Keywords

1 The Quality Dilemma of the Media

From economic theory it is well known that markets of information goods possess virtually all the economic properties that could cause problems for market transactions (Varian, 1998, 2003). Mass media markets are a perfect example: here (1), increasing returns to scale as well as (2) insufficient protection of intellectual property rights both lead to dramatic difficulties for media competition as well as for the management of media firms (cf. Doyle, 2013). Additionally, media management and media users are confronted with a problem with information following from Arrow’s famous “information paradox” (Arrow, 1971, p. 148). This paradox results from the problem that a consumer of information cannot tell beforehand what information she will actually get for her money. This is because if she already knew, she obviously would not need to get the information anyway.

But if you have to buy something without knowing what it will actually be like you naturally might fear being taken advantage of: the seller might deliver too low a quality for too high a price. And by the time you have reassured yourself about the quality you have received, it will be too late. Now assume there will be no punishment for selling low quality at a high price. Would you buy? You would rather not, even when trading high quality for a suitable price would actually make both of you better off.

Information economics would say that the two of you—the buyer and the seller, the principal and the agent—are caught in a trap, because both of you are confronted with a fundamental problem of moral hazard. You may also call this trap a quality dilemma, as it is usually called in game theory, as the quality trap exactly displays the structure of a well-known game called the one-sided prisoner’s dilemma (Rasmusen, 2006).

Information economics now says that in the presence of moral hazard problems, reputation mechanisms are required to make satisfying market transactions possible (Stiglitz, 2000). As far as mass media markets are concerned, Shapiro and Varian (1999) in their famous bestselling book “Information rules” are all optimistic about that: they suggest that most media producers overcome these problems by building up media reputation (Shapiro & Varian, 1999, pp. 5–6). If this was true then media reputation from the viewpoint of New Institutional Economics would function as a typical institutional market arrangement (cf. Furubotn & Richter, 2005) which can solve the quality dilemma immanent in mass media markets. But is it actually true?

As this article will show it certainly can only be true if media brands’ core function is to bear the quality reputations required to overcome the quality dilemmas of the media. But firstly, let us now take a brief look at the standard economic theory of quality reputation.

2 The Standard Economic Theory of Quality Reputation

The central idea of the standard economic models of quality reputation originates from an influential article by Klein and Leffler published in 1981. Formalized treatments were subsequently given by Shapiro (1983), Allen (1984), Tirole (1988; cf. also Furubotn & Richter, 2005) and Rasmusen (2006). These treatments are too complex to be presented here. However, the gist of the models can be expressed with reference to very basic concepts of dynamic game theory.

Assuming games that are involving infinitely repeated quality dilemmas—which, as remarked before, display the structure of a one-sided prisoner’s dilemma—quality reputations according to these models can rest upon self-enforcing implicit contracts between consumers and sellers. These implicit contracts state that consumers continue to buy the products of a specific seller as long as this seller in the counter-move continues to deliver high quality goods and thereby maintains his reputation as a high quality seller.

If a seller should ever cheat on his customers by delivering only poor quality for a high quality price, the implicit contracts contain the threat that all former customers of the cheating seller will subsequently terminate the business relationship and start buying somewhere else. Using terminology from game theory, customers are expected to follow a trigger, a grim or an exit threat strategy, respectively.

The standard models prove that under certain conditions, subgame perfect reputation equilibria exist, whereby such exit threats can credibly deter all the sellers from cheating. In such states of reputation equilibria the sellers continuously maintain their high quality reputation simply because this is a superior profit maximizing strategy. More precisely: the discounted expected profits from future business with regular customers are bigger than the one-time profit the sellers could get by cheating their customers once.

