Generic product advertising, spillovers, and market concentration.

This article examines market outcomes when firms produce a product, such as an agricultural commodity, which is sufficiently homogenous that advertising expenditures by one firm enhances the demand facing all firms. The potential free-rider problem that arises in such commodity markets has led to the creation of a large number of government-sponsored generic advertising campaigns financed by compulsory contributions from industry members. There are currently over a dozen federal and perhaps as many as fifty state programs of this kind in the United States alone. However, these programs have been highly contentious, as reflected in numerous court cases. This includes two recent decisions that upheld both the imposition of a $1 fee on all sales and importation of cattle to fund the "Beef: It's What's for Dinner" advertising effort, and a similar compulsory program that funds the "Got Milk" campaign. (1)

Not surprisingly, generic advertising has also been the focus of much economic analysis, both theoretical and empirical. (2) Our article contributes to the theoretical literature, but it differs from earlier work in the way generic advertising is both produced and consumed. On the supply side there is a well-established tradition in economics of treating advertising as a quantity of homogeneous messages, which can be purchased by the firm in a competitive market at marginal cost. The assumption that the advertising industry is competitive is reasonable in so far as there are many advertising agencies competing for clients. However advertising is a "creative industry" (Caves 2000), and in a creative industry there are important advantages to having more than one supplier of advertising messages for a given commodity. In particular, the greater the number of agencies working on an advertising campaign for a given product the greater is the likelihood that at least one of them will be successful. Advertising is a highly differentiated service but advertising models typically ignore this fact along with the industry's creative nature. Yet these features play an important role, particularly in the context of free-riding and spillovers, and we incorporate them in our model.

Similarly, on the product demand side we consider different ways advertising can affect consumer willingness to pay without changing the underlying consumer preferences for the product. In the same way that consumers place a greater value on lodging accommodations the more beautiful the surrounding landscape, consumers also place greater value on a product the greater is the advertising of that product. Advertising is viewed as a complement to the product being advertised and as a complementary good can increase the overall utility a consumer derives from consuming an advertised product.

Our analysis provides insights into the conditions necessary for any generic advertising to occur as well as to why producers do not always favor compulsory advertising programs even when the programs are introduced to correct an industry profit-reducing inefficiency. This paradoxical state of affairs is well noted in Crespi and Marette (2002) and Crespi (2003). In addition, our model provides theoretical support for a well-known stylized fact in industrial organization that expenditures on advertising initially increase with market concentration but subsequently decline. (3)

We analyze advertising in a homogenous product market in the framework of a two-stage game. In the first stage each firm chooses whether or not to launch a costly advertising campaign. The effect of one firm's advertising spills over perfectly to other firms competing in the market. This creates a tension. On the one hand, the greater the number of firms who advertise the greater is the incentive to free-ride. On the other, the greater the number of firms that advertise the more likely it is that one of the campaigns will be popular and successful.

In the second stage, after marketing campaigns have been launched, firms compete in quantities. We solve for a symmetric perfect equilibrium and identify the critical degree of concentration above which firms will themselves undertake costly advertising despite free-rider problems. We also show that in less concentrated markets firms fail to promote their products and as a result, a compulsory program to fund such advertising can be welfare-improving. While the welfare implications are complex, a sufficient condition for such programs to raise social welfare is that they do not induce any exit. Even when this is the case, however, we show that compulsory programs will not always be supported by incumbent producers.

Advertising and Consumer Demand

Firms in homogenous product markets, just like firms in any product market, advertise in order to raise consumer demand for their products. Yet, why does advertising have this effect? The answer to this question has been long debated. In this article, we rely on the basic insight of Becker and Murphy (1993) in modeling the mechanism by which advertising increases consumer willingness to pay. Specifically, we treat advertising as a complement to the product being marketed that affects the demand for goods without changing the underlying preferences of consumers. Consumers may place a greater value on a product the greater is the advertising of that product because they enjoy knowing that the product they buy is widely recognized by lots of others on television, in the movies, and on billboards. For example, consumers may gain from knowing that when they serve their friends fresh fish for dinner their friends will know through advertising that it is "in" to eat fresh fish. Advertising raises consumer willingness to pay because it adds recognition and prestige to the basic product. In this case, advertising "builds value," and although advertising responds to, rather than changes consumer preferences, this approach is similar in spirit to the earlier persuasive view of advertising (Kaldor 1950).

There is another way that advertising can serve as a complement to the basic good. Advertising can inform consumers either about the product's existence or how to use the product more effectively. In the same way that hotel advertising may include information about nearby ski or water recreation facilities and interest more consumers in coming to the hotel, agricultural producers can attract more consumers to their product by telling them about the nutritional content and providing recipes or ideas about how to consume the product. In this case, the complementary function of advertising is informational. Advertising informs consumers how to extract utility from the product, either by telling them of its existence in the first place or, instead, how properly to make use of it. Advertising works in this context to extend the appeal or "reach" of the product. This "extending reach" view is similar though not quite identical to the informative view of advertising associated with Telser (1964). We now consider more precisely how each kind of advertising, "building value" or "extending reach" affects consumer demand.

Case 1: Advertising to Build Value

When advertising builds value, we model demand as

(1) Q(P, [alpha]) = (1 - p/v([alpha]) L

where L is a size parameter reflecting the maximum number of units sold when price is zero and v([alpha]) is a shift function indicating how demand responds to advertising efforts [alpha], with v(0) = 1. Equation (1) reflects our assumption that the effect of advertising that builds value is to increase consumer's utility or willingness to pay multiplicatively by a factor, v([alpha]). Note that the inverse demand curve is: P(Q, [alpha]) = v([alpha])(1 - 1/L Q). As the level of advertising [alpha] increases, the inverse demand curve rotates upward as illustrated in figure l(a).

[FIGURE 1 OMITTED]

Case 2: Advertising to Extend Reach

When advertising extends reach, we model demand as

(2) [Q.sup.D](P, [alpha]) = g([alpha])(1 - P)L

where L is a market size parameter when there is zero advertising and g([alpha]) is a shift function that reflects the market demand effects of advertising, with g(0) = 1. The inverse demand curve for this case is P(Q, [alpha]) = (1 - [1/g([alpha])L] Q). As the level of advertising increases the demand curve rotates outward as illustrated in figure 1(b).

Suppose now that there are N identical firms that compete in quantities in the market for this product. Each firm has a constant unit cost of production c, which we normalize to zero. There are no fixed costs of production. For a given level of [alpha], it is straightforward to show that under building value, Case 1, the equilibrium values for price P, firm output q, and firm profit [pi] are

(3) p = v([alpha])/(N + 1); q = L/(N + 1); [pi] = v([alpha])/[(N + 1).sup.2] L.

By contrast, under extending reach, Case 2, the equilibrium values for price P, firm output q, and firm profit [pi] are

(4) P = 1/(N + 1); q = g([alpha])L/(N + 1); [pi] = g([alpha])/[(N + 1).sup.2] L.

Note that our assumption of a constant unit cost c implies that a change in advertising will alter output but not price in the extending reach case. On the other hand, the effect of advertising in the building value case affects price but not output. Despite this difference, the relationship between the industry advertising level and a firm's profit is qualitatively the same regardless of whether advertising builds value or extends reach. The real difference between the two cases lies in the implications for strategic interaction among firms and how their individual advertising efforts interact to make advertising successful.

Advertising Campaigns