Superstars and fairness: Let the sky fall

By Guillaume Allègre

Are actors overpaid? A column by Vincent Maraval has launched a debate that is in essence ideological … in a good way. Indeed, it seems proper that high incomes need to be justified based on arguments that can convince the largest number of people. Pay levels cannot be  fair unless they are publicly defensible. In this spirit, by drawing on an analysis of the economics of superstars, this post supports the idea that a small number of actors, and of artists in general, receive collectively constructed income, which justifies an intervention that is designed to reduce income inequalities.

How do you explain the huge revenues of a handful of singers or actors while most artists struggle to make a living from their work? The superstar effect has been analyzed by economists based on a seminal paper by Rosen (The Economics of Superstars). It is related to the structure of demand (by nature gregarious), technology (which allows broadcasting productions at a low cost) and the legal environment (which can be used to exclude stowaways or freeriders). To this we must add that the special position of entertainment stars allows them to capture a large share of the collectively constructed income. These characteristics justify a high level of income redistribution. It does not seem that this matter can be resolved satisfactorily just through the tax system: a tax rate of 75%, which is already viewed as confiscation, is not enough in an economy where superstars can earn 100 times as much as the average income – not to mention the risk of tax exile. Intervening directly in the institutional environment and on pay, especially for projects that receive public funds or assistance thus seems legitimate.

Is it fair that artists who are successful are subject, for example, to a tax rate of 75% for incomes of over a million euros, or is this just confiscatory? If this question can be raised for all activities, entertainment (artistic or sports) can be considered as a case study, because there is little doubt about spectators’ willingness to pay, and there is no information asymmetry or principal-agent problems. The issue of the compensation of artists does not arise in the same way as, for example, business leaders who are engaged in activities whose contribution is difficult to estimate (i.e. their marginal productivity), and who can exercise control over the committees that set their pay: are the company’s good results due to chance, to the work of the CEO, to the entire management team or to the effort of all the employees? Does the CEO’s salary depend on their contribution or on their ability to convince the remunerations committee of their value? In a recent note, Galbraith makes a distinction between the case of Depardieu and that of business leaders [i]. In his book Anarchy, State and Utopia, Robert Nozick uses the transparency of the entertainment industry  and takes the example of a superstar of the time, the basketball player Wilt Chamberlain (the book was published in 1974), to justify very high incomes. His argument is famous: if a million people are willing to pay 25 cents extra to see Wilt Chamberlain play, and he signs a contract with a basketball team offering to pay him 25 cents per ticket, giving him an income of $250,000, which is well above the median or average income at the time, is fair and legitimate. Redistributing this income would be immoral; freely consented inequalities must be respected.

But how do we explain the fact that many individuals are willing to pay so much for a particular artist, and nothing for most of the others? Where does  this winner-take-all characteristic comes from? In other markets, if a service is perceived as having a slightly lower quality, many buyers are willing to buy it for a bit less than the high-quality service, so that a range of quality levels co-exist. In contrast, in the world of entertainment, a relative handful earn astronomical sums. In a seminal article, The Economics of Superstars(1981), Rosen explains this phenomenon by the structure of demand and production technology. What matters is that lesser talents cannot easily replace the greater talents (people would rather watch one top show than ten mediocre shows) and that the cost of production does not increase in proportion to the quantity supplied (the effort is the same whether 10 or 1000 people are in the audience or buy the book). In fact, according to Rosen, it is technology (especially television) which explains the sharp increase in the income of superstars. He concludes his article with these words: “What changes in the future will be wrought by cable, videocassettes, and home computers?”

The high incomes of superstars are not simply the result of a slightly superior talent (as in the model proposed by Rosen). The studios are not simply paying for the incomparable talent of the actors in Friends or of Depardieu in Astérix and Obelix at the Olympic Games. Some actors actually manage to capture an income that is built in part by them but also by chance and by the behind-the-scenes work of many other contributors to the entertainment economy. The actors in Friends were able to negotiate significant wage increases as the series was renewed again and again. While under the season 1 contract each actor was paid 22,500 dollars per episode, they received $75,000 per episode in season 3, $100,000 in the fifth, $125,000 in the sixth, $750,000 in seasons seven and eight and one million dollars for the last two seasons, more than 40 times higher than in the first season, whereas the audience only doubled between the first and last season (source: Wikipédia). In season 2, the pay rates were negotiated individually, but the actors, including Jennifer Anniston and David Schwimmer, whose wages were well above those of the rest of the group, quickly realized the importance of collective bargaining: while the studio might manage to dispense with any individual actor (by replacing them or killing off their character), they could not replace the entire cast. Clearly, the 40-fold increase in income is not due to any exponential increase in the actors’ talents, but to the fact that they have benefited from the commitment of the spectators to the series, a commitment that was forged by the actors but also by the work of the scriptwriters, designers, and directors in the early seasons. Because they embody the series and bargained collectively, the actors in Friends managed to capture for themselves an economic rent that was collectively constructed.