3 Application Problems and Proposals for Theory Extensions

At first sight it seems obvious that the standard models of quality reputation should also be applicable to mass media markets. After all it is a common feature of mass media products that production and consumption are continuously repeated in a strict periodical sequence. So you find dailies and weeklies in the print and mobile-app media markets as well as in broadcasting schedules, and also in the digital world TV and film entertainment is still dominated by series and sequels. Furthermore, customer relations between media consumers and media brands are typically long lasting, often building on subscriptions that are prolonged virtually ‘forever’ if not canceled at some stage by the customer. And last but not least: people love to talk about what they have seen and heard through the media and do so almost daily. Such word-of-mouth communication between consumers facilitates comparisons between competing offers and this may strengthen exit threats.

And yet there are several problems that preclude a direct application of the standard reputation models to mass media markets. I will now look at these problems of applicability one by one—and for each of them outline how economic reputation theory could be extended in order to solve the problem.

3.1 The Infinity Problem and Media Brands

If actors have finite time horizons, the reputation equilibria of the standard models can be destabilized using the very logical concept of backward induction (cf. Rasmusen, 2006). The models therefore implicitly assume that actors have infinite time horizons. It is all too obvious that this is not the case.

This infinity problem is a general problem for the standard reputation models. Kreps (1990a, 1990b) however, has proven that this problem can be overcome if it is not the producer himself who is the bearer of quality reputation but the tradable—and therefore in principle infinite—firm’s name which is attached to products. If therefore, by selling the firm—i.e. the reputation bearing name—the producer is able to sell future reputation profits at any time, then infinite time horizons indirectly enter the game again.

It is now very plausible to suggest that in mass media markets media brands may function as such tradable—and in principle infinite—reputation bearing names. This would in any case also account for the often considerably large amounts of money that are charged for goodwill in media acquisitions and mergers.

One might now argue that media brands are not immortal. The opposite is true: since competition in many media markets is fierce, many new media brands are doomed to be killed off in their early stages despite having been launched very recently. TV formats are a good example: many of them vanish—almost without having appeared on the screen. And they are then infinitely dead—and not alive. Hence we have to deal with a second difficulty, another infinity problem. But this one is relatively easy to deal with. As Benoit and Krishna (1985, 1987) have generally proven, reputation mechanisms which are stabilized by credible exit threats are also possible, assuming uncertainty about the last period of the repeated game. Such conditions are certainly brought about by dynamic competition in mass media markets.

3.2 The Information Paradox Problem and Media Brand Cultures

In general, the standard reputation models assume complete implicit contracting. Though the contracts need not be explicit they still have to be complete. This means that everyone is supposed to be fully informed about what is promised by implicit contracts, right from the start. This is an elementary assumption of the standard reputation models. Unfortunately, due to Arrow’s information paradox, it cannot be met in mass media markets. Actually, to know precisely what is promised by any information good would mean knowing all the information before it has been produced, delivered and consumed. That would indeed be paradoxical.

Kreps (1990a, 1999) is concerned with a similar problem in his article “Corporate Culture and Economic Theory”—acknowledged by Furubotn and Richter (1997, p. 333) as an essential article for the overall development of the new institutionalist theory of reputation. Kreps examines a crucial issue of New Institutional Economics (NIE), that is, whether reputation mechanisms could work in relational contracts which are necessarily incomplete, not because of an information paradox but as a result of unforeseen contingencies. However, Kreps’ reasoning is also applicable to our problem, as he argues that the incompleteness of the implicit contract does not make reputation building impossible if the implicit contract provides principles that can be generally used as guidelines for the determination of an appropriate standard of quality. Combined with the reputation mechanism by which these principles are enforced, they then form institutional arrangements that are necessary to overcome the moral hazard problems of relational contracts (see also Furubotn & Richer, 1997). According to Kreps, a set of such principles valid for a specific firm is called the firm’s corporate culture.

It is now very straightforward to interpret the creative principles of entertainment genres and TV formats as well as the professional norms of the journalistic press as types of just such institutional cultures. With proper reference to Kreps’ concept of “corporate culture” these sets are indicated here as media brand cultures.

