only holds for the primary market, as of course, advertisements are excludable and rival products. Other mass media can be described as network goods. In this connection,the Internet should be beyond dispute, because of its physical net- work properties, but also newspapers, magazines or television programmes can be considered as some kind of network products, namely in the sense of social networks. Of course, media products are also frequently characterised by price discrimination. Newsstand prices (i.e. of newspapers or magazines) and sub- scription rates are typically differentiated. The same is true for Internet portals or for Pay Tv programmes Finally, also regulation of mass media is an important feature. Because of the existence of economic factors like scale economies, barriers to entry and relatively high fixed costs, but also for political reasons the media sector is usually(still) heavily regulated. Some other features and aspects of mass media will not be considered in the following, even if they are interesting fields of economic studies Therefore, topics such as the convergence of the media(see e.g., Nilsson, Nulden and Olsson 2001) or the essential facilities characteristics of mass media will not be discussed. 2 However, the focus of this paper is basically on the pricing behaviour inin- terrelated markets and related topics The paper is organised as follows: In the next section theoretical consider ations of media markets are reviewed and in the fourth section some empirical work on interrelated markets and the peculiarities of mass media is discussed The last section concludes
only holds for the primary market, as of course, advertisements are excludable and rival products. Other mass media can be described as network goods. In this connection, the Internet should be beyond dispute, because of its physical network properties, but also newspapers, magazines or television programmes can be considered as some kind of network products, namely in the sense of social networks. Of course, media products are also frequently characterised by price discrimination. Newsstand prices (i.e. of newspapers or magazines) and subscription rates are typically differentiated. The same is true for Internet portals or for Pay TV programmes. Finally, also regulation of mass media is an important feature. Because of the existence of economic factors like scale economies, barriers to entry and relatively high fixed costs, but also for political reasons the media sector is usually (still) heavily regulated. Some other features and aspects of mass media will not be considered in the following, even if they are interesting fields of economic studies. Therefore, topics such as the convergence of the media (see e.g., Nilsson, Nuld´en and Olsson 2001) or the essential facilities characteristics of mass media will not be discussed.2 However, the focus of this paper is basically on the pricing behaviour in interrelated markets and related topics. The paper is organised as follows: In the next section theoretical considerations of media markets are reviewed and in the fourth section some empirical work on interrelated markets and the peculiarities of mass media is discussed. The last section concludes. 5
2 heoretical considerations The literature on the interrelationship of markets is strongly connected with media markets of several kinds. Most theoretical and empirical studies deal with the newspaper sector and focus on United States, Great Britain, and Australia While some others concern media like television or radio, but only few of them investigate the interdependency of the markets. For this reason most of the literature surveyed in this paper is related to newspaper advertising and reader markets Theoretical considerations of interrelated markets are rare. Only few studies deal with the interdependency of reader and advertising markets or the influence of different cost structures. Moreover, none of these articles analyse features like quality provision or habit effectss. Only few analyse separate(for example oligopolistic) market structures on advertising and reader markets. Actually, most studies consider monopolistic media firms, which is surely an appropriate and realistic assumption for most reader markets but(also due to the possibility of inter media competition) not necessarily for advertising market As stated above, studies on(inter)related markets are mostly concerned with traditional media like newspapers or magazines. Correspondingly, the first arti cle of this kind by Corden(1952) dealing with demand dependency, analyses the relationship between newspapers'advertising and reader markets. Corden con- siders the optimisation problem of a newspaper firm using geometric techniques within a static framework. However. Corden does not consider interrelated. but only related markets. On the one hand, he asserts that circulation is an impor tant variable for both markets, but on the other hand he neglects to consider the influence of advertising on the demand for copies, even though he is aware of this
2 Theoretical considerations The literature on the interrelationship of markets is strongly connected with media markets of several kinds. Most theoretical and empirical studies deal with the newspaper sector and focus on United States, Great Britain, and Australia. While some others concern media like television or radio, but only few of them investigate the interdependency of the markets. For this reason most of the literature surveyed in this paper is related to newspaper advertising and reader markets. Theoretical considerations of interrelated markets are rare. Only few studies deal with the interdependency of reader and advertising markets or the influence of different cost structures. Moreover, none of these articles analyse features like quality provision or habit effectss. Only few analyse separate (for example oligopolistic) market structures on advertising and reader markets. Actually, most studies consider monopolistic media firms, which is surely an appropriate and realistic assumption for most reader markets but (also due to the possibility of inter media competition) not necessarily for advertising markets. As stated above, studies on (inter)related markets are mostly concerned with traditional media like newspapers or magazines. Correspondingly, the first article of this kind by Corden (1952) dealing with demand dependency, analyses the relationship between newspapers’ advertising and reader markets. Corden considers the optimisation problem of a newspaper firm using geometric techniques within a static framework. However, Corden does not consider interrelated, but only related markets.3 On the one hand, he asserts that circulation is an important variable for both markets, but on the other hand he neglects to consider the influence of advertising on the demand for copies, even though he is aware of this 6
