text we emphasize the importance of firms' market shares by stressing two extreme cases, which allows for clean analytical arguments. A specific example helps fill in intermediate cases.
Similarly sized firms share the same incentive to provide extensive horizontal match information to maximize perceived differentiation and relax competition. Then, comparative advertising has no specific role insofar as full product information is provided regardless.
Dissimilar firms have quite different incentives to disclose horizontal match information. Hence, allowing comparative advertising may have a significant impact. This is best illustrated when the match distribution is such that the lower market share falls to zero when information is only one-sided (4): we show that shares are always positive with full information. As noted above, the strong firm prefers no information to one-sided information, which in turn it prefers to full information, and it serves the whole market unless its competitor can achieve full information through comparative advertising. Hence, neither firm advertises (provides information) when comparative advertising is banned (because the weaker firm can get no market sales by using direct advertising for its own product alone). When comparative advertising is allowed, the weaker firm will provide full product information to give itself some market share (and hence profits).
If firms are sufficiently different, it is socially optimal that all consumers buy from the stronger firm. Then, comparative advertising deteriorates social welfare by letting the weaker firm make positive sales: full information relaxes the strong firm's price incentive to capture the whole market. Consumer welfare is improved, though: full information brings down the price of the strong product (which is otherwise consumed by all) and some consumers choose to buy the cheap weak product which must therefore yield them higher surplus.
The text shows that these insights may remain valid even when the weak firm always retains some strictly positive market share with one-sided information. Under fairly general conditions, a sufficiently strong firm always prefers less information to more. We also show that there is a range of size differences for which the weak firm uses comparative advertising to achieve full product information, which the strong firm prefers obscured.
These results indicate that the benefits of comparative advertising accrue to the weak firm, and to consumers, with so much damage to the large firm that total surplus goes down. Although we noted above that more information can raise prices (and may even hurt consumers to the profit of firms, despite better consumer matching), this happens when comparative advertising is irrelevant in the sense that incentives to divulge own product information are already strong enough. When comparative advertising is relevant, the strong firm is attacked by the weak one, and we are in the regime when more information actually reduces prices (and enables better choices). This substantiates the FTC position that comparative advertising improves competition, even though one might have worried a priori that more match information would entail higher prices.
We provide an example with a Laplace distribution for match values with a complete characterization of the equilibrium outcome and welfare properties as a function of firm strengths. Results are fully consistent with those where the weak firm's market share vanishes to zero under one-sided information. This example shows that the weak firm's profit can be substantial enough to cover advertising expenses for comparative advertising.
Section 2 gives general background results for Bertrand duopoly with product differentiation. We outline the model in Section 3, describe demand under different degrees of product information, and find the corresponding equilibrium prices. These prices and profits are compared in Sections 4 and 5, which paves the way for the equilibrium information disclosure determined in Section 6. Section 7 shows some key surplus properties on the desirability of comparative advertising. Section 8 covers the Laplace example, and Section 9 discusses quality revelation, other extensions, and interpretation of the model. Section 10 concludes. The longer proofs are collected in the Appendix.
2. Some preliminary results
* We first give some results for duopoly pricing that are used quite extensively in the analysis that follows: demands will satisfy the properties used here. The results pertain quite generally to differentiated product Bertrand duopoly with covered markets.
Consider a duopoly where firms and 1 set prices [p.sub.0] and [p.sub.1]. Define [DELTA] as firm 1's net quality advantage, [DELTA] = [p.sub.0] - [p.sub.1] + Q, where Q [member of] R may be understood as firm 1's gross quality advantage. Demand for firm i's product is given by [D.sub.i]([DELTA]), i = 0, 1, where [D.sub.1] is an increasing function taking values in [0,1] defined on R. Further assume that [D.sub.0] = 1 - [D.sub.1], which may be understood as a covered market assumption: if heterogeneous consumers have unit demands, each consumer must buy either product (and total demand is normalized to 1). Production costs are assumed to be zero.
Assume that [D.sub.1](0) = [D.sub.0](0) = 1/2 so that, if Q = and firms charge the same price, they share demand equally. Thus, Q = may be viewed as a symmetric case, whereas when Q [not equal to] 0, one firm has a competitive advantage over the other one in the sense that it may charge a larger price than its competitor and still serve at least half the market (firm 1 having the competitive advantage if Q > 0). This competitive advantage is presented for the exposition as a quality difference, but could also be (and is formally equivalent to) a marginal production cost...
|