3.3 The Signaling Problem and Media Brand Advertising

The standard reputation models are based on the assumption that all actors involved could perfectly anticipate the firms’ reputation profits. Thus consumers are supposed to be familiar with their cost and profit structures, a severe and unrealistic requirement. Indeed, Klein and Leffler (1981) have already admitted that this is not very likely. More realistically, consumers are usually totally ignorant about the potential reputation profits of firms. Thus reputation mechanisms are bound to fail just because there is no clear cut signal indicating the potential value of reputation. Hence this problem is called the signaling problem.

However in 2001 Rasmusen and Perri extended the Klein-Leffler/Shapiro-reputation models using the tools of game theory, examining whether reputation mechanisms can also work under the more realistic assumption that consumers are not perfectly informed about the cost structures of firms. They discovered that reputation building is still possible if firms credibly signal expected reputation profits by conspicuously incurring sunk costs. Theoretically, there are several ways to reach this goal.Footnote 1 Probably the most obvious one—besides boasting huge editorial departments—is spending on advertisements (cf. Kirmani & Rao, 2000). Apparently this is the method of choice in mass media markets. Massively advertising media brands does not only credibly signal expected reputation gains, but it also serves to communicate the content of the respective media brand’s culture. Unsurprisingly, one key feature of the media industry is spending huge amounts of money on advertising; in fact, there is no other sector that spends more on it in absolute figures or as in relation to return (cf. e.g. Nielsen, 2013).

3.4 The Bundling Problem and Exit Threats by Specialized Audience Segments

Studies in information and digital economics—particularly by Bakos and Brynjolfsson (1999, 2000)—have impressively shown that the bundling of several information goods into one package which is then sold for a single price, can be a very profitable strategy for media firms.Footnote 2 And indeed, you find bundling in media markets wherever it is technically possible: by definition, newspapers and magazines constitute bundles. Cable TV programs are also almost exclusively sold on a subscription basis as program bundles.

The standard reputation models, however, assume that it is not a bundle of different goods, but a single good that is repeatedly sold. But it is certain that bundling would rule out reputation mechanisms if the complexity of the composite good precluded clear exit threats.

There is another plausible yet simple extension of reputation theory that might explain the quality reputations of mass media brands, including where bundling occurs. For this purpose, the heterogeneity of mass media audiences needs to be taken in account: not everyone is equally interested in everything. A media audience might rather be roughly categorized in such a way that for each kind of bundled information goods there are specific audience segments that are (1) particularly interested, or (2) quite, marginally or not interested at all. Now concerning the quality of the respective kind of information good, exit threats could at least originate from those audience segments that are particularly interested in the specific kind of information good. After all, it is reasonable to assume that these audience segments would turn to a competitor who has done better in their particular field of interest. Provided that these specialized audience groups are just big enough to hurt a media firm’s profits significantly by leaving, the overall quality of the entire bundle can thus be assured by efficient reputation mechanisms.

Please note that in this case—and this is a very interesting economic feature of mass media markets—considerable mutual consumption externalities result from the existence of specialized audience segments. Since the specialized audience segments enforce overall quality, everyone benefits from their existence.

3.5 The Journalism Paradox Problem and Exit Threats by Expert Audience Segments

But let us turn to the last problem: it is mainly focused on journalistic media and—with reference to Arrow’s famous information paradox—I call it the journalism paradox problem (cf. Lobigs, 2013, p. 59). Fundamentally, the standard reputation models assume traded goods to be experience goods. Experience goods, according to Nelson (1970), are goods the quality of which cannot be observed by the consumer prior to purchase but only after consumption. Indeed, concluding from Arrow’s paradox, recipients cannot feel secure about mass media quality prior to consumption. But when Shapiro and Varian (1999) state that mass media are generally experience goods they are only partly right: this is true only for entertainment media, it is simply wrong for the general journalistic media.

Shapiro and Varian do not consider an additional paradox as inherent to journalism alone: to assess the quality of journalistic news coverage, media users would not only need to consume the reports themselves but—beyond that—they would first need to acquire an independent knowledge of the objective facts concerning the reports. But if they already knew the facts, why should they then buy and consume the reports at all? People rather typically consult journalistic products to gain information concerning relevant facts that they just would not discover elsewhere. Knowing only the reports however, they cannot fully ascertain the quality of the informational content.