connection. It will be noted that the circulation is the link between the two prod ucts sold by the newspaper-printed matter for readers and space for advertisers. The circulation is the quantity sold of one of the prod ucts and determines the quality of the other A link in the reverse direction is also possible. "/.We shall assume the absence of this reverse link, as it is generally not sufficiently important to jus- tify the complexities with its introduction into the analysis would 2700e W.M. CORDEN(1952), p. 182 Furthermore, Corden distinguishes four different types of costs: fixed costs of the plant, buildings and equipment, costs concerning the editorial, costs that vary with circulation, and the costs of advertising. The copy price is assumed to be fixed, and editorial quality increases both, the demand for copies and costs. Total average costs are assumed to fall with increasing circulation. Corden, therefore considers economies of scale in the newspaper production The results from his analysis are as follows: At first two products are identi- fied, advertising and copies. Introducing a third product, classified advertising an interrelationship of these markets is conceded. Moreover. as circulation can only be raised by increasing the editorial quality and increasing quality will lead to an increase in average costs. Increasing the circulation could, therefore, lead to losses from the reader market, if the effects of a variation in quality raises costs respectively stronger than revenues On the other side, revenues from the advertising market are directly linked to circulation. And the higher the circulation the higher the revenues from adver tising markets. Nevertheless, the advertising space and, therefore, revenues from dvertising markets are assumed to increase less than proportionally with in- creasing circulation. Furthermore, costs for advertising vary proportionally with
connection: “It will be noted that the circulation is the link between the two products sold by the newspaper–printed matter for readers and space for advertisers. The circulation is the quantity sold of one of the products and determines the quality of the other. A link in the reverse direction is also possible.” [. . . ] “We shall assume the absence of this reverse link, as it is generally not sufficiently important to justify the complexities with its introduction into the analysis would involve.” W.M. Corden (1952), p. 182 Furthermore, Corden distinguishes four different types of costs: fixed costs of the plant, buildings and equipment, costs concerning the editorial, costs that vary with circulation, and the costs of advertising. The copy price is assumed to be fixed, and editorial quality increases both, the demand for copies and costs. Total average costs are assumed to fall with increasing circulation. Corden, therefore, considers economies of scale in the newspaper production. The results from his analysis are as follows: At first two products are identi- fied, advertising and copies. Introducing a third product, classified advertising, an interrelationship of these markets is conceded. Moreover, as circulation can only be raised by increasing the editorial quality and increasing quality will lead to an increase in average costs. Increasing the circulation could, therefore, lead to losses from the reader market, if the effects of a variation in quality raises costs respectively stronger than revenues. On the other side, revenues from the advertising market are directly linked to circulation. And the higher the circulation the higher the revenues from advertising markets. Nevertheless, the advertising space and, therefore, revenues from advertising markets are assumed to increase less than proportionally with increasing circulation.4 Furthermore, costs for advertising vary proportionally with 7
advertising space and circulation. Thus, there is a maximum of advertising space optimising profits from the advertising market Overall. taking both markets into account. Corden concludes that a normal situation is reached, if reader markets are characterised by losses and advertising markets are highly profitable. This situation is due to of the high circulation generated in the reader market which is, of course, higher than the optimal circu- lation considering only the reader market, since a higher number of copies raises the demand for advertising space. Due to a less than proportional effect from circulation on advertising, circulation and editorial quality is naturally restricted A full model of a newspaper monopolist's pricing decision is offered by blair and Romano(1992). Starting from two interrelated demands for copies and ad- vertising, profit maximising pricing rules for each market are derived. In contrast to a standard monopoly price, the(feedback)effect from the respectively opposite market has to be taken into account. Assuming a positive influence of both, the demand for advertising on the demand for copies and vice versa, price decreases on both markets lead to a stronger response in respective quantities in comparison to usual monopolistic situations. Decreasing the copy price, for example, leads to an increase in the demand for copies and, therefore to a rise in the demand for advertising space. The increased advertising volume, in turn, leads to a further increase in copy demand. These direct and indirect effects can be summarised as the phenomenon of the circulation-advertising spiral. Furthermore, because of this spiral the equilibrium copy price is lower than a usual monopoly price and possibly even below marginal cost. Finally, Blair and Romano conclude that vertical integration of newspaper publishers and distributers are a typical conse- quence of this interrelationship because costs can be decreased and copy prices e set optimally