Similarly, if Depardieu has succeeded in establishing himself as a national figure, it is partly due to his talent but also due to the work of the many directors who have used him (and their scriptwriters, etc.). While it is difficult to explain the success of any particular cultural product, the element of chance or luck should not be overlooked. This is related to one of the characteristics of cultural products: they are generally more appreciated when the experience is shared because, as André Gunthert emphasises, cultural consumption gains meaning through its socialization (conversation, judgment, citation, re-use). Success thus breeds success, which explains phenomena of the type seen with the films Les Intouchables and Bienvenue chez les Chtis, whose success cannot be explained solely by the quality of the films. If in Rosen’s model, we replace talent by audience (people prefer to watch one show with a large audience rather than ten shows with small audiences), a small initial advantage, which may be due to chance, but not only to chance, can turn into a phenomenon due to a snowball effect (Adler, 1985). In addition, the fact that televisions require stars to get co-financing for films, as Maraval explains, shows why celebrity is self-reinforcing and leads to a concentration of wealth in the hands of a few very well-known actors. The small initial advantage in terms of reputation is not necessarily due to pure chance, as can be seen by observing the number of sons and daughters in the profession, including the offspring of producers and directors. Stardom is also a status where you can enjoy a reputation that is “ill-gotten” and where negative buzz also provides visibility.

For the superstar effect to be converted into a high income, artists need to be able to exercise their intellectual property rights and exclude freeriders. The artists need a legal environment that legally recognizes and enforces their intellectual property rights [ii]. The fact that actors can capture a significant share of the income is partly a consequence of incomplete contracts and asymmetries in legislation on intellectual property. For example, California law prohibits contracts with terms of over seven years, which explains the jump in the remuneration of actors for series with long runs. Actors can also always threaten to quit, which constitutes a credible threat if they have gained enough reputation. The studios cannot contractually retain the anticipated benefits of this reputation. The actors also benefit from the fact that other people cannot or can no longer assert their right to intellectual property. Patents on television broadcasting technology have long been in the public domain: industrial property rights are much shorter (maximum 20 years) than artistic copyrights (70 years after the author’s death in France and the United States).A certain number of ideas that contribute to the success of cultural products (films, series, etc.) are not copyrightable: we do not copyright a joke, a story, a way of filming or editing, or a concept or idea for a scenario. The fact that some players in the entertainment and cultural industry can capture an income is therefore not merely the natural consequence of differences in talent or an objective way of measuring the contribution of each, but flows largely from the specific provisions governing intellectual property rights that establish what is copyrightable or not, along with the duration of the protection. It is not at all clear, for example, that we should give celebrities the exclusive right to commercially use their public image (see Madow, 1993).

In addition to the protection of intellectual property rights, government intervention in the film industry can be considered to be massive (whether in the form of subsidies or regulations): investment quotas in the production and broadcasting of French-language cinematographic works for TV channels; the artist  unemployment scheme, whose  deficit is financed out of general taxation; tax incentives (SOFICA, tax credits); reduced VAT; aid from local authorities (regional, departmental and municipal) for filming, festivals and local cinemas ; and the financing of the CNC (mostly from industry revenues and already partly redistributive). Moreover, Coq et al. (2006) show that changes in regulations, which have favoured the goal of defending the market share of domestic films rather than pluralistic creation within the country, have led to a greater concentration of resources for expensive films, while the requirements placed on television exacerbate the superstar effect, as the networks are fond of stars.

From an economic viewpoint, two arguments thus justify redistribution of income: the capture of an economic rent constructed by many individuals and the element of chance (to which should be added the weight of public intervention in the cinema). In the presence of chance or risk, redistribution plays the role of an insurance, which can increase both the equity and efficiency of the system. From the viewpoint of equity, before the winners are revealed, risk-adverse individuals would be willing to socialize the risky gains. From the viewpoint of efficiency, too much risk leads to underinvestment on the part of very talented individuals who do not want to engage in an activity where there are too few chosen (and where they have too few connections). From the viewpoint of both equity and efficiency, the structure of the entertainment economy justifies a significant level of redistribution. This redistribution can take several forms: (1) universal taxation coupled with sector subsidies, (2) insurance, for example, based on the specific status of the entertainers, (3) minimum and / or maximum wages, in particular for projects receiving public funding or support (France Television, Regional Councils, etc. [iii]). Economists generally prefer the method of taxes or social insurance over direct interventions on wages, leaving the market to operate freely before redistributing income. The tax system also helps to avoid the arbitrary effects of thresholds when setting a maximum wage. However, in practice, fiscal redistribution faces a major limitation: once gross salaries are determined by the interaction between market forces and the institutional environment, they are generally considered legitimate; a high tax rate, e.g. 75%, may then be regarded as confiscatory, or as representing an “undue burden”, in the words of a recent decision of the Constitutional Council, even though such rates could clearly be insufficient to reduce the inequalities in a superstar economy where income differentials can reach ratios of 1 to 100. Reducing inequalities then requires direct intervention both in the institutional environment – for example, by reducing the duration of intellectual property rights – and on the determination of remuneration, which is all the more justified in a highly regulated and subsidized sector.


[i] “In reality, the case of Depardieu is very different. He grew rich thanks to his talent. This is not the case of business leaders! Their revenue comes from companies that have earned money through a collective effort.” From our perspective, there is a capture of a collectively-constructed  economic rent in both cases.

[ii] In this sense, we must understand the libertarianism of Nozick as the absolute respect for individual property rights (which have a natural character). This is a long way from the libertarian liberalism that seeks to minimize external constraints, since in this case it is necessary for authority to enforce property rights. This explains why a contradictory mix of appeals to freedom and to authoritarianism stems from this doctrine.

[iii] As well as private television channels with respect to their obligations, as they benefit in return from the free use of the broadcast spectrum, which is similar to a public subsidy.