Consequently, because of this journalism paradox, journalistic media in the terminology of information economics do not qualify as experience goods but rather as credence goods. Credence goods are, according to Darby and Karni (1973, pp. 68–69), goods “which cannot be evaluated in normal use. Instead, the assessment of their value requires additional costly information”.

Just ask yourself how costly this additional information will be for the man in the street who wants to assess the quality of the news coverage in his newspaper independently. Costs would obviously be prohibitively high (cf. McManus, 1992).

And simply to rely on media outlets to monitor and control each other for reasons of journalistic competition, ceteris paribus, would not help at all: the competition argument in this case ends up in the well known aporia of an infinite regress suggested in Juvenal’s famous question “Sed quis custodiet ipsos custodes?” (Who guards the guardians?).

But how would reputation ever work if consumers are largely incapable of assessing a product’s quality even after using it, that is, if the product is as such a credence good? Usually this would not be at all possible! The extension of the theory of media brands outlined in this chapter, that eventually also appreciates the credence good characteristic of journalistic media, follows from the line of argument of the preceding subchapter to a large extent.

Again, it starts out from a rather rough categorization of audience segments. Now however, this categorization refers to the different issues that are continuously covered by a journalistic media product. For each of these issues two audience segments are distinguished: the segment of the (1) merely generally interested and the segment of (2) experts respectively.

The generally interested—as the term already suggests—show some general interest in news on the respective issue but they cannot tell if the quality of media coverage is up to the professional norms of journalism. Thus for them journalistic reports on the issue are credence goods.

The experts are confronted with the issue in a professional context. Sooner or later—and very often at the latest in the course of the next working day—they become informed of all the relevant news on the issue independently of news coverage by the media. Nevertheless, they are highly interested in being fully informed about news on ‘their issue’ as soon as possible, and so are also keen for journalistic news on the issue—despite the fact that eventually, they are going to be informed about this news anyway in the course of their professional life. Hence, journalistic reports are for them experience goods.

Since the experts also have good reasons to be particularly interested in the news in question—having up-to-date information is important for their professional careers—we can now refer back to the line of reasoning concerning the enforcement of media reputations by specialized audience segments, as already pointed out above in Sect. 3.4. Thus, mutual consumption externalities could also result from existing specialized audience segments in markets for journalistic media.

To conclude: expert audience segments can effectively enforce the quality promises of journalistic media brands. However, it is not only the experts who benefit from media brand reputation, but also media firms, the entire audience, and, last but not least, society as a whole. However, since the experts essentially do nothing but pursue their own self interest, we witness here a very special manifestation of the invisible hand of the market in its full beneficial sense.

But it is important to note that the market mechanism described above would not work without a professional journalistic culture. The central question is: how can the experts’ control of journalistic quality become representative and effective for the whole audience? The answer being: simply through journalistic norms. The experts’ control is only representative because the norms of objectivity and impartiality are the main professional principles that define journalistic culture and quality in the implicit contracts between journalistic media and their audiences. And it is only the fundamental journalistic norm of topicality that allows for the integration of expert segments into the overall audiences for journalistic quality media.

If journalistic culture is enforced by brand reputation mechanisms, then both components form the institutional arrangement of journalistic media brand reputation. Finally, journalistic media brand reputation plus the institution of market competition constitutes the institution of workable journalistic competition.

Using NIE terminology, this institution of workable journalistic competition overcomes the quality dilemmas in news media markets by reducing the transaction costs of quality journalism:

  • Agreement costs are massively reduced by referring to the traditional professional norms of journalism in the implicit contracts on quality.

  • Monitoring costs are massively reduced as quality controlling by experts is performed quasi en passant in normal professional life.