advertising space and circulation. Thus, there is a maximum of advertising space optimising profits from the advertising market. Overall, taking both markets into account, Corden concludes that a ‘normal’ situation is reached, if reader markets are characterised by losses and advertising markets are highly profitable.5 This situation is due to of the high circulation generated in the reader market which is, of course, higher than the optimal circulation considering only the reader market, since a higher number of copies raises the demand for advertising space. Due to a less than proportional effect from circulation on advertising, circulation and editorial quality is naturally restricted. A full model of a newspaper monopolist’s pricing decision is offered by Blair and Romano (1992). Starting from two interrelated demands for copies and advertising, profit maximising pricing rules for each market are derived. In contrast to a standard monopoly price, the (feedback) effect from the respectively opposite market has to be taken into account. Assuming a positive influence of both, the demand for advertising on the demand for copies and vice versa, price decreases on both markets lead to a stronger response in respective quantities in comparison to usual monopolistic situations. Decreasing the copy price, for example, leads to an increase in the demand for copies and, therefore, to a rise in the demand for advertising space. The increased advertising volume, in turn, leads to a further increase in copy demand. These direct and indirect effects can be summarised as the phenomenon of the circulation-advertising spiral. Furthermore, because of this spiral the equilibrium copy price is lower than a usual monopoly price and possibly even below marginal cost. Finally, Blair and Romano conclude that vertical integration of newspaper publishers and distributers are a typical consequence of this interrelationship because costs can be decreased and copy prices be set optimally. 8
In contrast to Blair and Romano, Chaudhri(1998) models reader and adver tising markets in a different manner. Instead of interrelated markets, he first assumes related markets where the demand for advertising depends on the de- mand for copies(simply by multiplying the revenues for advertising with the circulation)but not vice versa. At first a "naive approach is analysed, where the newspaper firm is a monopolist on the reader market but the advertising market is of polypolistic structure. Marginal costs are assumed to be constant. Not sur prisingly, Chaudhri finds that the price cost margin is smaller than for a usual monopoly. In a second model, the author analyses competitive situations in both markets, which results in a larger equilibrium circulation. Introducing a new cost function, where costs are the product of circulation and the square of advertising Chaudhri calculates a zero copy price. Furthermore, he concludes that even monopolist in both markets would set copy prices below marginal costs. And under the assumption that price elasticities in both markets are larger than one in absolute values, the newspaper monopolist would set the copy price equal to zero. as in the competitive case Finally, the d by the interrelationship of the marke Chaudhri concludes that, first, in a competitive situation in both markets, the newspaper firm sells the usual quantities according to the price equals marginal cost rule. Secondly, in a situation where newspaper markets are monopoli- tic but advertising markets are competitive, the copy price is not only above marginal costs but also corresponds to the standard monopolist pricing rule. Un- fortunately, the author neglects to analyse more realistic market structures for dvertising markets like duopolies or oligopolies An interesting and methodically different study on interrelated markets provided by Baye and Morgan(2000). The authors consider a media firm, i.e
In contrast to Blair and Romano, Chaudhri (1998) models reader and advertising markets in a different manner. Instead of interrelated markets, he first assumes related markets where the demand for advertising depends on the demand for copies (simply by multiplying the revenues for advertising with the circulation) but not vice versa. At first a ‘naive’ approach is analysed, where the newspaper firm is a monopolist on the reader market but the advertising market is of polypolistic structure. Marginal costs are assumed to be constant. Not surprisingly, Chaudhri finds that the price cost margin is smaller than for a usual monopoly. In a second model, the author analyses competitive situations in both markets, which results in a larger equilibrium circulation. Introducing a new cost function, where costs are the product of circulation and the square of advertising, Chaudhri calculates a zero copy price. Furthermore, he concludes that even a monopolist in both markets would set copy prices below marginal costs. And, under the assumption that price elasticities in both markets are larger than one in absolute values, the newspaper monopolist would set the copy price equal to zero, as in the competitive case. Finally, the model is augmented by the interrelationship of the markets. Chaudhri concludes that, first, in a competitive situation in both markets, the newspaper firm sells the usual quantities according to the price equals marginal cost rule. Secondly, in a situation where newspaper markets are monopolistic but advertising markets are competitive, the copy price is not only above marginal costs but also corresponds to the standard monopolist pricing rule. Unfortunately, the author neglects to analyse more realistic market structures for advertising markets like duopolies or oligopolies. An interesting and methodically different study on interrelated markets is provided by Baye and Morgan (2000). The authors consider a media firm, i.e. 9