  • Enforcement costs eventually disappear completely because reputation mechanisms effectively prevent implicit quality contracts ever being violated.

4 Exemplary Application: Journalistic Media Brands in the Digital World

The coming hardships for news media outlets in the digital world have been investigated thoroughly in a number of international reports (e.g. Anderson, Bell, & Shirky, 2012; Currah, 2009; Downie & Schudson, 2009; Kleis Nielsen, 2012; Levy & Kleis Nielsen, 2010). They have almost all focused on two obvious problems for the digital transformation of news media: the implosion of online advertising revenues and the users’ striking lack of willingness to pay for journalistic content provided by online media. Both problems add up to a serious revenue problem for traditional news media, and there is an unprecedented and accelerated consolidation of press publishers and of editorial departments taking place (e.g. for the case of Germany Lobigs, 2014). If one acknowledges that quality news media need substantial economies of scale (because of the high fixed costs of newsrooms), consolidation of the industry seems to be not only indispensable, but also an appropriate way to maintain a quality news supply, although, diversity certainly would suffer (cf. Lobigs, 2014).

In contrast, it has rarely been examined how digitisation and the development of the internet will also affect the efficiency of news media branding. However, building on the theory of news media brands outlined in the preceding subchapters, a rather disconcerting implication can be inferred.

As argued in Sect. 3.5, the supply of quality journalism can be made possible by quality brand reputations if for all relevant fields of news coverage there are not only generally interested members of the audience, but also specific expert audience segments that provide credible exit threats in the case that quality expectations are not met. However, such exit threats can only be effective in the establishment of quality brands if all of the expert groups each represent an amount of revenue which is at least so large that the withdrawal of the group would hurt the publisher economically—even after the deduction of the additional editorial costs due to high quality production in the field of interest.

Formally the following inequality must be fulfilled:

$$ \frac{n_e\left({a}_e+p\right)}{i}>\frac{c_{qe}}{i} $$

According to this inequality the net present value (NPV) of revenues directly attributable to the respective group e of n e experts—expressed here in a very simplified approximate calculation using the NPV formula for the perpetual annuity with i as discount rate—must exceed the NPV of additional editorial costs that are directly attributable to a steady production of high quality journalism in the respective field of interest c qe . Attributable revenues are the advertising value of each expert a e and the annual subscription price p that the experts, like all subscribers, have to pay.

Now let us consider the effects of digitisation on the relevant parameters of the given inequality.

Thinking first back to the ‘old world’ of traditional print media before the rise of the internet it is very plausible to assume that back then the parameters of the left-hand side of the inequality quite easily could take on values sufficiently high to satisfy this central condition for securing effectiveness of high quality brand reputations in all relevant fields of news coverage:

  • Since expert groups for professional purposes depended strongly on high quality newspapers and magazines their willingness to pay substantial subscription prices p was pronounced.

  • For the same reason the potential of expert subscriptions could be exploited almost to the full and, moreover, exit threats were highly credible—which taken together means that values for n e could have been quite large measured against the overall size of the various expert groups.

  • Finally, since advertising customers could reach expert groups not only very precisely, but at the same time also almost exclusively by only a very few high quality print media, values of a e were luxuriously strong. The per capita value alone for job advertisements could often already come close to the distribution price in the ‘good old days’ of traditional providers of high quality journalism media brands.

In the digital online world these ‘good old days’ are unfortunately over. The favourable conditions of the old print world have lost their validity and are replaced by developments that depress the very same parameters that had been pushed up so effectively in the analogue era:

  • Two key challenges from the online world that traditional media face that have been described thoroughly in communication science in recent years are the fragmentation of audiences and the cutting out of news media middleman referred to as disintermediation. The research on both key trends clearly shows that for expert audiences they are particularly severe: especially when it comes to their own field of expertise news media that also cater to the needs of the generally interested public are getting more and more dispensable for more and more professionals from all relevant areas of society. This fact means that those universal journalistic quality media will face a shrinking number of experts n e from whom credible exit threats originate that are apt to enforce implicit contracts on journalistic quality concerning every field of interest e.

  • Along with the quantity component n e the price elements p and a e in the digital world will also deteriorate. The cost per mille (CPM) prices journalistic online offerings can achieve clearly only come to a very small fraction of the comfortable thousand contact prices (TCP) that traditional media were used to in the analogue world. So a e is strongly depressed in the digital world. Concerning the subscription price p the situation is to be considered as similar if not worse. In particular, there seems to be a vanishingly low willingness to pay for journalistic online offerings that address the generally interested in every relevant field of news coverage alongside the respective expert audiences. Noteworthy subscription prices seem to be rather more possible for online offerings that cater to the professional information needs of expert groups exclusively. However, such specialist journalistic offerings obviously do not solve the quality dilemma of journalism for a broader interested public.

Taken together it seems clear that it will get more and more difficult to satisfy the given inequality for whatever field of news coverage, as the digital transformation of the news industry advances in the future. Therefore in coming years journalistic quality brand reputations will only be sustainable because of print revenues attributable to expert audience segments—which are clearly shrinking but which will still remain far more important than the corresponding digital revenues—on the one hand, and any radical forms of editorial cost consolidation that might compensate revenue losses, on the other.

However, in contrast to what one might think at first glance, in this transformation phase of the next few years the consolidation of editorial costs can be a necessary and suitable weapon in defending journalistic quality brand reputations. This is always the case, when the “larger-than relation” between the left-handed and right-handed side of the fundamental inequality on page 380 can be strengthened through the consolidation measures that are taken, be they editorial cooperations, mergers of newsrooms or editorial departments, the multiple exploitation of digital journalistic content, or any other measures that are apt to extend the cost reducing scale effects of journalistic services. For instance in Germany there still seems to be considerable potential for such kinds of cost consolidation (cf. Lobigs, 2014), and therefore potential mergers and acquisitions that could foster them should not be rejected too hastily because of the very traditional doctrine of maintaining as much supplier diversity as possible at no matter what costs. In the end, even a very high diversity of only pseudo-journalistic news offerings will obviously be of no help if reputation mechanisms that allow for the preservation of effective journalistic quality brands are about to break down completely.

5 Conclusion

There are two paradoxes leading to quality dilemmas in mass media markets: Arrow’s information paradox and the journalism paradox. As outlined in this chapter, economic theory can explain how these dilemmas can be overcome by brand reputation mechanisms. To achieve this end, it is essential to extend the standard economic theory of reputation, which is solely based on information economics and game theory, to include straightforward arguments derived from New Institutional Economics. However, the emerging economic theory of media brands does not only provide new basic insights into essential aspects of media branding but also into the economics of media product bundling as well as into the very special economics of journalism.

The probably most interesting economic feature of media branding that is revealed by this rigorous economic reasoning is the utmost importance of mutual consumption externalities between different segments of media audiences. The economic theory of media brands thus also provides a basis for a special theory of media marketing and branding based on media bundling and on audience segmentation. Whereas general marketing theory views market segmentation essentially as a tool for addressing the different consumer segments in a suitable differentiated way by separated product offerings (e.g. Wedel & Kamakura, 2000), media marketing must also take the signaling value and the beneficial consumption externalities of audience segments into account. The theory of media management and media economics has so far tended to neglect the manifold strategic and tactical as well as societal implications of this very special aspect of mass media markets.

The very brief theory application in the preceding subchapter was meant to emphasize the importance of those implications with regard to an exemplary relevant question in the context of media digitisation. By analysing a central prerequisite for effective journalistic quality brands it has been shown that maintaining such quality brands will be sharply aggravated in the digital world of the future. For media policy the analysis suggests that it would be well advised to reconsider the traditional doctrine of diversity maximization without concern for the necessary conditions of effective journalistic quality branding. The old doctrine might be obsolete in the brave new world of the digital media. Rather, the primary focus should be on the question of how effective quality branding of news media could be effectively supported in the future.